From c5c842cb4c81871d760963ca3f078693f9f2da79 Mon Sep 17 00:00:00 2001
From: mattfidler <matthew.fidler@gmail.com>
Date: Tue, 28 May 2024 04:01:40 +0000
Subject: [PATCH] =?UTF-8?q?Deploying=20to=20gh-pages=20from=20@=20nlmixr2/?=
 =?UTF-8?q?rxode2@659749fd56edca5fb76a81bf60b3d01c62ff58b3=20=F0=9F=9A=80?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 ...ng-User-Defined-Functions-into-rxode2.html |   4 +-
 .../figure-html/benmchmark1-1.png             | Bin 40425 -> 43430 bytes
 .../figure-html/convertC-1.png                | Bin 34032 -> 40398 bytes
 articles/Modifying-Models.html                |   2 +-
 articles/rxode2-cmt.html                      |   6 +-
 articles/rxode2-speed.html                    |  40 +--
 .../figure-html/unnamed-chunk-10-1.png        | Bin 35849 -> 36489 bytes
 .../figure-html/unnamed-chunk-12-1.png        | Bin 65510 -> 64330 bytes
 .../figure-html/unnamed-chunk-13-1.png        | Bin 69408 -> 65303 bytes
 articles/rxode2-stiff.html                    |   4 +-
 articles/rxode2-syntax.html                   |   4 +-
 pkgdown.yml                                   |   2 +-
 reference/dot-rxWithWd.html                   |   2 +-
 .../704a78003b911881e13e3865d8be8e09.md5      |   1 +
 reference/myapp/mod1.d/mod1_.c                | 234 +++++++++---------
 reference/myapp/mod1.d/mod1_.o                | Bin 112848 -> 112848 bytes
 reference/myapp/mod1.d/mod1_.so               | Bin 90360 -> 90360 bytes
 reference/myapp/rx_shiny_data.rda             | Bin 5591 -> 5588 bytes
 reference/rxCombineErrorLines.html            |   2 +-
 reference/rxExpandGrid.html                   |  12 +-
 search.json                                   |   2 +-
 21 files changed, 158 insertions(+), 157 deletions(-)
 create mode 100644 reference/myapp/mod1.d/704a78003b911881e13e3865d8be8e09.md5

diff --git a/articles/Integrating-User-Defined-Functions-into-rxode2.html b/articles/Integrating-User-Defined-Functions-into-rxode2.html
index 3d1e47bc1..ba0f0df8a 100644
--- a/articles/Integrating-User-Defined-Functions-into-rxode2.html
+++ b/articles/Integrating-User-Defined-Functions-into-rxode2.html
@@ -253,8 +253,8 @@ <h2 id="r-based-user-functions">R based user functions<a class="anchor" aria-lab
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/print.html" class="external-link">print</a></span><span class="op">(</span><span class="va">mb</span><span class="op">)</span></span>
 <span><span class="co">#&gt; Unit: milliseconds</span></span>
 <span><span class="co">#&gt;      expr      min       lq     mean   median       uq       max neval</span></span>
-<span><span class="co">#&gt;  withoutC 6.724595 6.871821 7.296725 6.924844 7.033558 12.172650   100</span></span>
-<span><span class="co">#&gt;     withC 1.979874 2.033378 2.336496 2.132534 2.630452  6.166024   100</span></span></code></pre></div>
+<span><span class="co">#&gt;  withoutC 6.867416 6.970227 7.706955 7.084244 7.519517 12.342958   100</span></span>
+<span><span class="co">#&gt;     withC 1.961846 2.025228 2.333243 2.122846 2.614116  6.562157   100</span></span></code></pre></div>
 <p>The C version is almost twice as fast as the R version. You may have
 noticed the conversion also created C versions of the first derivative.
 This is done automatically and gives not just C versions of function,
diff --git a/articles/Integrating-User-Defined-Functions-into-rxode2_files/figure-html/benmchmark1-1.png b/articles/Integrating-User-Defined-Functions-into-rxode2_files/figure-html/benmchmark1-1.png
index 484ecebec9355256674c8d499975131239221629..0e96e334c6043866250e238e82c7a000a7b3c75b 100644
GIT binary patch
literal 43430
zcmeFZcRba98$T{3l}b`1t8$DGm6ffC4xuEHT|#!p$R4SXRmaL+3CRd$uarHL>`^w^
zd!Fz0a=P#P&)+}a-+$lF!~Nm*d5_oYdR^mrJ)hU%b4y9?$f47RNJvPIU~kB%kdTns
zlaTDyKd>KuqfOeT1K;*PP>_>>@9@vMbdei;J7{%7!<vMI(F^@=PvFS>BNCD`Bv_fN
zs`gQT`<z~fEPIR*njgn$-oJhV^S<4C-zljRS8rdvzNf-lb{-$cH*_;2D7;S1_5Fhe
zHVe*3RUW=ac6Pl!Lk(;>Y1DkDZ}_o!_iB>eK0!50d+a33p=-WpjtI7lt$Ixz+{e!u
zmaM!wbvLAL(JEplQ0?s8=x-<a5nKVo?gBq3DYow2{b|i=A6h(#J16TLDf&V6yaz4%
z9w&dW7=0(XD?dnvzNe^=@}lojA;<0apdUU2ms6ncx03!pAMwGub+7vu{mPd@3!{-!
zL72Xx)f*id%VfL{bd4WBZu*t3iN|bCKb0~u5jk1x?V>;3efRO>$5_1cFpF36`>NR0
zujHR?r!*fw?ux>srcXySMVnIS>}xmb`I4?tvJry8YWKMc55Ja%Uy`(A2g&dl(~gWL
z3SO~D3?095N_o#!exb5Aijn<CH?<2HSm*4H@s_S>{mOI`$NtC}lT*3Fa<hVb`Ily+
zMwzQEemKrk;@sR<udGM)p<;^-iCDZR<;?jx(sHliH;ib<|Nk26|AM{$-y{C7y6^up
z=l{$}3}vJG&<hUZ@m|?>)7@tM#ln3>Pn+kfnPV;b3Qvk3K_N1IpjhgA_M>_}%l>VR
zFMnSYLwJc6t8s>0Hg1XATGl1l$_6q;tE8&7$fJ<zKRZXNQR?haO4w!|ZIxqo>Igg^
zR`2k~haJ~`SN+joxySmbP+9m@&;8StA?Sv!OCGc$`H$P)kC02QRUB3R3o8*|kG4_S
zSecG2EG+zZ@6$O-u{kNVOr2<q!(?jmLP4_&kGj^A>uib<!;ewYbANWAxaDd{{m51`
zVO?@DmarZhVmZ}c;v_*>Zyx*O%d2E=K7EvVpu0Olbhem}uwqFF4dGHZocfi%J|E~7
z#UAT?d0@Ti>5tSq$u8T0XjLu0+fRs-a8xO~E(9_DnV~a^dNZ&!Dpc?{yWZ|q5mx2S
zognLGaiP3koC4d;x3|_;&;~RIHq^d=<&^DEPM#wt<+YI#>a)3HQ9NH&w>_}kH!u=m
zno;XP&a}o}X6fSMQo7kUI~qB#WjJE|QrLQQ82xx(Q**W35Us(EZSESm^=!%NXzb2J
zEa$4*qL5oqv)j&MYN`W)Aj}aWU$FMacd=yFC39i>G(Elb^yiM5`4V(%u_zezb$gon
znCQT|(#X!8*bUXvweP&H<IPW-gvW-R1~zp@c4*7iuDdQq4*dNki*D+dgqys7{rcs+
z*)z~oXgh_+NWVN>8=|hGA#C`&a^oAj#GA71p0e~_!u9~cCB|_wwQnP8_jQu6)nmvQ
zOpfF67_`H;%v*!x6F*binAUq^30q@}KFZGLu-WE4Uz}l0SOtZ8_z%1HqNTd4>$(VZ
z=(@R#?DP!m*c}!9n<c)ou8^dXQMxh7@Z=j67C%xSHm18XqPx;LMh^Gsb?nm_C&?%n
zjWFxZJ&(O?_qWrJj?XaA%|AoCXtZcx+hU}~OLyxJt;15Q-0E5fs&s>t`*e)QW=c0_
zdR<ogu=tIQ4K0U}5T|L=BXpNBVY;pk_;*@7V{(!kx4*zbJ4%`0CMh}gxoe_dY^Q!{
zIZb~K*(5CJ?kF4;Uur4Y5GY$p;vJc*q*TUpYZvn0d4IDjO_vhxiDRL*k7wL$v~&$R
z$)$$DB3D-6DKN-%YXpyW-k#FU?A3MKUe+kGo7N!|!W|@xZ^|jARjsY9T{qfd)n>{T
zLNedIHyqe{EjgXrhO448!e9k0ddH7+po>?Id(d{LMp_h}S5#EQ;(u3q8GZAn+FY7)
zn>N)|FgG{H4?iv2)DGp<Emba)fE%Rb&cG&ep>Tsm%2ImWM!0mbsjA({^rugszWn&d
zI}Fy;)O2S(mf#TU>H;^o2S^RRurpp5uI<Wy@(^#ZG~P;g#Y*q2sKjPh^Jw{A%A5Fx
zNFiQdT3#LlbR1Gntnh-i-rnNMwxZb0-mybHKhibx7)y6-JG8@a1zMfRmhLOE@YK<c
zvK`xBnYytUI$@lFzVdA8!yoUk6X31*;Wu(jwVsp`_r{5r%fRIZkfKDzN;Y5cI>(m_
zSf-_?HyTMShSzNqb_NJNcXq*zlQ&rJ#>0mXe|^*l!qAD>7zc>X<UgY=G5_<;M|i14
zItZf>%3Ygq-S-e63kYus$QS`z3t?p+wR9aPlE`?vZiCs-=G11~-~?#$5h9G+K4u>u
zZ9f`rv{1tmh`~B8PloU{i{*4)bu)kmPjuxN;SJ{he0SKIuSSpsFw{1<82cAD$5p%n
z4>nmT`jA#mPIq_r7BcO~j^xge<RLAT9d=i6^rLQ>L|d}5GTvt52P-wZ61RP8hPF-X
z_0-Hv<%Jc83NE#e9QIUjxs)yPsJ9X>j`#21$76=7{OiwpF{sQXXp0WI@5Rl3r)9_1
zjo>T_C*D7!MZT^o<U#vEt3a#CJg-l!z%mW2Jd|C}qcGMK9fY})tkekcO|-?Wl6^WU
zIAzIgn;4gsE=+B;#LB0`j3T*%*Jsn!Gj-@BU7hjol9E`<wx=!sZF2XH9?24?3NqSv
zs+=eDN4E=>ekgEtKnNg5@-T%JkbDR$r?_jjINBH@q@I?VTI(<ttv;8qYSa>URe1F6
zLp&z4ctL)&IaYV>99-=_iJ7~0($vhXp~{~=2$LM^>gc#Mej9GE1n&r$(zGZ@DrZH{
zZC4n4!6|6oom;oeF1cPy8s=<e6m6?^RAOxqQ@fW^>@}^f^QXb<yO+wrrJtdqaBJmX
z#3GN3f~-ZxyP;=dYHI!a1+ChQWy$g{(%q<1We1b}ge<~osRXf>4`-4zc%8ZTn-4XY
z<sl*4$)CBoxhIMbz=Puo4;0@@n#msus9P==i_sj!Q)b4;pG}U!$>WFph5BoH3#|Xy
zN*k<utv-ym85}1FBfK?NY5n(8HBIq&Wzl$oFK0)<W3<pUJ#-nkLZ+7T+l%qIhN}W1
z+e>Y2IY!Nx+CSfY7r#>Qo)F!A^(%PwBRFfr#)!KGlAC{XD711mMV`(?QSju);RujF
zzmyB?ZvHxKezAeCc{nLdM;^a4)xlfhw8~@G!j$DQ7%gIJ0=4S?;Ayzjee&$wwUBUO
zYr_%8-ux#J4b~_xIIqoNuy^j<q2a!F>)#nc8aWB|JY3w#Zp3wKZJsQnE2l}4uvND_
zoj0H}c>F&PJBccOHg2R2Q(Nw_*ZLck+=*i2W=POCk6syc$J7=(EF#E3^wnG!QeM?J
z)3N>Lyw!Zgi5VFg3ooCfr>BPtn%}SRMO*2XfTNKCZw9+^t)kQRFkK0l4txwJAgaO>
zMnR0Jm4n5%SVMWV?+eH1m%HzIY;4S7-x?zXK!w5f=WQ$=`hZ7psARmTLo)6GlOG2!
z7By^mOB(i9JW1vJSd5d!i0kCfRI&^dM8=!Dmyq|jCd$#~8_Y&r)W0U5q=GltSeoei
zVy^jbaXhfN@>eIhvYpo#Fxc7t5^JbAw8h4=!hf?WabO5GY(zu^Zop;rB++}OkK4x|
z;ujQD%jlV!Kit!^^Y$gAt3;xEn@2*}3}zI~e9($@7!7|h-Kk#XM;8Qz15+!$+8Z=F
zI{MFSiln@`PoM&z^b>?Baa^tgTT|%duC{k{jQR03-N5gHHL;4ru$OXXNO@I_v67M_
zIAJWV`*HH!e)o?hn^4tn7sNnf@lb%PNV;J&-T4lSqjxjp{7>H<#JiY6X_-Ksox$&~
zFCLqia1s+z56j#&`d>eO;5TQAc(E$#>S0iZVABI&V1*`;sgJuQ1vx3vb-q9-3YxVh
zT<<D&u*MG$2PtP1&)<`K0&qk%W43s)i8KtoKNoFEf3??Id8Ckqv3V}wtn|xa7UuOE
z_+h}M&2=GM#JFnt0^5-V%*WU?APyz+kk%bPFtQ)BWGQa;ElS?*j0fS03%aFZX=>^6
z_`?iWE{Z#?aNFO$^o6pR&#-~R>L6StwYv{nKv`-TqJP?oW=~zzlM1-<#~V|d(<~AB
zzf@xGRZe04b8CH(t0Ota(ZoaH)m_dIJNs8pM;pTbd5b3Ge$RgA#YS$8Yz+1X1l!@$
z{3(Oizl%VWa&^EvZxm?pSRq&~)dLYnpKJB|YwnjXY*^U<hVY1pfNkmJA!dR-cvld{
z0kYFVuXXc3moXpjY(LP=F*ZA61`!%E1*L~JA61;VX{b^D5R)2NdROu9-oIx<S*XId
z`(;j(NE%Mu^cQzEKU_xwwN%gu4o4%$P!Vr1H&}rV6<wDYC?VpqxuS?K1~{ql#Y}_f
z)~^8f&@VQ%71j2P%`TAlFt=Sa{q^zRDII}Ry0qH&RCnb5U|;J>HGe~lc+aKXoq3`A
ze)RT6u_}aOu!T=&WA^Obf1)_WTzE1i@kgd^8KxG+3S?KJ8=ru5E;w94&fEZ?3y%RX
zjx#aYI=1}~z0qmqFY%2hl*OM3ew#n+*4_E(A=kgGDIx;S>WwQXuB87MjjKCPmhreK
zKx8syX}U+Kwx^n34E%@~>>L4bdfyayOM(ITvtc(@XJeVf9br||-8<*|0ehby8uK3P
zi)wYi&26L`d|^5;a5p=Tr@+awfZMRh`f*rcdpeSpW0xk{*s%*+i_J#s3nTwt@)BNB
z4pz-EZcP9#LtFdhVLGVMUz&F<Ljun}7$~`89}yWDgkg`ex&}pvLg(j-IkP>KWlduC
z`~!ZyAIeo0&SjyNoi_iSqx04fb2F;Sm|6&}oCn`u5e@g(2_CqoUvJuzD?}cp){*{Q
zG$bVlm9#?8tn)Q{jQu$vMvR*4BE9C5h2Nu`Ha@;jhociP)>}aO7)$2nTbv<>bX3Ln
zA3pW^<;$0t+8||#a4-_hZmQB7fK;VNDY2$E(!E~9d^1(GZoD;-Yt`T1pXl_RtLSkn
zmRWW`u<R%E---NtOAY+G<-^^yY*;P^8!O?0mhp`CZfu1ox%y6t_{Ml5azvu5tuT7f
z9$ErashS~PvibYx&!4IRk2%8X3#KxPUNIss&!U9Xe#YZz^C`5?Zlb5jotwLM6)GX?
z+^?TMho4g`aoFEpJ}qc=@2JRl90sAig(`YWV#U@QfG|1JEJ<+wm7yIuJZ)_Nr%Zqa
zWhzCBjO->$h{VyEqFJ8Z{8nBH0eY38Im^hOQS=LLdN-xSC8CEwI{BbQdFo<ZIqP)g
z-0#Xex&UQ0@~!mYT#3mkYZX$@{+~bB;`56T(AbzrX2*Up>%y-uj?ohBI(sL?-R@z-
z+e;x+RaI3)AOA~F%G+Y$ck0r)&!0bI(Z$J0NdQ4m^g02U%B5dkUZ%VBKn8En_Fl1o
zu(SE^bx#i#OBJCOsBSq**!qwbDsN~FQbj1wCjs2U;vuC!f`*sqshv*ybmFv80(eQu
z9D@P2%!P(mM~j%LOXLR0x}z^WtWwSRpIeqFpkMMeU15PTiT%;piQMO(A2PzSXP`r1
zfp!s#|EOIg7%0AS@88iL^`JElV`xKru-OL}^7u1}7vlQh0nil?(MO3iJUc1Sbtw_`
z7<hw`x)6tzE~9^&8$<i}ynUzGdbHvEZq6=8B*XNtm9aTNK|z|`z}1^QX76E__H?#?
z0zedgczrCkd9hjYAG8!li~lU?nuBnW`-%+ue*hldiaxa<R=pK5SY-#w3cDfDrYsr#
zzYu7qf=8#TaysnZ9@lpq{;%a`xu{W7)W37nKn6I%BzDD%$<Y`;475k1X=aHU(VXIF
zcO|Gcp^X!3Q(Orw_o)8n!_%Yx4SaDhyU6o+X|MNAt20>qcVF7E<qln5qBqV^!}alS
zkir1yf-nG%P;4CPL#;#>J5V%Um(cNlez*(gR;1ujeo&8-HS;X6cycEDGgBRz|9KQ_
z=V+q!SQDLqi7MWp=ZiTYSSMmHkcIq0x(o2ge^a6!@*^)C<f7!#QVA@Mu)R9gC^W!L
zTo)vJQbu`botmx;`C)29FK7~vT2;V<cHD~ib1p5I=Qp@@0`6udG~QwU^g_h!GFXY$
zlAMvjgW&#dmjUep*S|-zz@vX?7)9SLhyl1sJSYh`sPWfOE;WST>sLG@du(CxFJ8&f
zN>m`{Dv3_#LSAdLY8GE5#{QruvR*wgX?#XMPyM{3tk{vE9aQXcvl99ZqN~kSAVb~-
zD04eemepbN?aQLs(jsE~G5eKM*ccC02WD!C?&i@HLD*HiKLqRMY?&K=7+8h+mq+M{
zcAVQmMpM(&j6@NG1^SGJScLChhm$FOGE{Xp1KdZ0nDQxQ(J${P3A4gUf&T-}av1U#
z`p=`#qB@ck!oYt-b{Uk0Xwf|4s%8VrDP9l%&1uf~&#Guhd8P2w%3_zunVqfh27tHz
zmu7ePJ!sFMrWE%}!-y62i-2A35WQ4Vagc1yttyaNw;<-yLs_D=`<Xpx2MKhS4Q`;$
z7lWvj9KiI%urG!VbC8gln%ecd@t4?lBq8M5lJb_@NzD5wt7brn{MQelM|Ua_cDCD!
zp3Y)wfj0|--O%iIGlSv>ib;CThsekLLE^**sGxIlUnISr0tR|)Zhm1GER`2g+wa+_
z0d621ss}q(E&YB7fQx^L9f%-z?JHwZ)|?%X6}gBR0^MP!4-E=XP<QYKD}TE{hd57k
ztft)?)&@QKaKkm2+DOaNaHpAqWTHhy;l}>&&ks`nr!(t52iw&4J1xLTJm-F=+5h#<
zSuE&3{%mg_`Iyrrk2mZ2aseSr;(G$%J*!sz4r6ySz~=XfXB`b{%<S*Ap}^+fUXXba
zp_y&gl}$A1mIrK~Nj?dmjFcI7K}cI`)J*{T>y~pLw;}S3*s1Pb$Ij2sKm3ha`6ixU
zK;RL?0x`AQL;El%Vdi$Ta@KXVh?t;uub1#+6cvO5g2jVZi$dE#y#5&MHw<xR;6+Ni
zJ9ygfU_?ybXamy#aWb_ATsNlyV-jmVx+M=z0*+XMNX+@L#}GH&(u($X2dJgr=lgps
ziqeU#D7XdrW&N+uZDx9fv6mlK`SA-05ewls00no-!PnA9Yr~CV!ZBUBkMFz(*ZF$}
z+#yaHPD3J&g10#9KNa`}dK5o&rm%YqHpY3ORt)-N;#NdB;WYMsKpuI%U(>SW^X|)~
z<k;p&oezw~y19zjP6lJbq1S=-{@$+N$FYD2Ch;7wplZKadO?9Oav{58@a~qguoLDV
z$&%X>b<4r$@BFvRw2^@&c=4NcW-+)dwG#IaXGu&$Ev?|s%)o}%B}*)))@={h1`&JH
z-McK_LIYW7Hywd_3_QnVdm1tC*kIs^7SI>(#my9qv0(uRhw5%ms{h*+kbQbW+Q3qU
zOz9FfIvl3b^N1W0@xDpm>}CB<y}_WIfox0kCB_%!6j|h2c3qD`rohF8K{Nbk4k(^~
z_n;;LAY~vc7W(Cg>%KI^K4bg<9EW%ubP;YlV{SSHK%l$n>{e-nefF%pUThNOC1<p`
zg*O0?Lqz|sr?{^uL5xeU^jMZ@;c8h!Hi4T={r9>v@Vcu(Y>M(>Jlc2+axNh0h*>92
z9L{$oPWq@YhMMi>4Lk$_tI&Y6^}lD{glAUrlx-~p%O_!KGhLU{iq}UYiB{?60;^C?
zp}m9rfI%$f`<vk^#6vBCLtS}Rwlh_hp)~;g*#AtHs}82tv1+(fZLl`?8;gJc{(UdF
z_P=LBH0$xM|22}-EGYf`lK#h!A1cJXi=zi3o^=ce2L=2`oe~kK#TRztyrmzG+zHs!
zI@;2?7j@f5gnC~C(L|Gv*cxL?uxE)ew9uemg3$ol{Fl|(1tOikJ8S3rV|SL4n|3qc
z$|<x4QSfOl<@wTuI<A80FFnLcfa++u4k#}RBfM?}I6_YBHg}|;Ij0pO2wQYu6`SQ)
z_UK@0r7LH#KYsoU*4<eM`?s^b@Koltp=5i@{NFD@j-V&wK_Ibv_xHj;WL-A-nL^Bk
zZiH1gB2>6F4U2af=h9u5H!<noQJw{FUR<c>B}Q(FAuLDYC*<^-QCBWuvw<Ps8fKRy
zCfQq!=+?gqx*ICH{vIx1s>bBH_M7ON?I6T;Rs+5FrrC8RCuEbEuvWdB?a@wdrNgbH
zTi<oJt^)~$$25yA-h@mxKy=U)StxLOCMPFx-%qNm&bTcE#Trz;JVM;g2O!aJy#}&o
z44MI8KXx{z3gCalC{v~Ppp82N=yhPNN(f0km|6(y1waBPi_t(p3L8Ax1NGizYrbVS
z@WG%bjmt?cyhLPX`6eWh6Mrdm*RByuA>bjtM|CbqQ9A#b1qlAPw`+rB9QTm8v$LG3
zzPrF-kWkL=p3<!gus1<H#HK?ZvkT+bTLWE#0SeD?!fWEvAt1WF(=rM<iBJH7EF73R
zVnBR>ATwrhT=^T0Vj*PHG^3#EHtM;{oxnf4Un8+f7`PBTrm?YceU`9YmJE%=;9a;v
zN)yHca?hgO@2uomw_GJ>#Uiby*QQO8EOXZtAbH05Pk`}|=2Ewfczfs{6UtHhLA?Dc
zB$rw;TCLX@M1GYnr(2HfBEgezMawVnKl7ik({K>p*s(gU>m$yD%>iX#JnT-O3*Ms(
zQ~`7?CNghx>~=UnVvLPOTJqTMJ|PrcaIegD8!3f%XEIAxx`9Y}_iq8`VEs>kB5wd_
zA3Hp~XrEa)`2hm4@ZbI*&8T`<IuFsW=SXhm?DCeSE1xqT0iYyq@D-e(st3?$TY-dm
zN!(loAPJ;lIfHL00%=`Q536)P2U-8_pRCHG^#UDQK~2%3>{#HiLI7-M5MT5LPLLA^
zjN2HJ`$l%4!)yRQI;4f#)DtkXka8K78l<+V&18Z;)es?&M%=?~bnB83=y>Z8DB&1z
z`9`-*OTvZKekaR{H1v8Gz)C6C;Fv%Vvz7Hf9YQoUxY8JiY8>`-a8VX7TH^;m@avmp
zgv}_MA6P6tG&IzCIZgMZ0|1A(y;SylKHLLT?h4d_(RfQd7N6Je$c1W34p2_d!VnR#
z`3!<(WO8b1S2sN7keUDI6>yR?#fUBJ?18n~4Mx|9LY>>i^$Y%%&MZU?0I_R-fv1xa
zy$!7|Dilqr$uUsWU_Yy+t0h~oEGIyh6;p7bd&yPsS~-Z=1>m->AO!8)1C32A5(vAp
zA1Qe3+I{ybqe7br{Hz;cbG^Q7tG+I;&#uRO9YmTakOoW2AHmbgW8f5Xr3C~83jVy}
z!i|KezkfeEP58;HQ+x#pcnA<Y+ZFqqUX@d{1fa56u+nX6UA~7LgVk|aZg<{UjBQTO
z>z~wu$R#l*NBgvZ{9FJD9|r)J*r+q_%y|`TOs0XNr8Or47wSC)?_`uh<SvxinYmF@
zKBK0Fu+k;M$V74&P5!RYpb?fLM>auwThAmcm%ab;MGy;FJad$OU`@w$XR#J-xq<@0
zJ4(oc<5IvU^Arr@d+mq(Ls;lR{pE95m|pK4D}4(+QQ_bVv}tlE*)~$!8_V^ILVdyY
zJVhTD%s~~a*WGT{9qBgBv=u}XJ*o#Gf;~T^rPX(U=WBt!2n6_zg_X8qNhsrssmd;#
z;4_L4koK%_`J4`H{ip?$!Ge)cP25CM7&2un4MNu_0io)#xJy7HIAkvbQ$rw2HKPcp
zw=BNox&O!q=S?KHz<RhPw|-OR0HIGxi*iX31naF!R)gieR+UF~%?Mm5^kh9&fL#nh
zIa}{=+wRyc=9wsL`dbUe_mJdXK*u`wm4Xji*{xW@q8qOMLY@JT76=qVp@@A1dSt=Y
zLOmZSmOnc>XvJws2Bn?gnb{Ddh<AW%>o=CC;MAfTBZbg#0~W868Vmv)VDv}Ox;7(e
z8py9JUJSS!NX6Y1dE5?>AS%&adlDFt42T2g$vd%iC<BJ`8L?rJRtX7!2Wj~{MSruM
zK_FDhv*<gdLqU>d3UoT8s{2Sh<JSDbjVLCykN4hyHwpvDE?O=C85@;4mVhhgzzj-=
z3oCOYEX@uKWCoIS8o|(2e;aBBkk73*7)b=Vt%teQ0@ECXu`HgK-GLbk0m3{p;SJ#6
zi(%S!8b}_0)<H_bsI8!&fKr$ks7el?DlK$A44k&BaS?(w0asfaAz+$qP>W~O5e5B{
z&wg(3FzDRUUPl{&q0n}a)+Ff+LT?Y-+1?yOGX@LLP7Dw1Ob=x1RXoFBks3Ykt2+!S
zx(WL55OmPhOipGMrhD8260~>_n_Mtkx|t0*1ySA?E?jV1*ldSx2r|4es*}*Tj}CfL
zB4gno0u9QxSJD_tT3`?Z@)eplp%=1ffUX&#mCvRfhV38jIxM%VKV4tefxs;WO|bv~
zXAlORmXg@~H^hx0_WdXfiHUFWlDZ#-7HM!Vk`ouob~*@kNMs{yx)CgaSBV6M3OGGO
zXmXLXwg1TJU=ZOkkb#;3@~JNwki<npS6B-?A+AIF>7$*kbqx?vid@Zr_G$y7`tYfX
zXCUYl(1`)eK@3-t%HzB|^iZ_Yo%OLCfUIJ`?(Yt!z@*8{91>(gbh%h-2P-_SK{+;r
z+Ys|@=R{7zsGh3D6Xwgbhq3i@hfXA#i@@UDwyND4Qf?;#$YX9o$!}}SZN~^9o=4D(
zh!)#{;fABEn`%LaYy-y;Ux?(e8~F2`BHML)3X8v^%m~2R1Tap!(NI-Kj_Y>M)d?7U
zgXxADGTvvCISz~Od))}VnA!^R)rQw+*~`YN^Kx?$d`BwYXk#SeV`dhHR)7yZOfSfd
z`~!^u0c-#6i`@W0D-Xs7#8x^VYGfPiVg|9DwIKiiJv@?MEC)*Ige<gxHi%^j6vHv|
zlErlcQG3iIY{d{zS;6*752Ul<k6Hz|<c>|4f<QwvMzJn-;2d|$uK|zu6__pk)sooq
z9&ifaEH0u30ft$g(%C}X?5AXkdcz`Ti10H|gV4X_cC5PH0(hg5YkV8DV?vBFBeQkG
zr4U^{Mw^zan~>SsZw$VGfuLM<M3MnVx|pWB&AB+%tZZsJbu=?fJ_&6SIROya(UR&U
z9<+9`<rGD;C8V>6ngC#E*q!?rjb<nf<SB&l2(J$Z@j(Dbxo%9{U7hm56+}U176Aqv
z*$+bfQ(HCaYZGPt>)?p2>&_?JdkT{Dhw$0&=}a^+F}VyhgCB$NWt1h}D@8}Z8`PTO
zxnL>^2Co)cq$x+3cYZTpwi$0h3=cd;URqkZVrq|j0@U8Kz$esLx^4D|!7@Zk`U|Yb
z(6i700QgKc?2k4J*9@~j$UggzGQ2I@EGnx9oq!MM(wngo;+drzCU}%Y!-4HYQ4t8k
zm<5B+C@w$wBjZMW$UWNr8SSAz2X#d7urs6>pc6ot2YMrIS1oXUL?Vv_-%+Xsx6rL|
zaA5Rchck0{Mr$_!?gP4}HRvF`qrbgkYosYJpxGWJW+7NRIIS?9sGS*%8mN{3Z~-tX
zfTl2jk>LYf2PlBqIY)>VCCTk2E$8{~F_UC**LKg&Mip4edTC1WU_gk;pFUm6EZf=O
zkpy9w2fV5tICtx>2iXlk;x%6oD>1?R^P-QMc|2w7jp^ZH4mcer^G2miA)rV=TC9iB
zgARSdOJ&GthRre2KulZ+l-L+o%4CL~1gftB4kj6LH|LI&JHedA8B*Tj?ch4_XnfI3
zAs6^gCc{t`7jP%S5C(h(HLn?*{(eG9XUlL@`Z(C)VDVf9Xp!82LdpQEwE??*LG#Ps
zIY+=?NWAr-W;g82(vMN<YeGH-+=Kx1k9`yVqevbA3Df{hzzmeRU8UH=fiI^Zqq5gG
z;w@QH0X?D=qIVnIy;aHW!h9ej3`6Z6HOMuPwv%<sToGoMPx>KoUn_c_PA6#T#crq$
zGCu8n123`-b9>tZ$qY4VK#~PW7H?C-@I)|FHbW?<pfDTRIl2Hw4CjBW0=@LKrNkba
zGyMif+y#eV=%oR%Zpttc(^Y6|iqIAJ{ag+h`)LBzu<*{@_V%n>>;PeB6*YJxJxF*#
z;+xCHQZ3Y&=)?dKPB}(fRS5SH>=}|yIY5P|4eWKCN{0nBDJf6Q<PEr@855qu@dO5l
z<Q=OOm&xwWrabwB&!XIR){(dkomOx|lrSH#*^2nxVz>6&t1V4E6Oj!FF<1kdr!7A1
zUeZVn*++TqD|rr(WoRf1j@g*WZEGHi2ernh2iN;&l^;UyA^gXi1qahcquwk5HB@p8
zZz?Gz#f}^~0*Gn+mgV^9K`>;@;0tZ1?d1+WnEdDkBwFNrKMxXK0gP=RE{nmcd=DA7
znH!S!0fOvKHR-^J7^4mvQCgcHLMR1P=d8FFB!A6Gc~5`<+t&fe1Z9tAiA}Q@u>CG#
z0D_Naz++Px%@aY(y8s=bjTgbv*tkaDi}6<mX%?g><-r${#(=NCzM!z~G0z(T$H)*<
z+z5bvrmn5flbmuTcSZ;G-~i!ALp3dG!F#jYrck+-*q%@}`rG+wda^@+<VwOGcTP2c
zsF2hjflG#1n!-~-R?mUqg<>!z(w&iLhvqIJIA=Es&`gzp>94;F+8-b#BhbVEM;JmV
zv5xu%-Z4=uMI!PJ=G+WmZUQx{FsWJ6HigjY3tFAG;N9$~m)%1l8Vc={5vbRwWW+!=
zMe|n;Xx_$cZX1xx$fs(idNd(5`iz37<~pr*Fo-eF)c0nw|I1LFtDvxPjjfad1aw#*
z2}4|TG5Q(W0EwSLWxJAlokp1)vx^Pt<0V`+hM|))4N^Xjl2t6b{X`1Fa;WWu`LNEB
zI|<rFcIjg2qA(#MxW1?-mkraK=%~XlKavL={JGK;S}_*6fQIk1x>0wO#oHsZ_o!Y{
zo+EA12mG1rM%aO%B`9Tn{_^F0YU+73(4+&akzaBdBj~~p(fwye8Bs7o3bfnspV`@*
zjV*QAR4W~I-?MiCW>0%{efPzEfVm<AKp%*igql`XFL_KHC>fNK48UA~x_R|ji_koQ
z^C(|z?MRqz3{vw@^bUHE<BPXt?NJ9gQ?lB>G}+FD+E19taaLX?RXqzI0>Fcw`p&31
z2Dw*&_)2HAjwKv95~v(DR$(-#4rFmv;T?VGlto=)0St`7sdk;+vV)=qa{-NUS3Pk6
zE)8L1009}8ea%rPtdk6S!pMK^_Ugc{Jz-u&-r9Omi5Z3E&9@iNjzvG!aa^4#nUz~8
zg#bEl6yspn1XK9PugW&sGcUt0(fAuS8<N8oaBlU83qsB8ZgUJd%?QvUm%T8_tQX7a
zxHAF9?khvlYnw0#wX5zejMO9W1p#ChR(5vhVPBzb{gZ0Ou%SRn2?Swa6lQhuPAS~h
zH|D*+I}yfaxbJ11L1kOWvY*H8Xc{y#h;l}ZCL%f{<32usVK6m-i;IMur(?IxV~qOO
z%8uiut{pgb=F_ZRPG_gRP{8z+%|BPnRfKAXUE0FG#FX-x<5#BN@h#dlHH3%9*wilN
zhU+Gq<7=aoemusDG?fb59k;xs^US=V^pzs%9@2fs&r~E1FgI8S8kAR^I#$1G-DLgz
zo!pNL^zX(BW~NGc7({Fi(DLd;*_0_IS6;~LeFi1Sm(U?2<IBKj@8j)l@meMId6(JC
zF0(37+rCZXG4VM$)R59r($lXiDjxap;RBn1leI|n6^8|kQjFL^y@Z}M$fuw5EBC|b
z2o)I_*{!s8ij5d=Z{tT<_tSSTi@M^05Ryi~jdZn#P<!*iD4(pylniZosN>_~^I^v4
z2~2w=rluYQ6Vdd$5nPs*6>pLj^;tl96}5-MK1NT^bvi#?Bj@Gk5Fc+EI>Wi+WCvlz
zYHYF1UNbWms;a7S2??YlBO@svJ`|dG3Z<+0{`vC&j*twF@T#HV=#l7qmpsG9tKeUu
zoMH7Zva>IPn!O(+@T_~>cV^hhZCC$xc|ydVJ$LSbKDqqW-^-3q7KX9K0D`VO`Lhpn
zbhhb-f{{<4YaE20MfUb>Dn2Jz%bxxAKF^--(K&~ADA7?>S093bIu0(b$`=&0gSdkC
zeUdypM_~;WE7LunK&g|FmDTB(x>T-UGjwV&Dd~Zg)iGE#x&fwvJU)Ejyn*K*DRnGi
z>dCXL1_I0vk3TM`p!s|M{{2?NHB`<}i|&M4k+4PNhXe=fRR=J37TR*LlZH~|_4tD!
zY@R-?{B0FiR>r({?_MjB=u->~&wHj{){lIxsW~7kCkMOQ%eL>@wQI+YA1D2M|A@b=
zXI9q5=U!ecVqy$$-n_YZ>hQ(W{_y4P>+4x-aLvYsKc~{i&}WZM!Wn({Fz@#)*yF&q
zSnX^%M2S1m(DG=N&&|yRZ*rbEbmGJbJwV}|r7ofii;J1z*h-SwAH$357y0?0dwA?w
zT3xN2uz3Kaftf`fJ$$jSusm3_7n`#eJ4H`V9f`rd#vD{<B84Xb;DoOfRnF(npGzgF
zTuoP#hJPov?laTZ*AEg*SEGQhXeIJ!C3dCZB}eg}Cd=6(j~}{+(In2#KLO~Mn4Nt*
zBqRieERRD0W!o;<tgoq&!eS4~%gblEZab!^e*7aBWNT~t8v?otdVpKEZ(kabkWV@e
zjco<A>FA_}$Ho>{A5&QuXc=2t(jIx9uEu31NDp676*D5tF6#21KJsU1=*{cb?(g1V
z-o1Nw=E8*|6B83R<mIp3xKZ;g(8oAc&pnLe#;HNpgzyE&l9ZH`a=;~48%vMDPn08F
zPSVl&0Fuy0R{LU3pOBcyJ`G0uGh|oP+STRL)TCr%Yg=KF=le0dzhCDH!_g}Y0q4W&
zcutypJ`Em1MMd@Jhu+0<y^CiqUp@|p=l$x{aZNZecu;)1iKqC|dGNCe!7Eq1tEyx{
z(u&K-P%=~JTHi}@>C&an{3nSVo8Zy3#@}+aZ0zj*z+l4nPY<hasi_H0U-E$j0d?h;
zmX;sX{j<Ehln|;EV%;I;W@d+{r>7xmAVXhOP@n)-skOUXB$fT!gXC7@*P)>=KYrvk
zFfw8R6;4=K*xTP<TaCI#Cg(c|@9ozY%N1mDMr=f`XRX4F)A-Cx3{LXJHIL$Av0!j#
zNDds_+|N}~4cvteI7=|Y4E;B3NYo`3qsGIB5ASpG_Eo!g@9n*>boBJTZ{ME!R#kP5
zheyebE$ATZp2G22H3{#9eSy_`)7oFHtk~s#{v7}L^GO<CGHB_>CnmyMtubtD4<MHV
z4=fI~Ixshvf4{VG)0JqY;Ly-NAV6?jxKLu!aw|#YEG+)Tix(Um98x+uOh4M%G%BAS
z&DY|Y@6!8p?^`n%;1n}+*~}Op#7ku3#qgVdiqvD={r&u&>tuTAWYTgyy(TUF{T1_z
zYkOjw$stKWpJ893;};O{z~258%#=bG2C*sQ<|fH-l-$wDNk6X#zFOPcU&O@3q-BWY
z<6Ditfr$6XaoGeS>x`r%^OGk}I<{kcLPh?QaUv)rVkrSvUUAvrhe4wkfSi?_&RAYN
zmF2K#h+&I(9TDN4!m)R6vGq+n=3crQNV)N;sn{sv&$+pJ#>U3Kn$ud1_4rRyF*06I
z$~e5=v*u;{)basJN?P{&3$>xLi5~aUY1p-K;$j61o(#ipl)quHU)R>IV6cZcrUsWC
zf6v>0!$wpORg90PT+yeeqx0yQu9}<6>g=`1<6t}+_6BwsmB$);+JC=c4jp{W&(A;b
zF}ik*^f)y&6*IGBvce0x!6bSC5s_Em;7dzOaYaR<XUf6(52m$?44uhn0N)1Jb8v6~
zzmU<_Kl|X}!?6j*WANF%gU~TufsxCm7;#^;am@$vG^bB<W?$mt<9ibl@?vSp=G*u0
zPtN^0gvV4E=J?#ya>%*I-KMmn4BGaI+S=M1SnO3C?lL|o=*`==-p`-!`&L~|ZoBN)
zG4-O+keXT(S6&DesidTY&oLlB{|X#JTwI*gO^&ShS1vp~j0(`+Vi|sB;9P2E#BC=}
z{I%Q6-K_fl{Q;;Dkd*0P`TFW#JVj&mP9d^+%(KrzPFmXY-nk3pwwa&BHstT83l6@M
zKdMsKmHO$^C$PD$AbTOKbOfglB|#~fS}FO?&l5AVvvS}f5C<(C9bP*-u1hN`37MIe
z-K^z&NvfRsm&0Gf+J}b?Fs!UT8?LY`NZ=jImJKAlwPH-lQsQRoPc=&!PMtdTIcIZY
zW6<gfJcm55M|`S7HC6S7qT+QsJHhPi?AJj-o>5T@U;*C9=0o>4pxWKKdlxy>AL#Zu
zczNY?bqD;I(%L3o<MGd6f^I(if}46oqoFZ>y)~{I*m%R73X{*T{<;XabhWlV=}U@e
zFyy;@`7)#fHj}RRNl9mUcr37nN|c8W$Hm9*fwN2CR($x7rC4Y7E_3oNj@NKv&<5x=
zMhL)38?0QpAo^bu4?rdWZ2kmrUX+B(X{b4GS4i%~-_iq`44RC+FpTI9IqK5j=D^mJ
z)`n$A&>OVSn3w=qD9o(BihH1<a$Hol*&TyD1*Yf<5EmEE>Y4td(hq@`X{wQ2=Iwo8
zo^X!*q>73P+8|_VEJ5z@9MY8a4Gb`G!twPZ$0e0NuMf=P3P`_9MgI&x?=Q=8`bcdo
zEexsLP*8Z7bM`(S77F!&lZ)%Nwzhvn#Oa^Eel>l#`KHzQ#K_*^<D`&v6I_c3CSiIV
z+so_DRs~uKn5(+Fy5*}HFc#FgpPW2%bG_X}0IdR64<2k_VBjq^BVTiW*|Egf!b0}$
z-ID;pZfR)vfJ>Mv9iCct1i$olw}jA0NLbTWQlJH=y!w@dm+C_9@89~tAtCYbu>wFj
zaChnCj)sN?2Pfw>z%GX<DBj@l*&q2BUwV0cKQ2XZpJ~sI%RF)NBx~qfLwIcvw;+ke
zDM)2~eSHb~(Y;4{O+G_m*&7`lo%7*(O2^c7nsA19f&Tv2R8={aHmqc>f9ESNaxd1&
ztnpiJ4qYDD;xieE2*gR*+1uOyqTXYDHC6Q-)G67-+dqQ?{QY0l8a%5tI3B*$+S*#b
z!54Pm?EzQ2r%!!nwFsxPa}L@mU5A1n%=cr$rp4rQ%Y@AV3JPgs<CK)MQ@THHq^c@$
z8ezgN;}uRhh%woVG5PuV=@}Tv1amB83W#33XjPeZi=^D(vd`=pVHP0)fe>nmJEs^K
zufxy}Cnu-X`obeXjOB(o+KW%HSnRn@&0DwJL5z9+;>Cq>ZyyVWy~TWUphYJX6dWAb
zo}rBEklWY0`p38Is+^oCXi#Oc_W7=L)r3iPNw3P^<#&{~kHyFP5-RoCW@}@shps9r
zvYMEtwfy?^>bJho;g=i+hKBl$5w;L-d&`5mxu7b`R2b&<Jb@l4O7!VbXrD?9&)muE
zc(^>-UQ=IR38SsABO`sGb7QrIQY5;VsR_Nk-{`2}z4Tvnl7ULe(rRj<<HOtpGyc<K
zPnK*FG1zPB>a-#vA}y`0XG2Ts!g5k68JL&?f`bhp0(?YX;=YWiXp<JbM!CEGw{DDd
z|M0?oB1RNa?bgfSIB{s)Muh4j-cxSB=h?x_oScXDdtNo5t_#~LIaBxr02~n0ubR!$
z+7CrGvSn+_$;v(x|0aBrit22g04p2YEBv(u*1i5h@AL9#!+3RN9UO$g(vU=K8wWrc
z_GxIiIXp7rBWnSWfa@GzL}*OG6#pkbMSl8lksfLMyT`LdF)lPgQ*@RN>trC_TwQba
z;KLu4qhttRijs8OkHwx}n^b4MF0(Ie<2cm?HmLb9e@zJ^Elg|95>VXWA|^-;QW_dv
z@7%{)T8=fvN(MsT;oI}mL;!{@UWbSKK)wI;P5ja&D%r$K99IvA)w4#NSU0k4V)Ldw
ze*78ySAZ?#W>BRLw{Twt?TlZi2Sz{2U7*ZvZ*QL)1RD9(uU~3JU}4)mKo5;XM8th)
zft3oClT^s~9fOp6M-D3n)#m=4ydjT&UQu!N)8Q^RY_{9B#`>6thX+~x`y=4U_V)Rm
zHX_maQ6@q3v^oWr_V!o4e*OA3M47+&91F|!#-CiZk0=X^iduVmxCGbv<YEHYvqk?i
z3vddbvo(U}rnn#`rWO3^wSwHuu#)Du-&?PId@5%8iat55T98mbi2=;_=C>2L`aSMY
z_v#yXOm_W%!x-ZeCwcipI2`WV6MaKNHmH7Mx#3ed_AmuJ-vAt{42y1HHFeA1rvUTt
z7@zm$fM{!)n<-Z`@7b*N$mQ4#*N|VaoxDO4mv2uEJiW)bF$=U4-zr?RwFLqg)(=1n
zC&J{ir0gm7;M}>0fNSx`6qBL%AtwQ_R|TK5d;%-F%{yCX(W#TE&akF~g;p&n6bh`2
zoLr?b|EmT;;d{qvX!iTd?)`lKLa0m~xfLWK_)1Jne9)EHGfh*UwH--Q8$E7hO`5FA
z`6f6RQ^y4nc1~239&&L#4~g8JI~*qx?1rE*-8&P+SGK*kY@t~_k{|jEX-!Rf5bWN&
zXnR(iIXzh`fbo}2Jmp)VCk1K#+_`hdsHi^a(Jc-Q4L$Ig5Am-Yey_jdKfje%aPO7a
z$q;$gY<-q<=VV&G0WLXr?wVFzSMH9cz5N>gzPXf)I@Gam15agTtNfUtO@h37J(Y$K
z_XE%oN5Mr22_8X0GvAAAXX^x^wSg~V{?qVvd-5s|;fylo$z}WbLkACDg_%>Fwe>vA
ztCPGPFS-SV<khP~&@Uu@_;3gYP=lDb-16LDK;(}gEXU;RrqF?#N=h$l4er~OQ7mA5
zwJz5Z7V3Ur7oI%V;>1eClu^Nmi{F7?ZTt4y8x4M{E~Wc>mRlq}orbCxK4-g0UsF?a
zC|@`A@KpnY6TsZQw4D-6C@$tcRf##5K!1#u)(bw?qSqWFPQs?-2cHgv^;{^V6%rIo
zNKQ7Yt9B~U`ScO@={5c1TJiy$zt(=-gvp6RWPY+NezHTs6$3jvJMKJVfR@i$7Z(+6
zexUoX;B)KY!y&&aIAWUD+Na^)7qJN{sd1{DRM3$ETm}QN-E@{1xN9o6(GkeN$XF8Z
zmJ{AGwTzllXmIwJ6ZT`(E$*j>YUIf9n?jWC)A4zKoQlc=MA?TxGyM_tsyus|6sGo}
z-?6Ck=>h0{du2-g?Afyb)gO>4YiUK*42H7XjvSYu2q)uKc3e{)t&4Pv`z2{LC0Q$~
zOI^tKWd6_f{q~tz+<S|0jXfZ^&&IXxvuoIb4>>=Bm<BcY=;n&}DJxxdb@mvi83G6g
zkyfjHo8PAIv=4s1-!nf;*rC5&Hwd38o^apW1e4n*goV@C{mrdfe1g~m<G*}4WRXV!
zIgR6d*NeY@d((<pm?tMqp6qPVL2@Y{3wj+D_3F=`#+rJrw>EK8i|cdEgsrC}?x4T8
z#VpV@$4Y|CY)`(qD*5bg+Ct`{_`212OB#uLUy)_>iQW?Y{q(ArM>I1EDXFL~xS~XM
zbaBi^ME}#hmLET!L2E?yb9!>J(9p^><NEdMWc-}?u<H7*T4<sUA8U<TXx@r8q5D)l
zbkKO?S6A10d2hrjSmf!dcO4TxHf42XbiusgAT-~h%$wvU5SCQb)Sf9R9IYb~oSd(M
zIAvsV$LaiK%?3&tfyAKAE9kQT-Z-JOlnE$bV-u5W*4BKG+mf9Fpj$@;ZP^J*ytlXa
zG3o-5XeEBqV7{u6S}npyJp!9Mio}yqyFG?+xj{zMReHZ{)&(a#Fg#prA2Xc;7bYbf
zfMx#f#q63t_)v#4^pU#|Z{{&}C1)s9uX1w<!Uq|d>>ldHF_Qpp3UJwEa(7$IeE)v`
zjUQLCbJWSi25bcEw=EafF>A49O13<jckg<DNW~JTlm7k3KHsE+@X2#>m>CeCxCCpl
zUmu~>|HMz$+uwOFu&T-t8p<J4T~U(H_r<3>-~U-u{Ml84i{f)oc5yj6Kb!wZq4ut>
zm$!GQq-TXYXwXV^X8YK(>4p~$;`$VY>C30>d`I3e1(Ksw__j_vZNyJTSm6>kH+Nk4
zn+m;S6_~`<D-)ZqIXd}I`;R+HM8A1M^^f7n%M*%oO^;&b<LBR(+;%=x=qM#+^ZGFM
z#*G{03l9o>-s)sha(J8xllAuVJK<k>DvR}8Lddn`Hbo8Q?rFQkq@?RVXT#K&JIw?;
zr&s7671xbO)9|fd(Xzly>Ppm>nc=lU8{LZDsGMY1iqdmm-8HMIC@q-vRaMo7J4t4H
z_fZmWA0KH;OQ9$w_QNl4JaK#UNSNfz`Sa@$nb1-B!Sqn_{MKUGv-a))q?)T4tSXI7
z=jm2g`D;Rrx&S>-b!X=lPOkd-(MY#%t%=O~LsjWPA`eRmJDOSdA^>y^#(V8RWyZ(d
zBC&I#1@fEqUi!5C_Okk1i!{NsiXb7h5%ckvPtuo9pPpF%s%@KKaIYQ;UyumlG|-5-
zXkXH=N9o~h+@A_n;<kd!9SW$tQ2YR18*ZJ7ZDt4U^gRYD3w*TuoRb_+J;_aj)Aa%w
zBR<r^3bKi<Qfckk!iTA-T8{wy1Xw1mqtpLx6I!E|o*rLN3!z{2Zv1Uk{@coXfgI=a
ze4R7o-7OB%Y~y7|aVYL7%Wfm@G8j@q*llUEnakUu{m<`{u3)N)YBRnTuhPh)|44)B
z`}b;g5)vNWQg3l53);Mnc10VZ($)UW`Jw^wz+Y+<hbVkWNQQ<Uj*lCAs-;N-?5xUU
z<z0-b#WjEZdJYC?2*IwmQdI%e%RYQ4j-K;HP(ezAJk0dKqeo*TB=C17bocK|Y#iOk
zaAnRMG2z$D%}+7}=$!Ju#K*MiU(8<Ql2Bv+;Lqp1zxkGTHZ37FIo56Btkqf0p0#Vz
zG5s|DH>&jaA>f)?+W{lKX;+c~wm4Eg>Dew#9Migk%ehnV@hd8FMQ0$9KB60r*j$jw
z=-qcKU5y1=ZIjO;afiR+IMgARyv#8QBy&1tZPEVW^NkyC*PoY_J)#?IH@>FNU-k9(
zI1mX3u1T{Wb>`yaIeY#*+2h9(w>W}<d#2f=BW$d3?}~%NgjT`C`^e;Cm-Y{0TBXiZ
zN98Qbr-7L!h`&#8AI!L!)~;V4I?ZuG96F@?!2i1ZkXWo33V(}*MuK9&5f;Du57+e@
zH@s_NSR;Jk{bcR}rmw92cHZtZQ+t<e^mW><V%ko>I-n0;Uel-z5KkU3ChVT-W@BS?
z?<w;P8JU~fU$3UMlfErBC`(JDCdue_zre>wpOnO|pL0RW3s>-@PwUqUth~I3o4A1+
zAWERYVsbj}bLZaQUjsxt`bX^?7C#77Sf5LQ^7p~PGg>^fy%tZU_Roflyv^YL4xfF$
zE`1p|+FhEKB_bF2D6(K9M9)NOMIf3Q8(j0#q>_qWG{<Z_3B!F;bgG)Chs07%?Pvg_
z=quPiY$hS`;R9i3Lqi&OA6c`{pW9EJ8kbX5<>%yNo1ATW)aAjYsC(0KvV3%IPDV~T
zn#Z|&EZDQsVzr@&cOWG{-$&#XPI0K<qPE>4&TeTu6~z-r3N-ES*i5PY6<MFWo-8ZA
zsdL|srIF4%^bqh|kfB-)8VS#pB_@Iht{G~c9kFgl`X`jOwnR1S*`0UTU3}wbfVymb
zFt)gk_=wilJ<wACV+RpMtTVnFMSPEozqCNkGyX6+pmjIn`;=}_2cb|YSu<;a^3x~7
z6EqfDuRHE@hMK!~n8l=2YTkZw5a#)4p@fo<eELLPQTy|!4T+R#8X3@}(a~zpy{0lx
z-_$n7&8Zz-QhH=yFkH4;QfhV!8{yJ1^87i5MCxfCBNVa&M~*}<Tf^cO&S8%O&HL@D
znUT@hJGSGVo}Q;ppN_E2!3xh|m~R4+4p8+R8{2;1Zh-tts0!DSTrAv>p}3020MVTY
zq6xIk52hw2u3B310?Y7wmw6E~-`lMRj<|>2`u6SHL0=%M=bzHi(bdLXghH(eq&ST)
zs~(Jt0!)qc?%)<+3Y23A(xL>NkhEgQzk9g$K|y?El)%S!lkaTGg7pHP0;i9_pGXLn
zAIVOkD?SGHP)}?<!_)M9#zM5$f|irB?oDXu*`pT|6A}PkJp)Jn_U$JBTtxFmQI7uQ
z?|a{-e1}n#jq|JUQKx+*7!2lB&z9RND>mYCTdl=x)N=lV2M_cv%6y$*C@-*126A{t
zLy{Jjst7d;#4ckqGnqSgsJUq<02;zcdVo0`T+H)&j_j%tGo~|X9aD&1`}R!+!_Lma
z%1U<r{CQwYq_wmday%Y4tvgE~Mk_Iq<&0@HP)LS3@o8zMzk=v}J+@Yja=Hw2%dW+{
zSLnI78XrEerL=4f?DpaPdkH)~-o~wOFb5eK8B@~Iq_wpfVbrv>rG*4&!?dwBV}1y?
zlLMY$<Oig8kG`xMIWb(GOyn~ae-4kD;8OQv@!vRO5G%XOAAIUIGBDUzSy}m~NK1zb
zef;AK>fLbTj`g^@pib!iayEp*1ByS))`ehC7<cWr_wX4th;`k#B(I<h6w#^mb<=My
zhhC7Tsf{GQe}CxqZCMg2rKaVMfe=wzC$Yo>l(9YB@eK`TB;`p8kMgI#6tRgDAlu0+
zu!pXb2WWfbUqP6trO(`}i$QYY2`z7^;$~rotMVb|2?b&`NR*SjcK`l<6_p?FT}ol$
zs%oK$DJg55v!UMJG-<7^x1kM<1K|Va98*$LJ-$B(U&dcvkGmUFKuJMCP5u$^?}Jbs
z{5yBgHeiMhQ&LtExkm8&4(D|E0~1fqxF&%Z1b>d>thhKMxMfX41IShIMwK{~3scgV
zLhx>+2g2;rIfdks5=H>_<KyFRTW3exp`W3eThgVW10X&(XZ2<q06v?b@DM8oBV$lp
zqrtnw?ut7|UO=7a{8dOqgi~IkhTXL-dQLdHiB5SK(kl@BdU|>!K-G8j>H@zLPWQL$
z1mr3Ct;m`ll@!_=GBOID7%Yq_Tn82mCZAAi)ZKkQzppJbqf`YP12MV5!OxMeVK8ST
zB$$A42F@H>2VV(3Ahy3w*t}BQkqC2JDO(~q0ROF24IgEAcUMs>NvJDZ!l|~Qp$f()
z{HWO_$Vh;BBkPR~Q+NOHp&t14eI#ctU79G=Ws#8B<L|$mZ~y5q1dfM_@Kc4BrHup7
z5HO95(9Qwe`3l1x`-5r^zaSl-p041^EZ&DbEvKfIQFARcUAK(Ni?Ym1sNeo1rR3JB
zb1aF2i|cxMPk#?q5T0TD2U?uum7Bl4SJZqmqjg@#d0wcq@bkV%Axmj!&VuWXftB*L
zq*F38zyGeBWjTNT1*nb*tO-R$^hk$;kG<W}(Sex_5x+oXLdoxDpuar3fY*Kk>gldt
z0*Be8G52$59vK=wMDTzvytuan4iq@OtS&Re)-TM1EQiEim@O)wh7WJU3}a1GldPH=
zO?FPso9Jl&wKaP~nv65JV|nT6iu-UQ^<9P^xK11buD!xc&{yRgThRGIOMu0Ot+7iG
z-r#4x+>Gn9w@UjvI@kpZdcWONUcq3G2bM75N~8#lZkKD_G*Y&5ocjp)AW<2G$lhG}
zqcTShANKxn3pA1kq^5-;qa7a}@d$mw^Kpzw?71V5^<!#zkFCPUaY2ahsNnksj*gBX
z;Ca;5$s-*Y`edGt&zlVJxj&S1xVR|5m2W63%Q-rVfGeX8pM)h=*W~kkhDa?+EFSbp
z%(D{pog;->8SXb95NZvDKS}Rs7B9@UjxC}{fKSXuCqFH*xiH0Ys^Dnm-?}p?Zog7>
zn4k37n(c{8G2RQ>OTbXIWX1eYtW^Blnqd3!csZjGD(9pUr!v3(ekL26+}}cF>m%<r
zhyFBlmJyWM`br<S(Kzp{KLLRc{$|n|w%q05K`flMi~*r(MZC{t)#YO3&^*jt#OE&k
z%x$J+j&+BJ*@#<4La@K?v#5s8Z^Pf^(H->VEtPWG{2XaqRb35=oBF=1K)ZwR!>Q)M
z{oDt4^9cyKMU$>jz(9^iLgv-N&`_Ft+%ynxH@n@;%x(t9&Ly`^kdl@z(9qEdC(ruQ
zG<9~K5)l=>zG7}_nzknX*wQlg?s>Ar)>E%P4aF8byc`~<mNxb~nwuMXb<$HlK0Y0;
zBy4O3!-JU&OCh|;-5Kl(@~%5wKPigwkaN}2PI4XI;@$SFP`6_$%idd{lsxqOSS!QA
zy7Tz#ENzm-Upvvic1}`9tQ`^_UlL3&RFRR{2X*J$6Te>3GiT3UHNz=0MQB3Bfy|cg
zynY!_t{-qBHg-Bi&yzr={KuBwO;bC|&Q2a49{v~0Pe!<3oNE<o>#m-zR%sR>2Yq^W
zHja(a;JU2rHTc!k@`J{R1dS3|t*C?Wxi>;ZZBT4{()VX?-srD6murRWzn!YOw$VYL
z1RfW}ZaN@8|J#)-a&T)=ym+usktm;3^|W-+B}#0auCAW9e8T3dSwJ;NKUwH>=f8Ie
z+bbV}QAM&BFRsH}f|R+r03@u7Vt-#&9P{z>;|8*wMO=Ju`JcvXXV2Y#4EPUmmN2<C
z*bvSK>T1E{4iwaTIfh3-P=6B}8wh{JiZ?aR5bOi<LoZ;%iiMdV^%E6NVEAbSUL=2*
zow6(XUh7^40x8v6V?)D<B&Ja5(JB>o@_irGxPT(o%jrTavHbZ?vo9KCq&L~9L5|bA
zO!p@EL|{XMmVr^GEa)XRg;UC>?i~b?8!ouX*5<de>R574v{Ukq{Rh@~+{6K4zH!-_
zaOpdWI|!g0EYXpCBrTHI%3;D&*LW7hksPnb-xopUxWbT-;bv;e1v>F{@x`fhaz~aE
zhptPd(mOcg{M+aSK%RzD3nb;^)k~-Sk$Mhqhd>1i#19ELNlBNi1f_w36^vA5DAO)=
zrh|Eu4y_l7?Iv%+!n{B_2Bn?_R6<a{qVf~3EB<~M@*p8@4L--_6Q<*MkYo=2Xdi><
z)0x8yFqKjO$|3mC8hr5d_Wk>>78VxL_J6+2$9J13bjPQp900WL>sYYn4C50}B#wm_
zgTx9y<Yz=YfN}zanUt*T%arBFh=@Z$wcn_p3;(ezx=s@o)!GiCJ)l@1qd`t@D)h89
zGvkJVCeT@qTZO}Jpt$C1xmbxzlepL1;Ik=ADR4IBSAgy_K`GDeh<;f8$*fv|m)l{c
zo4oXlr%&H+;g$kI3`|;Q2g;l;cPY#npp+&i=3G%F%|9T<L<z9xwx%Xm#9aYFL8xy{
z8^X3CjKICU>+ZwVhH@$7u611MFy?PD=1*Aq+FqJGB9*E2qhe@&AAc5J(rCE(tOM>t
zS()xfEhttc@vEol=*|aR*vf+!wcc}O#X%SQ@#6{2>({>vDD1#YhQXsp@mX27%pjwR
zsU@=hpZ2~pD5`9Ywm}gUg)!kcA~@O#1_UK%R1B>Mm_U)BB9cV1h(vV+OrVW|k_F=+
zAd)kPWLrQ)a*`Y*h~)Iv?q=@XAFp23`|;}4t9o7c*7a)lIs2Tw_qV^WzO_95frYgD
z<TNXrn7mootrr<!uC4utq(#F$fW5+!l72^hhfDg)3xhD<blv%mZCMsTa@h&DVN?gz
z)OZ7Cm(Kv^u}a`0uDPuE*W~16?)zF)gJYIxMa4U9to8>g<6^ALda<I}pW25F)->rh
zpv;Cc)55W$-REWx<E{OsGGcMCqw&$@D78fRl6c=Fm%q~i)vw%%rnpWT{V&+O7`_ZZ
zgy6MSK3=qT-8u!s@U1?D)b~xR->ZJr0SW~Of!+4@QgHiKX5mL}MV0UQiHHi6iI*+C
zfMIKE*StS-?z^S5hQ7Xte<(Wn;W@SVd*Ai)%0n;2N8gZ;%-$JuW5e&V$PI@XD$}Vr
zj+#Iekq^+E7f>@MHy1D1J0HlSccv$K`5ju#Voi6YEZ?$n@zIUZ&!4N*DJm*@p}c@H
z{q5yEBv)o`9%Q{JG{0rpr0Tv`<$uDZfjwYjB(`p<x7MV?E!K?CJyBgTvwP?4`X{dI
zRk>XX*+d<Z)6?d=+S=NF&NC4IV_%a_lt%zCDx^qErEb4_ElFHy<>Fffov9CLBFSa<
zbS{3#<c`T{{%yr-R#{YQX?wv60f3U&5{rkk+fp6?bGIL5;sEKRr3dMMO-x)n+Bgf4
zlUR{vg=sw7HhmOVK7OpVUXo=Z{s#_zwKK&!J5K*udD{Z`5X3P;`a3H+ZdLXKjr^Vi
zmt;8)T#EB251l{%J!~coHBcZ)zB?DLX~f+to!)9k?nee3sL-1qKJ0;8#k*+HqPyKy
zw~eDjZvsp6B>S+6)ZeHvp`?j2#L7d555H1yAbE4tG|@5OMesD1@7C_!yJC3L;X{{i
zYNuc2>GpdMe;IGYQ~L(*-F@N?sb%{$cA+SRvTs>0EuvyiM8$^f+f{9cH6tp3UqeD=
zA+CrF!RswT7C)_zCntoa71rOK_Iq|sp1-WC%9Zwa-F>HLzVQe4?YrG8>>w7v*buu-
zQPD+jZ3u?h?qAZ8gMc1O3AR>+VLZF}p;!=QkjIof5HhaFLk~@H8*hrO{lUa;gecxe
z^OSC^mmT6Fm?V|^Q>eycx<uDHyQ8#cZ=Wcg@gNZ>!!45wb?)zZri>i+lleyu9{jB_
z{O1K51hi?>ji`Zox;wCZx$id7@$q*1uj`CEKY#U#SjRaui<;W!FmAmSj7|T5NK*80
zvSPNs&;J|g@*3wOn^zE&l=I{j(n52#+VymGho!7w_rsl8YxxkNBIG(Bs~rh{v9q`L
zulmcGrnLKQhfu`nZAC2L#y|<@;t5o`ksDleZA5rpT3jifKwaH?;rD`w1sAmi<WYYE
zVF=QIYhz5xHxtZ3)@u`DmrtHNDeW>Ptp>m#%G)NR6~^KK;QY3N8?ZvSF=w*MqPAP-
zMNFM8Llzsw(741z{n+FxTXb;0HoqkLJ6t=qAT<w!A4l^~Md@!h$}jTCn_2O}dfFy=
z2bTDglgXPNAG-DF(<isbT*=AF+cG|_ioCSFOc@#LgD?`Pm=RDLxHER6iWpaHe{$<8
zfWbU|)Fy^Me?A|cg^o*tXP2#PYisLzK|!OU$+i@yE2-<aJ?3HYo<E)Fbc|X#`e^Mz
z#y+k7Y7L>48_Z5&alI->rs&SS^Ee?*?jp-A=(*tLxud=T>$NXo<OQ0IX%4YT^&Lb1
z1+}!ZlK_f9P=c3kb`9x3M~d5{M=J@M5eTWAmq!A8+m>JT2Oajaw{Mpm@aE4YsBiO&
zFAlx<K){%It8-w?PhmRB&CSgO#<Y9)?yMikU}KVgGvsewQtjo%4F@1%f$JvP@)w4@
zbk`%zDAFSCS<fn@?GEYOq<snCwGA7zshC0p`b9-W19MJilIdon^XJ2aQnP%9+S;`3
z#-jlBYZ|;NDlfkzu~zty5DV>e=NK36xgos0XrIoZ7rV~-h*_Q3`dDaaeI@mAu@dU-
z0QZTB9$u-k_|~575oqp}a=TqJKY-y5Qo|NNPm-<<ZonBgZ>A5mR}|FU0Y*m$>oc4y
zIMF{6%=c9F-r^7{4Iq_#^w3+i7CbbW=LBEie4B=q)u6=MlV3h?VK;+Y4rlQ^_ye=Y
zf9~GFW57gPZSv2O;p6@_JY26~Q@*Y6bT&?_u-0QvPV#{E?Mruxt^V|Wor>5;04kOp
zeSloSiIXRvrKRx!M4NVT#d54RoN{;WMPPasd|Q~M;&y;LW&ggI6O^(;q#8&qL!$&6
z1)fBbWivbR?m$!(787%{S4mP5$<UJTz16XV0+f{d>r?GpzMZjvhq@rsOmqF++#8sA
z9yF2sc}cSN`Q)m_9AKBa3yA7Dw2&c*($ZhJ{3Viv8>@kSIg%Oz;8J<l$_Ni2l7uqX
zSN8f9+Reg33$?8GvqKdC$TXt?Ut?~9K2V9HAYLc~rT{53@<_-Vz^l^I)r>y&=G{fc
zAK<-RgurwE?%iyD;Yf(R{e8)T-kNwW1ikDOB)*A<dR(&mEi3tg5||9k3>kecs2aVn
z!_F=Z;mtx79*r;?)6l5}KLV&@SS~B8otIO8NY;TlftZL7LR+4EHI09LOI_tI7qBZT
zDmI_GY{P&Ng4cym_W&!m1E^Wlgmm&E8h%djKEMiD-rl%u8RKqM)hVRG6_P+Rgc_?C
zda9a1;x-1t5#r>I0P}SfO$X>b5W$w^w6C{Y-Iq3`_C)pNdavO2J22GXLIP!hMH4!<
z01*j_Q^zxtuMns{(nw1`2;KJV)MoAj|69z8l_sKruoo{l(2^K-W>9sl#>bBzL*Kpo
z4VImtYXBGwc<|e@vKwUCVY~o>4T6yn)4nXk>E|LZKA9V!z=+Xzc2>aB9*mxyhz1yQ
zBY`qtHF!ug`K^Cdew)laI+J9#^A;m+Y-!1z590go4?GV%m%H?UQkkU@;2FhwX~ze~
zoe;mHhR*p`(=5vnWo|5*lf~O~wgAb26Fd)5k_J#NQ&Ub)7#F`qY}^pGZ{ITwve&Dt
ztEDZQC_u^%pa-i3y*GQ{uDjjtFTkPH3{uRoLUorLph)EaN?QQ9$&&u~eNhOrIeGX$
zq8&nf%Nb-HNs+iJ(Lw;TlQr=g$(n^^Hsj7O<>tN<8p*hLv%+*${Y_NubaqtQDIpn0
zb|itGA*X=DuC=`#SWYj&bo{0j)%@=$c$Qc-Ww8-RHHw{CxDiz|0_lX7di%y)bS~Y6
zN7GxI$b;~_?byt|k1c~u1kH=q14%>kdYCe&@nZcve0k~3?BD)cwtQLH{qsODd-@6^
zO%1jwGKqkPBGG}c_<*CM96}@ZKZ<mC;B&`T-@}e3m|h8qRRq(ElqI4PF+D^O)klR^
z0ugb*#YHjr>i0ziIZ8%^<R~DKY!u|g#wwy+ZV@(>B$O%!UWhm^LLIbY^!$UWA6NL~
zul|O{+XXn_*!zx0R3w4%oD};*U%&SBDaC68_6V<FJJ7yp0tv@HU=>Yo<etxNYHE5L
zTmtC9mM!p#79CVoRXKlt1JYi(C2t>NQ-i7R#RsJR-x=g{1Jmk+KBlWf45_19TAUPs
zs{t0z!6?rH#ExcVgT@T!OP>-baHe<YBoU76RY+%U0X`^6>CZpez=}=*Y{LV`#rpen
zPtW#0nB!{2Bb?CG3`<G*gX9^JD6+8Vw^KL(01KT?f51w=Eh~5|f>dgU$uDf@Vvp>z
zS!gDUah-KRhO!Bm_K#@ATwlaT@Yji%)55{7GY;q=nn!xVpH|56AiF+@mgVmiTKcQ5
zUvWiWRxbHH9%`m`#WLZ&00u5A3)@Uvj`lXI<_{b?DYlE?zz{HTnE}Ed0`>t48L{J0
zwLfKK)-1Sry?7aXz4lVzX$MJj_*KxR5;Q&GD#5wIDLUv3zF8|iB^N#fe@TG+TY*n?
zXT<1&(~+9Y8Os=NfBxmTY!ElLrN~Hsqc^dEOCBH@GoSfhUL!XLzr7h|QQE$b4+mvb
zb+0OL#7I8@@$^p$?fv_zs(YTELJQ-xg-818@1pC5m^u&cbss<g<&7jPMcK$mknpBA
zfjG@Tvwu5WZgNhbr8ftSe#jTH_o*K_Lf(iKct!U_S8|#>!h(4~3{+UCO=>+b%(;a=
zf>vl!in!Rh_3<HiD?qMN;6#xkCQcXN<=MTSpaR9j%&032uS64@l@og5q-{HYLwaS&
zdeMl=3%kx?iBzT9iy%F4Q@%U*4N?tg$ljsnI&BA(!jY6rI7Ni^AqFj&3yN=hYW@zV
z(o;wXYkQ#DD5RMH%F!0I6GVD#E&u_eS7|ToA#tg0++a73ui@YqMq#nu6M^x=iZfbT
z9^mI$fu{;S^#aE;ipM1_=<>>f!}HFfQL(oneJw!fyZId~3y+j@uao^N{DAj|p5X;`
z(Um=9!kR%QL#n$s#1C!Quvn4|#yA@rfe&Py0<Kd#asDd3i4~)ul82Om0ftlc-Vh~L
zu4iS&H$=4o|MBO6TX+?;S)dn?BE{=HCS9wIO`aAU1YUCvZOoOf=M5^Nbz~(VPKgaG
z+F@sJ--7<A9Q^ft^o7$#JKEdNgOiL92ulwhQBh$o*{Ppb<w-*fNWK&T(aKh&s*6h@
zIY4+)NCN<%UpGaMVHny6(dbYWXQBX(V(;d2BBrMPgXr!GkB;V{AYJ(vT9BGcpY4SK
z_!wm0a2xc#^MLH!%|AY6nOZM>rT2HzLxc8~i(gcd_J?PaBN<!nH>`kdTbnQ(Edb@o
z*-4I}DQjedzf1%BQ59>%b2+(<7a6PT61Lbdn(sb-oL^^GIXyO#n0envQ2jful!p#u
zl?Uaf@!#PIenk4w6Q-85Tqc3&0U*2N=ktOP7?a?PoOI}gHh@q0MG-HSxv|g;kJ{yi
zh2a1g<$585y}@4$HF$C|a`Vf3a;@H>nxg-z{2Y!GW}^_CS>y-ipTi<oi97YC!jyOp
zNuX*%No|0b2^<c}c6RYquYlGDD<2<mMvx^jATaD?8^Jqw?x1=xbG#1dX=GeGh-W}r
zFmRHn*W{lSZTc+N{}GNSfcD*+)_vT(A#6NxEOQ!$zO1LGC#NSm(@SY)&>Q)YrAg`l
z9fS6P?1_LEh&IaZ0p}tQ`h%<J_C#ZV<=DVEK1HYB5oGq3H#B&W-qkTQv3esUAs#tA
zsyu`}c-L$%j;%?P8XzegIQj$#O3o3qo#7n!89@Py3>|HH+-Qq<$H|7t`kmMraWm9j
zfI;yaM2qGJq;cw21EMA3+;un!JM%XI0fFZ@=$4gP6a4TzPVT0F#;MH&J55@Lz<P;d
z*@{WDG4RP<S%TwWp6El4Gbc{`g;1Mx?*M%Ygq44#)r6BG64TFM<k3Y#(BDV^J#V3&
z2IV!0vC;Uu3Un1KY-Q`yyu5Lu3X-kXnF4Z#X0j;IUMFZ6ie<#cv<_yKJABWQl!(j=
z0KEajBm+YS-9P<d!Sc6()8w3gjP&y$_5%-(XDyPB1U`<(AW5%Hn}8qZM)HG<@94L;
zz}TX$ip~jEhz(Q|JKSY=_XW~HpCc3%L4Ganhdz&O099U`<=KOTDRYFfAnIs_UCjYY
z0Iy{#PD{Aw*OkV7lmNTCY`Ne7*4Q>RoT<dZPfw2Kz^6bELM-|YbYB~`zpcOQf#f^7
zcwRawBId=svSOIM&oq$lnFAstSy*$at_v4Ci$R1rsc33`3lmzm8;2n$nMK<pa5~0@
zqh0Ukc<^x6MU{6AfXCuH7V~}`9i0dJ$uPu^=3D_o1B}25x7(_<sOpf~24)0qIx@X^
zV`F1+B|GKi<qgYYpTdOgRa4^zZty5vyc;)e>^sf{`|c~kG=bVE^<2fuADrlk<_uiY
zHGGY^WWf(`R`mB9(tEHQ!`>iwvyzuL49UGZm}_^^J`)$0Wg7nhCD~W6I1%3>S?^Ka
zg>JGuw33ku4v6@WS44RY|6w(aA+G`akup3b%pF!^$=Mw!7E3cnu{^r+u9ltyhe+B0
zJ0g@M-H3|sXa&uNyLksk(Rwc&rk<S>!0M2Mg@VHL-;?7`8E%U6mek<5Kc7b|76&=o
z&@X<RV3BJ9I~{=M)CBUI1^4zF{C0r(-e!S6M1J|52VS@MMq+rEQP5VhmeAA;!oTIL
z%fNyp9ov9oFa1)0yONKwEct4rCGyM`nbVRW(+S~xWPl3j$I;ALXOar>3<va+e8M6~
zC!8?{vNvP-u2ToFbMMsRK<DO@b^{9S_pPfkH?uD4$txC|ykc3Th6j&89!|siNOW7(
z!b2M%JWY68yZKNrWqued$;#OFX<$H<h;)31U0;@X%fq8(vT>Fl&9{n@D~Z{}M|%%B
z>a}<Ov3Bhs_{E+9%2lU%;)Ev{8B1bJE3PfDhv_CCU*>VF=?Q{miqb@xx9F~2>=fj-
zz);4Yqig@e{n(o&KRcOU#q&!tuiB2V5Ueay?5<-u0w2u*j=u_i;fGItzu$b$i1miC
zI8~%>Y&nC~{G`0=BYbZD-|FgNuE5G=h;x<G>H_%+KH{rMx-L$&`jRBkWkE@LS>iOP
z*Ek4wRt$9>>&eS@Wn*s3cEQp2$QN+s!&0MAu(w<%Ub%h%p82MnDC@yD_#Y>8BrHa%
zogk841j<MX@<KTxuH%OhKP+&#z`QMPT$!{VGqh$L9>ky@7;iQt1w4FseRj-y_N*mB
z0P-<@N}>VdrzGP9NPj<);uH!t5`k{^dwP5G{48L858uonU2Abmh<iw!B_O~k6mZLY
zBziU3XGQ1#d*2Y|OZ=`lJ2m9i)CYL<K(YSJa@n9A1snfaJmYzr&EV#rI(>Q#WQ_Vb
z=iAS=>CbTdaewo_-?^2wgj9C_#&%a%9Dn{gITa0E|Cv{oo8`M6NhP!vvm+-*d3WW9
zHrk)Xmh@q{lTV~5^HuEx!hh_uea1YG95xD^&^!^BL9TCgNa>J|_I2iP_oV5G%ydW3
zEJi#_q4*U3MXV#50RA1%S@bb_8sEm*G*(^W;*Jb9MHwDHScy)PK-wEFE%S4ORgwM<
zb{gVIQk37E_@1vcvlFG`oAO&(H{Gy!GqDH0Za;rolOcbf0HN%4q-{B@=4MYN?sw)3
zFF<AnR98w_XgQ-a;lRrS?ZEQZqs#}yA$LBDteXu81ZIF_c7^}HfP!W4-2xg+&oqQ?
z$nToFqxKT!bACiceIMz!poCWT43>w?tb~`fn;$I<%#ScH^OYSVs|yJQ!Km3BrKx#{
zu5L=tU1tL;9j5&k0FN%%AA9Ij3Ny4p2-?I=tQh(5t|QOK`z!>6g`M5q?4n2QM+*f*
zNcD;JG2Y2sc@Jfr#zaYagyIGt4FuI2jm7=fd;jM$8=04Rh^F4Dv3w=oN_enV@TBJd
z`<kARR8HJMgdT*Z6iFtE8^J)s@%*o+y2yO0<p;d@P@m3cqacQH1qmb13%Wr0KifrN
zKR3pW_5?P95M05$)e}bhdar+m?90v^%yqIdknN<ryIe0z!fWn-A8HF(BSHm5#l?e2
z1F=!ibwS?Id-1t#zKpSO<3bgc-Qq~n5K6}H69(Q2?j(#_`k%Y3BzL*diCBj~dI;oV
z8L<L?w9ZY@va(~Owdv>sB<x7&7@0J4Q(%fhf&#vOz9K9^9+;yr+I8A&$<n246fh0k
zMr^tsjtcAJ`v>zjK!<aaNSuNe<7T4i^njcn%+KLv#29mZ>KJN96aV#+j(HSvWTF(b
z=;!OXs~B%#&f@hTX)AU^LqiHByv!5#?AcBFa+x3Dlgw8iikuN)ar!=ym@zDjR%_0G
zXYdho2HhI2^$4RRPeEaE0fO$;b#VBYpE5c*_sZDU@bl+^r{|2-G$gmUaq!*vcTILP
z7x#Kl@oWXqjE$0;n@jQzH{}1bYFXc)00S^q-|r$A1;U?3_dU-RexG|3Ph%P824^Fj
zSjZyHr}Wk(uYfq>(toFy^^IUO?ltqcj@?X}iL`Lu<{3VBCWW<;4)`#K5j0I1puv@5
zIr$?Ia0ID>IkUGRtnCK#8($k6y({$`mr_DCga)8tCH7z6U|r*v{_L>+2FKBa+RF%5
z_y0R9#iZd<W$0s<Ei*MxIon$@v;Dt@;K{tfPf*sOi;|53TIQR*y}kc6?`z4tw`b#s
z1sR%>XV0gGIYSsW&NL<RS260HqM~3ZL^KAbk#+wrJZq8yF5?4|CI^fb)F>i`#no{3
zpT%$*-RGXhUhCleqX0u8`N>f9wEWk@-65lsL6UqRWOh7+x%wB}+OzEA+)Tq~XHGSr
zEP7hO(?X%3mjyw9(SJSv8uI)$1(_p#aGDll1INy-S8><h9oBE~--whcJR}Z^%D#Qg
z2!j^=JAKKl@lDN)X0BH1Z_E^6?wO@?H;<f8Rw{$dBj-ABNxxA-%Z!Ofy1-%X`YLhD
ziA5IYM9@7yx{`}RBC~Vif86|cjV&UpFl9F)Iakm^<_^OVlKyvY9y6zQtu4Y3;QiPs
zB!wFo82InpXf7smbGa0FsSPeJa7}MWFaO7Sn=miUhUyTK$V`;tS;Th#?z2cL8GXGQ
zN&|O71e^wT3q77X_0P=S$~icHS=n|ORyvFVp#FWLDf92dJa#7&bNU)sCQ!bhP=F&P
zo%sLtXw3WjbI|DQMPg3SHe2-Fm^tk;vm<>9pQ3IkuAWUbq8BP$e=z@GQtO9xDhsL2
z%g-;53Of~$Ut=1tSr!t1$2{KoZHWxOS#3*9{<p31U#_X#Dc#y}b?-8&^7=m)-PXCW
zsPZ@VMXl_~4;?SkpH5G1l-~SF_2Aiy-`+kg>2SCj43p>>X+P2+wzuK*WGbce&B`>2
zMHV5kTCI2WPntFKF^JBDkTjF7cf}TvXuFXEh*8|{P~r&6=*6mp&HvYl#d>agLHGT#
z&apq8Mt7N>Xv3BjVg2kle#Tt}O)p<G{ZMIyjSwt#p-n3IrLy{qCJyQO6{dEXgG&ev
zxhUa=EiETM$p*<CQ;x||B<c$2yjBGgcH+)LjM1YBhk>~sA)Nd19pGGw@-zG`rwQs<
z$(RQpDw2SNY;&>NeEj1tTMX}}ILO72`xK;+LKm38SF<kH#r$F8;F8QCxy<8dnQBn!
z<nNl1Yhu6+4W`OK_`?QybYA8!l(V@kHHafGC8BRvaq`WZ!Il-ciBPaRdy^eHQrf{N
zZILkz_b4`tDPog}N4&a@=s|+Q`w3MB4Tglp{l*HyXxhvC$4M_E%aI-Uzon0b@Z=;L
zfM8IK{v1?UgkdN9HvWy`0uoM33ZXi?(%%q7B}$5fT#t33Dc>8bn<#|)P+Vkf3HiFM
z1Rh%wJa)T?v<1EOb+}g%D4w{O@8q2lhlujp%b=@iZ^M!I1COzhJVvK>+$u`sYt4ws
zRcVN;46ifAWVBy_4>gyZLxe_}W{p%91gLg$2omN@V_G&>>0+aP@@K<PJnMawoI2oG
zsSy^R*8eO7<sQ7&Xp)sC@^~~l&1bVAIK1XA5`^JPxZv=Y6d{bQBBDg;DC;^F!U%G-
zUeM~1V=6WwPsF>HEQBTsE;mUM6vv2|i9>B|gD=vTtZ5B3$Fn>4$^|#y*%l?lx>nEe
z0rs?RaDf(rkjjS&GSDLM^6Fb|z{T9=k$*Hr;tee4s&qm^hb6=snzSf~NLIwf&w<rx
zPRNn0enQq%QxhhNPRM{@+8|nh2ay3)8Ee4}Okij<BH9R~2eojMpAH?}fvkB-Q^boO
z5h$^&HuPIHcd0E`uk%}=NBwl#RV(@h5C_V@gJsQ=rRB;rWp8_wNU+?%FNt{yC!0}-
zQ+PpXOFj5L8O3>G_zQ|*SQE#!IH0%Mz-)hzP8j!I1zpi*EgkOQ8>>6ZMid;G8h-$O
zGxu56v!es-3N^?s2l#=x?F+jPxe%W$he)YGB-Rb8UPtIeMt~{L#9V{i_i>>Z@@VaC
zII+H7CB(X+wq%g;7TJz?)|IS>x=$|sh0x$PUUf%UmCW4l6c4p*u3rbE6&x#Bzr06&
zc|o37k<xWScn=*&%@!c24Pq)Ss>!rHntN|M7wd)Q6WI@czxzgHoJFd*lyvMxhN+%p
zFxg8&Vi<ThS)uCkgDUja&Twx+@~h37VJUKFp~>KMM<5(oJLHi+9_HvsP$v1Sxp`$^
zAt>bG(H=3ipa@GnMw8AOy@eNkuv1Si_}R-mZ)Myvv@UL~vUiL{wjAi=Sg$XR2cy`C
zG}WJk$Bnr%9D2?){xb@7jA9|CQ~&P@a2$fQku`huH{9oSF4L29T}`ekr!l}@-vF*8
zLbbSwMkpoWWzQ{bXS}A%7Vtw@OEpMV0t}#AnYLDWlKDd<jce27{~eJCMilv@DrO@W
zFB1P?aaAI^jdQ=iJk{5r99Ldi`hZG^;-HJB#d>8U;Y}PV@et#X_{XF`;3Ks5Z?I57
z43AE%tPEmO)=DkSz*Hv2gM}4to*0cVfjtM=*xdHkCqLZ9oTkn`^2Io`rxTVy8rd4E
z!SZwK;|Jc4?P0Q~i1rZmskGI3JiX4`cwmN7UO4BAsRl?!W>N8`DC-X3`$QQc8HsZr
zDw<b8QYb7*=zo<t$)kM+rYIKTCDcMf2MehhmK6^*Aa60BwFqC4w{|FA+nm4T0k$>F
zLEjNvb{i3xBusKtnc=U;nF4vt6CkvO{MHa>26TH^hDtdf%cj04qh$?tK5lJKMtUEM
zbq8|CI`j)DM9F}HbnZ>XkQdO*I=IxyJ6_sCNV466_h)l)bisNK{e;JI>@2+x{rwUS
zXv_7X5=J*L<)2`^39nZmoh0m@jA9;zUmDkGDIS$xMnKDRumrRC7V8NWz}SQLZ1l{4
zR_4S>EVu02ML*!bndoKU<ls^z_PW$zWtzqVC>yuMm8(G{0QQWK0mocFbO~g5EPDhp
z44IaDVl2~EDuW+{qGTi~0-l1*t)jGKlqsDVM#}1#lkrc+<T?7(xY#*r5}R}(Fm;_)
zWR7>pC?+ZuOhc^Pacp4Eb;7I0(j$WyE%-8HnpE1r``X@q_bX8)f-O~JnJ!sxaD5_+
z;|Fa!Iof-$yOAs#hu~Ji;fCxygmgj622PB*i5(Jx`d6NKnhXwAB0hSZh9Oa8LvR&I
zup71;a`ZnVT<}9Sl4rBrri-k}h=XyebI9SkHmGh2(jY7qxK8fM%D~t1eVMZ}C=tog
zc7f0liC&f$5^A#rcR{U^VHAwSrlh{;y-FjN6@%~cj7#)ob1B(yXdFJwRIyyGW4U|o
zc`%X`H+Z_y4h|r+<=DK%o2X`?NJp~<qyR!9imW};>9km{lbskK9Dti?grFI{p1gst
zd&HRg2t(F&-?@}C8dYZ^|3-fflqkw6J6Icg=gPFweZ6uon+XZKD6LY;b!Lj}%)*Aj
zB}hWEoHJ$Sx1d0%h?=1qeF{^02<n-C4m2G22MdsWJ0*%}3OS5}D|iQM2C;xk_YrYU
z!i!5M$2p<A<Wl=8At5f_D;3-vc6Y~|MY6(IZxK>VoKxa)eHcsuGL`h+b!I#9`dPMp
zB{s2cK?X-Z2&B1Gzu46yQNk$p)3K(pChrbb`wl`YxJJ?GD!mowTLDO1al3e;Cxf#H
z@3^mC4oa?DSUW$P+!TnQ_Smby(-N-d8!PODu^^;juEbD<o4YR@D&2cBY38Y*%KqnA
z(HPj{$0Y0i+B^LX^ban~os;5@@*@sPL8cRunEu0@HAzA^SW}+Bon(nu60?s}Q&{?F
z<v`8gl6ec3uzY9bL5zsw2+W*=LqnPa$L22Ck#!}OATZ};9+#m#xb13N%)g{a38a#;
zsJb~2xEWbnB8p|l8m;Y8tcM$FvR%&!8!(L!XET)o1h;;J`Hlsug6dduxetu2tc8}c
zfJ4M~H!LDqd-UhnDKP&+(nn)huFQY@k5IFtxVIs9zaH=e7W~J{z`QygOwJ*<V4Rt{
zU%?Q39Ei&c7TL9^ep`l}C(amxLK=O`@t4m*FFLngWyz{o3e8H^s=(SO^vl_|?^q!U
zk`&R0yaLB&a!_|(BlC0Q>8=+J4xa;Mum$p9@4m>pUPS>%2Ev+iwk951nTtry`up``
z8?yVXY=Qlv0ou%zGkp^gv5mFXE5oordhtRSfZO;`iF7T1DR-8BEo6C=`c}EC=5Fne
zTjvgc-MN&IjU&RKQwcpU{E9Oy%Nj-2qK^8ya#6$!?QMiVTtdP{hYUN-H*bDQ66m_c
z5EH5~NnME{$-XU{kAhEG+7#-U>8xW(*AL4giFLBcFs0`}f{d-(0*c-3*e-|ZeD-x#
zaBhyzJ*ot(#(eDkhBSE$jNQl9iwJ=;iir=7Vmr$XPuamC@+b?$z<!_!rP9E=8G==b
zr5%XmG0}u!B!bSfh`k(xrP#s+7P=1R&US#5ROx_OZDY-UKGFR{sQiGs1{}6Tmu&9I
zu^Tq+ghp|DTc7;Pb2#4mkzb!XCx$UT>Sp!eKs*z*5JlkkIKOYKh|dIyu>+;&BXnZS
zK{jYLxA-J6)Q_Oy1ZDF_Y9WgALkRux2BwM6N=W|SkKSt|Dx0h$)$I7dp4)J{(;db^
z%ZXzppX7wUrCnkg)hzh#bN{TG1MemPXXI*JOCGTl^N?x<?e8VUGt5&o0Yg=RiX{m!
zHdV{LJO^zu4nb2TQmP5?BN7}?IUFQT{J=GFA&p%?^P3F)#NLMKae-UqLo}eqM^0+#
zo;=Hxru&J#S*S|<EMg<;$Z@xN)3#VxZLpW$r_o`pq(Ff^XYgdOXfF`rc0jb1Xzxkv
z5+D`bq+-$FjT}aMn|*qZE^93ct;Rbl%`o_bh%&SLq&Xlr#$a;jV|{%_$#c%_A@^T&
z1otPz?sVcbF$EcJHzW~KX81ldx_w&9I-(-Tdih0rF_T{jfpUPPFzr3^$OOyAA$ZLI
z)tw=vsN=>*`e*w3d7m)7cPAX_$H+n(`biX#&(ftX!$zG>ysS%ag+^ErfNw0vMPmn+
z)SEeCXc`>~Ook}#atO-W5-o)QKci1ij<O<XYaH>k3PgVhzl<|<6YfSLo_Ih0-4t?*
zR#p;t_ij=+zc!6{*{6j&j2kIUo*v3ADmvR3YY%25`_v@z;<MI<Hr-mzJe3h};sk`s
z1Cx(Ih0_RS=5Z_#-wXXu3rH@4wT(2?Fu-K!$k#K9A%dmOrD{DuUM>+iV;<l{{%X!}
z3*zP1ftR*HzM+W-q`})>tY&)_uX8+`$EDP1GQHo~g0=68;0<n)!m&ZPn`v}+clRBx
zG4XI#+F&;h(}_qKw4mlJy#tot+A4e193l>h_>aAuuZk#)w72oN4BSK^lI5ZaNx?eL
zBeKwhY@14Csr=HbYUrB|zb8tpg@iC1azS(3qYhRJExC`UFIG1|VS?kONiAeGzNSs#
z!c2v1W;~}nj(Z1O&6jI<h)5VJFyIYmx(vpy#3Yi9LtO){V8rf6S$qB`_LWo_k_c%%
zy8VZUDqEX}($u-^+asev{Z6kOBnKQDCmRSOG%EmV!%2MkLgUV*1iTQ7was2W^$qv}
z8{jWC(I8{tr!<wLG#6ToV)`z(;`K90m8oRW`pMk2#~hPVfs)6%E2rj+1t;Ex7?F<|
z`UxOcAdDJ&ITdEIw!gQY$Ej)Ft#;<Ip@z4>2m}T7Q)rUWFJSYHcd5H-8wUfO_pLe=
zONO#eC7Y|hG{b1H0pcVuGeXpSAUU33uL?55C?4m|y|-G)H#?)qI}HNYGl{l+s3MtD
z*1)Ows+E*N*#TEf7#}%+0w>W!F1I1Sur<N~><()TnKX5YMk2Qbl|)}P21!eIpctdd
zjJnBklpP);zfic2V*m-Fd?OGa)95h%t%NLE_D2=qXAG%~!TasCB%gL%ORLf(pA*06
z+=5Zv*R^3tKoV}!oD?F~183Rys6`JTQz9xtPv19W=%@Q_OFjqrz6tq0(84>2vG$?j
zSROCZ%AlLZ*F4Jy^||0sj=+p4?HJB)qAjNO;#gOy0rC7v&aSha`6T*vm%!vq5@&L~
zr^p4+JNN`l`qV;HiD$676LP%B*hpILS-@IZTjMBb1hv<a!sipnhFvL$fz3|Y4%zC$
z2CfL_UIT%i2QByk<sFW}Yi?vevflAp$x8)(nNw|71S$q#v588D8Xe2psq3)*ArzQu
zhkqwO>qQW#pp2b31A>o!9P_YxI5m<1YOFP%#q0Rop*Wh{{$&hr=j><2$=)lqO5}Hu
zSj{{{RMZK=E9M!~08C0?6*-nq*5jO>nJSr;hxphbdLVR{fzBBBkxJJmzte+De(eF|
zt%yiE4Js59L@wns+HV>LW^lQOB_Oe8*w~}q9vtob!;E3Hn3=F3ZtzVq{AMow0z%|W
zB^2L7z_so|Ay>ZD8TCiukcqFaQ04Ka(g?pd`8kopO})r8y*Q&g&ymBz0^mHRcT42f
z##^vT<Pmdnrh9{=yp>3kkdTCmR{^_LuGEGK{f`Hq$k^lHqA8RXJlr@$${xfqc|W?r
zL}0w&fHgpx2rHg3@)4)BAz{kkR2;b8KAui!&Y@M5Nn{3k9`%x|knG{L*9eo_faV}7
z6RedqQJEQx*V6hj;Bpam`SD0iA1a{+FL9k2n#vZK+KoeGW(OeiSox4Sff3M!FhF2f
z#u8l9jXzSWjLq#k*wQz{h>ZRm#kMD^CiDwGYU0Q9q%s~7?JE={Xg~xKW5Vkl1Q~>k
zUoS)q3m*I#v~DfAiWU}~M-RN(jHxz}jvHK?+b4P~3b;hxnr1}fxBHUiAavpMeMLwC
zJP?3KWaUUAl|nKhG&-bA6(CfX4|e#wr!!p+1Cdw=$ka{@`4foheX%t=h}Z%VOFV9n
z!i66yYaxf}525sX{LYu6(th^h$=-KLlWV4NziK9ii1G-+Y2<%HpEJXK3S|@0xa2Q{
zwr~(xBZ@K&iA(Z_Lx-}-MJXD`78C6ziW~jd0NH;3{}TWIU65t)|J=HGqqmQflHzU}
zzlF95ij60PC6@%fJobim<cN9M&dv+q8O<DbiJrLPA@%&&8}Zqxw<iL{*evcf9-x%+
zBke`xBv(U91$n=7Is4mQhwon%==oXnLczl|yIKpv3`z_6oo%i3oU9D#&=+|_>t4*5
zXZSU9_FPxv?7J`n<u?_TQsT3-9aS-c?Qdf@lWa^;VHdHb|67~zZ#ZVAWq-PTH?Mv#
zb0*=E8hy#SYbq0?0veVZLc<}@P{Y(}MR4@k-=AdeOIJup_CcH~!iS2yDr2B|Bk?%f
zp8vD$Hxh^Sf3@9DM2F3z1VzVKsl$$C%DgrgNwYlkQa$sn!92EgQQDE{)(Ix^Esst{
z=oLii%Gp$%^B|ItpZ{9afilU)J6;Qw6EzqB|GYP?xO+O~S0U7OT8O$?l9gclbpLGW
z?jNB1HiuGx8uYaW;qpox5K!3KlCNHHromd%)HE=BJ|+4^(=C3e%?^eptd;0~!mY+`
z8<&@-Fy3cBs}P};nbZguuF`ON!tV6DvzKA~?4cUsz_Dq5XzHnyUtJJKlu9_&2o@X-
zZ`Z-x$IPEqbXz3qC;uFOt26GzyCX0Jb^I2R!%S`CaO;DQub^RW9_=)EcQ%4CDP`fZ
z2^|bdSQ<#@_{i>DQ@RX=qTsiB)91<Q*f=}Q(+-DG9Y{uU?%v|+Ad=J2WI99CbJ_-S
zJb2fwc3Td$l0uVL>6-*YoJ_ssF59AmDF~e;<DvYiHXUA2a7#QMDG9NOH^^t3you-u
zIualW9P#BMOd!-Z>*#RE=dGJMNl`{aHdpBf<cR6ijzF=U>DR{PetT38m+=90yuNXI
z8i&`8`1o}Ci7%P6ZmI!SOuUp_AcEQT<<(-<MauS~m+NTb>5Y!ANN4u)U8!ou&1_1&
zb>E%U15vL{-Wm%A;<#YoNvV&aKvL=i><V!9=1`>>ywostDBqclqOoVO-dQDV93)KC
z=z)$=-e!BpPaY>ltu->%CnC-SI{kb@8wr--pya-8zs}jGPS>$5i{l*~=_;4%2$tJI
zQbC!0UoTKfD^N+AP8<7PghHoLX=t_#9FMzR^U_Qn2VVWYg+@m(5SNmR3<jzz3fC^Z
zdcp3-^iiM`h8?h0GOfZK6_+?_Axa+ACdri0B`ediTzI!X2_Th$H?*y{F0$3+BWQXG
z!Dl$Gv+cR7wIuZ~#ALW=AtM)rgtz3z1N=TM-U7OpG$_~J!aJlAznkvw|84f^M1ag7
zvK!rTIuwls?q=)^G%5{!TmeB}NU=w9yke^<UUpzfiIL*a_@_Vz$EeJil<B@6WTl{Q
z_csJOwmiO*ImxbZ4><=xnme~b7{h$xSG8P;Jy6{y;BJXsxWFwq&q5`%8oDRJ)HPCl
zACaWYY!X*?)XsF##VV;kKaI2+<U8(G$D3)fy`h|q#`34pfz(b+hQnz-WF24gwfQL}
zbbDYnmw14fktV(M16QM_<?>Jkg{gsP>soWA!UunPv~@=)Jq-8ee~NraZ_UXNw0pIA
z$+`GqMasL57`Z-1YJ)6Q{?`dtEF2~SWYQwX8J^FmigxUj$$>M=ARJuvP7z+>K`75a
zn$zHp*St8N`qZ`L1~j^k^RS>b6n|eNBxunOO+bb0_KK1c2%kK~S9aO7|HyUc2+f~h
z2>dQ(UQ^XBNwRSxzS*cHWm$|zE~HjFy6e+b{cWGko^!w4`oizq=4v?)d#s4a%Qh2T
z7Bu?y(?5Ht6`$`cpfFs9<V=U!wV-D=F?Or*QRabl67S2c*_789Te%@1Y}d;4$4XRk
z*^_PJHuAM9X|(0-<3qKzC!m-kzn5}n7vkby-SPG}n%tN6oT+c`dWAR$9=1u4FBhc<
z2A~x|GDx`%yrj>Mu|?}Fc?218o+I#5%`mfOb^(IHr|8r*JHNlCavI-Y3(YyP%jC-~
z8a>&j-^|Kl?_Otqr-AF#LR_}`X*MOa7=@Ona`wnN7d<lJ5M-;VZOY#3uT(Tq-#>E(
zg@*d6vPuN#+qHjbpxtgYZE6W6^b|6nM-jKdhlV;=e<39PM3NE=i`$`KY5xVnSHKD;
zk9+%3vh*IC#+z30|2hi^uo9F_Mrm~Tx)#T@Dql6JC_C0>IG;hNv(`K~JAi88P<;C7
z&X#nm1tJuSyuNx8c+Dh?Ux}}I)Hvl6`|@Ri1I$Oi`LJpo%FMxQEvA$r`y4-dhFcF;
zgd7Cvs@$suZq_{~Z85;ZzViMT`VZ7UN-b4b`n#y8>)`5;VgqksQoSOpmDqW#I>S#-
ztl(eYUL(<66=uL*dIa;6(KnIsS?mvNXT_nXB}NTt)hHO{$D(-?C19rcKW%isF`Bx7
zGGdL9;-gG^`hUYIWqJ;jrcx~Y9gP#;q_p2=qNTkZYDXgLkIG!CPa#LctV{imhS?dV
zBAjY=d&2hG2t_+;Rr&Urv=wzg#kA2RW=1Szyaxc1cR5~;LySg-O^L8llOz<oY@rRV
ze&y5xO6T9iEC11(w{})tGfJl{YBCMR>#6K8ol1-M$16*ydZ`_m(6ci&0AX_X&^PN;
zwS_|q_=<WORwcZclzq}4KJu-qYAhvRc*AOnEeAYzw<}jf_kDbJEEG<C<j$LO&Ob~9
zMPnE8$0Z=!$-LuLW{ya3R!l-dcQ(x<Y9Ph_){<YU0u!bABdOE9v4?FCkga^QfRe(C
zAVFDfvf#kXbH?N!Ye!oF4(1a6#=|d9#l*YTKa4SkG>I*Hlg7!dy0hN9D5a;#69xS}
zEE!|HP`Cvf?I(b8;kgrLQR9~&tvN9L4dwe9Cm`@H+0@Q0f8Xs6;v$)8k^njst}zOH
ziuze%ZUzKTAEW6}$Xg*nn*1y3X~j==wpf3bSz%LW70x5;N*5RZ+5zxOA@Vh`#8SRH
zx0l{ZM8`VI3~XvlEDCq#^VuIPIeegdaY@Uww7zw1_b*n#ociu?-UX2`Rpe8TAe(p`
z$wo2b(x!WV*j-YSZhjtU`GJP)Q`2svuGga)1}(vWrffLur}&u2u;EOBgrGCrkC2UC
zzx(0<@rf}|FTkAwYHH@q+L!{HbvmK93y?>#)A8N8NOQlWR*(~y?1g*l9QQzLvlTjS
z{ce*E=^g3X=0#s#sCUHVy99w9+Hkb5p6Ebf^hT|{v4YQ0WaAu;?{BDoKpH^R<zuK3
zn(0q)3P9^?1uUdd`*R;Pmz)=-2hwI>Qb{m(=%u##wCSL^u4iPVx&Ad1a~w}+s?#0q
zhbHc?tt4&=AUm2-DW?1Lyh9#P$zZ1{U7bZLOpgyG;DBXr<*S2XFpXWrWti>pk~O|m
z=Obg&#|V^()lPT)Q6m8f$0Imy{jzU*w8E!r;d5v`%2{V*v}M9U{%d9WM!swArpW3&
zMy@8c!V$Lr2|Sy2@|0A>={hgf(y6FLIMqxT96XTjf!<pH2)JxIf-;SV&FT`&&EP)z
zui6$|j2rH{_N$85V4jNs-VK!SF5<Se-P1ka(gW>!^oUemAQi>^!l#PU@1f9@TJeMD
zSDFQ!N~yYmen{8nxQwO`e7FPowL4IY7K0FMJ}Ttl+-N{KHK=nrrE5Z5a%Mz2)m{o}
zt_NZ-6o#*nwLPUcQKrskDG*7u7_Ahb7#DVEQz+L$A&<HN@c7C;twR~#U~gOGXQzH{
z*PXa9Go&<Yka_HV+YEmR%FI?-27PtO8^JzWS+l)e&#c&&F`no;x!-2IeATK!d0zl~
zhHAHuY7+_IcK5Y23UD?|hG)F4@kn4v*EY`h)mi12XbT_X$ps$P`bWIJd2F3IV;;{Z
zMttRr0X$BZgZbJSPN$>NM<Rra>Yaa}@CzMhU2GaVw`s_<j(J=3YxQM(%ezk`sAr!b
zNAFNNUNbq&Ng)_1O6T>9pQmzn2aP)v3rn?eDkXJ=j*Mr`Bh@)C&Hf0@fS)*~Y9j=W
zDv2Yt$n*Y+dF6H|*-3}(J}+P9eG060F|@<Hq2hQsZBWjk>qWeU<as!fLrBJM1Gr70
zH6Lnko5+w5mdwn{GK1W-I`>%qtwxY0ewrG$(M~7jU_e2o_`eQf*iTdYWRoi6()C>_
zq2~L6NBgA_zJFan<7E1%5ppcmfPy0F+5Yy_V4X|tRQuaZB0x7!VC>-j`?8zw-#-Ga
z+|*qbdc_}%L9i|}{Y&i3Tm2+U7qn|vo}Wge&Eqv<GwacBq)rF!W-HQ#qo~j)Oedta
z`3C&#QlB;SOM7?BNYUHyy~EHp;m__rKsKc~mK{OX3IuAS5r>hYG+B0=4)Ze*099Q)
z*KoA|nFqi2lZpuaOT$mfU+Vh9c1fYX3$3q(s2;y}8h;NVzQp*HY0RB@yqhBSwwOVl
zzy)9xXmo@y9mw3e+&$*dlD8DqE=bF{T1QM;FrZ*JlrcJebZDXbi^CzGAXQd(K9y5I
z=RsGw2*d1v0BS4gk-_oure9;o$u2WBu<=-=CLl(Cqhmm7ovlNCL$ng5G$Cr#wE+3x
zBp8g2y#f=b+?U8~r)@%Bt)$H_Go0~bwiDQ&ZjK1kFGyCZhPa%LsEyPf-gDOu#vnqV
zoL}_pS?mNQhUc;aBzAsr%;#vh7o$QeF>q+SXj=fXRYRQ9!*;%(o%oBf-4%qt=~`w#
zLdmd`)VXP(8##hl^#QSxU<$Jz8ZVhWLZUUKc~cu(WV;nkAp8NyGqOE%qFWuFQr8-B
z<AI15mYb$0c?#ATrAaUy4E!NJH{)W-Wf|Epi$)iyUaMai8(79z<G~~FFd}-UTvNnF
z(%L{KmO@#yn5MG(gyTp(Ctu5;Hjwrb#O?y$2OWsbfQMZJp22W;npx#%1kP~<X_n1t
znqou&+Ldg3_+F-_BZ|=ilU0H|vRaNAFscOFhrLhp+2Is_VLUQ@li^{URG$kT=JzH>
zy?luiUE+OiLL`G-7p^-3g)?t6Tj|@h(_epu`^Z52A9rfDU?We0fAqniad}^e=iStP
z>)8JE)fwR#mvH<36xAn|di(E5TW1ch8|H`q?~LtGeLj^erF=dcod6j#Gsm_-o6UVt
zdQP#cPEUQ7n}D_U#WashtGPw$^IT+tC*ZDRFbw+5rp6ZGC0gDcMxxTpZZb%2bXHsH
za$Qw-&rZXlUoeqBbApk6MO+$WrBPio1e3u1$#VhHpvIq`_;P?(vN(2;Pb(-J%<ZNh
z{JHo2l^@;H-+ok={|s+H?crF2w*cZ5$)WrF3X-VS)LPWz{H?9a5F$@^B-3SD46AZ7
z=zbbjBbp>;HyY&KE5_Rb^l;;@OH~7WRv+IHXjk+)fq}K;i4&uSr@aJ}o-w8qUPL<M
zvI&g*xRDv!2yab}`~9o+1_>t<Mi3rV3`Ym`r<xq^aXl!l(e9%rD*^aiRnJAU5}0TL
zexxY#h+nlVj-4?z)fUmJ+3L?da$W;Xr<d)fm1c(!%$rO<iDb-2v(^C|`Ke#?6Xk{+
zwuhtp$@^#Y>_o|=t~e4L7hynziz-o&CNI4`ZD_|Os}UnuJ4p+d7Zv&ZILggkkD42A
zL#N5S8mCWgY>|zX-!k;f2AaN=@G4dHKa4zcCgfGEW>k7UpS8xMT224IulRLBLVV^>
zqTVAki^iv9yy;@-E_tJE4lIj7#BBVFn3x~l_N$+5llRuGCN|({iIbw1-gs1{@DY31
zNVGu_E%9QgW4Cd}&q_`5ci-!Ah(D<;P?gPLW`6HOgTa;iuI+ZedUz<kZA*~ImHqnQ
zPfpi7fqHd>Z^)61KNW4BB0+cg%cVFR4dDJ5!1a6xD`RSPG$z1e<}!?Orpb;{xo(B8
z-&~LgX?swz!TiUiK5BK0f#ZTDzgM+8NA~wHoYOhaY)u@N=&OoJXDmUPNap$WAA=L~
zD4gw5n^E5~9UJ&2^5hrV&l+pCXLJUWv`M(vAJtChlDT#qywZQUpfY3R@$`->@L*v;
zwc3D+Q2%1p|14og`!EoZ7rwqW;4#rRH`IywJ}%kshz?F|U02Z|``R4rD=f`Bt4b2D
z4Q5i2VtOOxlqNMn?)^<dF?AkAIZ*-&4JOE1KD#bCmZy6V#4SZQbl;!LnHcff&xshk
zNMW=txpAuE+pj)dzvi4Xy#CQ?@838ihd1nnqYS@GSme$qoDwX5f638Gl-gkT>n-Y$
zQ6qxoytNJXjrEaUye4ps5UZKQWJE=&LvH<BdrWOPRBy|=(;84ii1&A>O^5{QD3>$7
zf5d5;jfvo;ls>}{fYgfP?OBbi1xZm{tWWV9-R46I`-}knLO@Y{p;+!X?%_Ii4Cs(r
zZfUK@Q6_zLlRq<OM;qH_r!!}Zi05lw`)yh;-`#zJgpzt!&+WUL52qoK`6RHvaJ8OS
z^fu0nWI~oaIoj)JmE4L%9KFSUN<QF%aQ%#<EU9q>r58v$o8Fy<Zh<gdryroe8LXH-
z5o=PONAgr>gq2>YKK*7QX1ndNziQxBm%GCx^t><S(ou1@^du@SLthS?{lxj-^I)V=
z52ugiS#pe*Ht$F}dx%%~nh<gbRTk4EPK0xL<5=10;H&3p_;2B#3Kv@o3`SDxahk%)
z7KH8N>1gw->`Tl*w(+x^>qEKzseKXNX6EgEt%VlF81I<r&WST{gddniPj~lRl<9q!
zV>2`E!}I5KfRtfTE^>k{cV&_Q>D@q*cyYx$Yg(`cYucGRLI>#+#h(wA3PX2C>U%C8
zkudqfA$+mt%2&tN$KDj8syw)4-9&r-2Amn~NS+M3x3lx`NegM|Pv8FfHIDCT=|hs{
zfYG?g8E>wu8$ymvPC?UYa)+02twN6@=o3XTM<0l&E{|VHm`MrLxkBP(U-?8u;25gZ
z8L9@9t@#1Jec8Tp>+!b!OT!s({)T_lN|A=I;KL%$j$0oW{rO`u9yPZ>rODV#Q@dZg
z&umte79khD*p#)~qPxoCD-dRMRO`Fv`4$tCE>Y<unJ=K!9y(IJjGX_nB-`654%lr`
zoQWUsHLoD=>Su(c;QNb=?zd+~t+CpIh^0f_ujuH3`@qT@*dWRC<(2p8Ktl=J&UZR?
z758+ka!;O4{+axrnC90Pf7p;$*d)aBfR4axl72)t=Fgst2AK4}A=1pTTD1@<VI#!D
zL%+(^FEIQoh@S{7q&-YQ03?z-o)~yB$P=IwE{fqijC}}%zb=V?fw?SqK*Rn|K>02V
zwI6w-IqG#@8lvE0S9%x8bUi=*C0c)Wugm4yTAZ(~HX)OH$O`jT%J?TdLIPq1q}7dD
zf0Z{Pr&q(faToXVje2k{UjRQF$-54<E!i4h;ea6!?L6k~na!Ua3kkY%bX%x1%EQ~y
z-aqYvxTm|*X{se;htM+0%KSBmY=0TMPA1NpXV$uoc?f)7Bp7%?<Bw1YsfICEsI(W~
z^<tyw(vPnqxr_fvLCyLh!IPk}_Mh~||H<0h#4MZjwnWN`v|X3TU1|H&RHF8rz41Q)
DTha)F

literal 40425
zcmeEuS5#D4yDdiA+_nJ)kycPtL{vm1tEgBYprDe$M9w)=BHhvkL<<xpt3*jEAOc0O
z4T$6{AXO+iBRSWdtML5ijQesQ?tQ$D(XFFu*V-$5Vb1w|d*3>%qOfW?_i_#nj#ZS?
zC)7ANmbh?mEV}&LQhcJfuu30)Exn?oZ~}kh&!l`d6@M+WIjv>S!NKcG{#)SR_S2Aq
z<4+FCiDTzoLVKIs@7Iimb+9zVEq%V72$VX$V)<|S_fOoY+oT|Ppir_fYG+1CjG3BF
zeWBrN9zi3TvXgO)S<D%=>?gf}wYoIT^sKxO!Nr>tnhFIfHYwi?`o~3napq@UZj&XM
zf8~Y_Py6&;U3X$h)^sD&eyjhU1X&014*Dy_zXbUh%owwD?N#iFMpx;Mu>Zlah0mLd
z{3|kQF<A`9b=n^;3&=kd6`Plkzjy8azrG;;(KAjC7k(*cD^(2*W7@UKC{t6bF=O@7
z#O&!yl2X(F0aHsE31Q)+C8{5ezBupp+H?Am0L8p#Fx08x!wtb3L!Dh+rSE*EX#%(g
zmu5EKij*GfZwk=RuMZR0EN<nYBI|no-MR(&Ki9c<Z?k@Xv!*_^qfMajuVw0s8IsO}
z&2zo=Rp9~>ZGn`9)?qzeU4>sBstQt6Bn;{KD+(ep3#S3bBZ19g-RHfz?!8Oa>(k1x
zGNfH&&{tqr6qR?dcjW)UPW<2R<^Lby|CLq#|DF6l?<6B|70=<;TY}-I)3t@&){}TJ
zhk)TC@7(05Av@^3rZpg9=Gxql2<zZO#puol<bu<cY9k!^<zr>6nX=5EpLoofb3xR}
zptQ8KRwaWewcz03<)JIdh2k#W&$Y80x-Q#h)XA}i<vMD0tc5RaCtpqdf$y@0OQp;^
z#&gGWGUujxFwER|p7n)A`1<I(#rn-+j*gCPTzL~GX;IP9I}9fN<r=E+_d2wKYu_1K
z@3%}*f4MoA2mA|h`5|dcSy@&3=LrOj5CJW^i|r}@tziPL<JnBT>0s(~236?VVRGNq
zYhAqS99XmCb)_2mu7-Ucqcs|_(QW0!n+I~n8C#@XY+fpm2{f~CA^-V34`*gN<Wy;g
z#KhYAtf}sO_E!t;mFx)8r3uvMQK^=G<XR3N)#fGlUV2Opky%mf7_J>{X;msxy~B0-
z?v{Ns&-rFe&Kk2vy|7Tk<}Xe@kC#qf1yT&#4;?9Qf84&iMdBc3=<8Mg=<ICwa-;vg
zFgC&~%PDuHQptZ}X7Kxa`*%yY1SqSw9JzlX;o_<W3vz2Zg7|9n`13<E!)1tfvXiZ8
z>FGJ`b;7!DagU)wLC&j($aP+K@>gszs)|vr&#*S2UF-jrX=>%w7qOF_6R!VQM$f<O
zW<9X8P@*H$zB)zc<)vdcSMfY*J=>ggF=3G`4;jK2q4zz#K40p}hkqJ6q$DL@n^wgv
zh#>jq`g=0K#YD|>Qe`3H1CDvM!6biPzp{*eR5@Kcsi~Rwz?Bc@_+Hi~SU2mvv8X?K
zk9?nR{=18Ja7l=G$3qQ$3a$8wXh+-{%^pXrv%1u3keuvFS#PemuI%vdqhkYOlB3n?
zqW0~@0)?2|YxnW~1tILa^W!!pA)fi#2E!d?DPbB(+BE~BB%<3mEEG9gB|ey2v8dLq
z<cVmk(?)jgTN;$VqB>D0V;|+9#n+eO9{JiSnf4vEr9v9XTmE^4A95U7?#*?Yw&jS;
zpM8CO6k1nRY;1INo3f{zID3HG{DTWAt7Qa|GzNUC7<X-3J{kV_aEr3KiKD-_w|Bgr
z4B=wAP~^UVR-Tt<-;7b6=t#F-Re4oafS7shNQWl5fW8{{uybh+J8<>W0Z#*SZt~&n
zZDiy%^n%yVw*8D0!zb6D{YF1Zn_EHuwIf^cE9~hKM#Xc5j%syPnV11N!!Ln%xyAMr
zKF!yOKR@Mpii~tmU~pl5y1B;K05hVsyss%HCg$qYTuDhux(t1Ke4wnoBxL%(uM|uR
ze_B^>9HA$bqLteImoeG=W<PSQ-Wf?reQ))zj`h~3r{=G~=lhSqJ74BYUHx>o#cdTE
zJzN^Z8RR9JJQg#**hhEjYgiY)3|C$k--$`Qye!}6K^J!^xW^k(r79;w8TxTYBqlpM
z+hE?<I~VJl*gRC+5h8A_LOT<DX!m>HHFEuRWW$=8!u)u|$BH~K7{`{d{1vOLrpNla
zy1FQ|Ga-lj{?aenN>I)9cRL1;`!r@bn)gXNc2}=+OUm`GoBzgJ5a0CZmE8|@Y3`ej
zOs`4Mh$uHF)-m{r+*jv~MSR$_=wQ}6_R<E)XS17K6#Td!76u3kQVv^uy<b{dy3S38
zOhG>aQ%JUMmZ|Vf+>yfgk#v3!PIz`%-d6Ih0{dR8=)%InPxfE_cUc}o%Q)=9{Eix|
z!3Ec!uXB0ojkp$_lOsbp=sMJTNIYUfb(!b*V6sj||6l*(BFn}%!Zl`Mg|`NneQn3g
z$?!f3&3`X_<JeQ%keRp8T1Q8xZ|}U5?Dgj2{)+Q7@0eRo=U2E_+*%kAAiyJTb;j7(
zIFG)M{8-@++<8>?l5kgEeh57)=BZlzwU$qQvtwE0+bt>>`f$0oIHfjGtI9fs>_z@}
z?8PO|MFVN3Ra1SA(@Pn+^=M-cF%6p|xTU6bv!$hF$;8i==|0CY>EYsrdY@V{+R}U2
zrS9f;KE(cL@;#f86Q=soo7jMJ)0TxIhvKljn$p=!syO8ocKX&j!JU-a6oZKV{(dh{
zPcjA~_0wactLOoOs{P}2134cX;XFuIBhx)3LjE~;{Cz{Nr+c5YX=Rib)pOo^mWJcT
zW-qM<cNPYVnNQ7*l5_su<>_S%w8+!PxvwF8c5aFc{~Nvh<<hC81HJXB=En0Td<maL
zPW3m39kIQVmp2!8?zx$`PaP(O<M*PnkM9%?c+rcC3F$oamI8lObMv0-bkzZ2MtFEQ
z0?Fm03^KLI2g~T?*WoS!KAdZIZhkzvGN)9sC-J{FXNs7RN#yvzSa-Is27607so0Xp
z+-YY{p}}PR9JhQ1CmH7)8E1;o+~@gw!%-7+b0u|kb=ozoVbPc@S%aS<4(*Vp5cjFO
zn;sxU!C0*sp^G?Iul(vx7Jg_u4ypW7Q?s(ZUe(#8i>X#kSqnTx@Xhs8{8$KgA`kr*
zXA-$|r2sBHJ0M@Ht*w2~Yi0tM&2LIGHsz9=o7-z7K7qmyx7JmClYS|;D%7!7D}8u{
z2z`I#06`h$Ca{(LQqKKnqSYt|2YWWtu6??@+0@cfnU<)XzNf~_F`u~(*F~Nq<1Mvk
zS=pJaIx!G@Gf3ZUceYa>_u%+qd~LEG8}XPin_c*H!z!Q4rJ)k(!!Ir*ENzn^ljQcs
z&c)>9<T!;NGOdiHcYMpVH}7~Tq_O&oI~xWp#3IhD^{!jRjtSn|;Q3wIKF%)}4zMFz
zHrvI^sbs32)`P2SS2NB%KV>*If@p#$W5k-9^%=^=jW|Y$J5=$ws7iK4-Gv{OGcuf&
z<<qrGx@!_R`#i~Hx}RV&QL(XtWYKn=V^-?wRgpOv^V>Nhhnaj%*2`9lKTli9m}}DC
znnp5SIScpxvoXUuC5+RYWhFA7P1qI@^$%y_Xz>2a$5muTs|5<NwIu8u^CW!px|;xw
ztIOTTd8oC}XNsY$tSmA9%|1XtPfxFwo!>_m;fALl2yxS@7h2P#r+ZU<=H~EAj?q(C
ziCXLJn<wTi%^DkgKUdsEP&xQe=YSk_Zl*zt{ml_u@(qKB{{H5w1ermL{eGK8cf!3X
zH06-PZEZmYTSqpN0lRNu-Oms_5X~v8lZ;KQ_O;8ga?QCPi5TdAX5+rm<G5!<Ma2WS
zg}}#NW}o32VsuA=sh%DiL7rYa?d-mcPDYz0A?+1{K||)}eD_){W5bET@08U`7>dzq
zU-yzL3`tDw#M1x$YuVo=6_u5=B`Y=xQ1o5f9wus~#=O)b7kwe;RUtm85-y#YPoZVx
zjwYoKr%-2m77CKH+9i(NIl;`=X_c>~y<arY;xDJ)EI@wPwiBaoQuM}dmMNQ?o71j!
zRYoJd>~`3SZ?C^b&^%_r%$1>3e)|>*OqW!yrLCirYV-YHfx=wRX#%^rBuQvEss(4|
zcu`~4w{)N@HX{7;`<ol2&)HR+<f<v1KO#cJx$}!e@<Nl7lLbDe7<bl)54x3;te^iA
zJF{SA^lWxcDG(SN5Wcpkf31C;7)^5tlE#7&gu<Q5du(VL2!ea>Y!p_c0j};ZdL%65
zvYpIzh&{$f9y-OY-Y}@9bdB`jQSShJ!x0#OBfEsWnLd1WIbB27_GI78a0E-b#w<PG
zlw9zIoWlO|)`KdcM-Hsmxc9W(uHz}XnXj8~vA;b@zFqG;IF>c^MaaA(f;H3ecL~|o
zv!`G)*K>uqRpe$z779wx*zmw+X0d*pJqCC3ah{p@i69{C^6Y(jdb%M(0)>VIYz!c}
zM3p%F*{usjsxv*O$J+9}7g283!c$<}zD+-73st?;zU?RX-+@>a#`6Nmut#seO{33a
zyvaFMJ89z)0{)J={zgCBG1brHec|nlc=1HkScOIsn`?J`?DFv6a7pAAG9SaY??MFX
z^qlqjJn;VwM+iLm^<zeJ*dbFj?*MwPb54(*si`T2_RW1DCq3sKu3@7N$K-)kjAiR~
zu0K-PgK(gm4A?GEI5#_O58AL%dMz%?FNdLq_3*d|s)p+}3s@&2xKU>cs3R}gAWj?!
zD@^FUe65?FUEz)=l8$DyYaefKaBY8LJU`1zCOGBN(WQ)6R|||j{Ihn6>hM%>NXYQ_
zHPp9beT~+6IrC?A6NtF`D{h_YJiri;pI%_@Ih#!#AFu=17o_xO59GeTc!tcIr08x<
z<`KzifKP5wqce<Y#`sPOV)>9NYeaQ^|MxpAqn|Cn_6ZlQVhBu@`^uh|V%iq)c6E2}
zu_2~;{i#y(-;x(=^BZexAKekjJwh28DiZ0aPzd_3Wg|543bow6sn0k0S;4Ghy5IfJ
z{}|TIon#JL*2u@S9ZClQ7zGO3OLSfvmE8Hbnw<as#pM29wz!1+XQ>7=KOQWMCFZp$
z7zcONQ{&K?$NQh=Q|^_yGp8AMx5`LThQ8SgrMi!Y-w-5mf0H)3E(ZYf?&g38{0bYT
zOJ4Cv*j-D`8EX)rAZo+kwFXb%N)CDOAV)I<dV~)RTA;OzyukpKHErIZz^50CnA6OT
zrt?=Jz@-l(<E;p7{Qzzv2x_E4i+bMDwpTAJZQeC*0>XEnf5P{2tZ|C*W0uh1`G~`w
zlKSX+7v^NCGdWHQjdQC@<Qq3K-^dCqI<}#)(Y)heipgmjl2F@=gk(jU0|@UsPQJbK
z4*m2PcK(8@xJy85cS1^X7-6WtMwi5b)kinF1nt!#zurOF>NRnB9s+G2Jl@pJ<5HCS
zY#RlkYIrOo577%an%sYW*0+hJHoURap<V0!`f4+F>Rr`{zj<<IjMH3~Dph+vG|Ky6
zt4x@=p7q(z*0d?jvhp5FYj+(><@eA1WzCcc(kTa8j+6JS6FtL@yHj=6DW)zX6$&fn
z{Z13U{gg4flH$39A&?0kYu@3@=k+$9LQaBAf)^9*F+C}7XQC`mu#LQT2K<`sPDgFY
z(4!tc&fnhJ5>`qR02701rw17jKVOfKC0XsvX`KUjLsQz98Raxrlj3z@e(oO;w^!E;
z!=y7Oe%$#5@Jz0;=ab)9T^&e#Sz4JTg}5*3Hqt3j_}Z|@t~uX(-k^5f#L{Eb;-lCo
z{qnzmLEBg_eU8O**g}&r_9kNoB`Yh71nB>9d3$t_zM_VP25m~Wtla$PkB{lY>r|!B
z{hAMAb7}Q7Sg4-M_rKGJ15{<T{`YT4fW&c7Xl+G7v8k!4^JldbXwfmox;d$XhYs$h
z_15MxP2Sxl2+4Ndq9H!~Em^e-R`R~?kPADVl>dXh6ZNz65r5@NwBMuTit%mcqh3DF
z47%%PSq~RlYkX`t?0xsVsS)}V)|K<Fv7sRtWOjpeNxqq>X`yRJSdaiw3@)bV>;1YN
z2?v$+^-uJ&opwk#xM#%2$2Z?v8ZJZ9`B8qvuzMp-c~T%@3k5ezbLrX~>!U*%`DM2r
zIDrRAIvKU4>M>__+tA*L{x<;O_QOiAA)lq&KVM9uVEbLs($lm=gFM+3`jF!Mv9ARC
z?LN)VXjWdG_c8m*$J@UK)+GbyO1bdt9^9w-GSbPbKeJ<bD2XtaY~XAPQ^_VKfuL-@
zocivB!CyZGKjH1ttJn|R9_lsKwM13arM)yPq`h5UE{u&$j`HKsR2-b0jS14BfehB&
zFy`NM;K~W3vhXkgQ)}z8(lDucQ@><$NM*qd$vOrDn@Wp%1+n5ad+SY3U}Cq34_sZ!
zm>lk?`vnI40Ffjs#uu3TZE~4rH0!;(a&kUbvln>cIkPzoOd4RECIC@luM0aAzXPBR
z>*R5XRSAQSemv3@6>!kJHc_AuK=ZYEUGlGJ=t6FJDi7(-zN0iLtmy{<=ma}EK7+s5
z+uQF{3Dq1t^bsRLD*5FxKJu8*+4%F2=EsPxzRoP$gvXDEeX7r~5vyUE{4z2!6mDl3
zc_P*P(o(LULSp0nN_VyIleTO~TPV1@7R2eG=hTQm;U_<yR4Qxk*R)79@CKHwDfQdp
z5};xI^I}3xqJB<RGaG^|eT4Z%BFi#(68+3zrRl@NWpW1dPO^yD_)Ej0M-BC>7#iTp
z+UYOmqX;2<j`^z@8YN*;8BSwt!6cG_6X$eu9(?oWO*@1*A(!AQVK`z5$VUMJ*f7d}
z?A{yDqL3FV4=*pkY#-Co%xsU#vnyhd`+8H$RA>%PPDa2IzjCF`V*Onb)6V`5z^OH*
zBJ{!Izs?N&Td{PmiGLJ!$yU@Nb1?<GoN4EE-WkxEu>Xf3C@Ge%V(^y)iwTvK_<Fgm
zW80w{0BGT8(>p2b7(7N2A$wlps!h@%8UEMiZCWT2sLSM`tH-Jc+t8jy16%+{aj{WM
zem>6U1E8lR=YVN!*rT=lr`AhXl-sgrIgdR@!4&^!1R0l<k_z~mH{B?0Lco*BVhsD3
zGm$rJ0~<^6gDQW3;Bi(e!uHeVB`OC;h#K=;!C%F0*WkrH59cRC*lOTc94_MOxDDXi
zFF^@2{uTIHSpAirx|EbK<Ko8%f${-j?D2CD9zJ|)_Vh%qUbjY@q>yf4(;XL*E}DCf
z2db+SetmUyMYu_)tq+M?an86=HJr>Iky)TtbV~`Ft+PoJn-Wp9$_76X{UBaFcBjP1
zKo2B``Lim|!&$`xDH9=atKPs?5DHqldEG{7_C#374^fL5;;{ohwY}eoe{>`I>@Ym}
zKBQ{=nE4o6Rw2}hX6G|r@@qE6n9bQ}wRotRa$X*|?S;V;@^yDNi&i3P&eN-Ie~gj_
zstm)~`6~Ro-%4P0<8W6a;D+b?FtT5>0|ni+t!MK+zxlA4gX#PweW~v&+pqM~%`VIv
z)(A+rqDfPhW4b-m9$TNl6HLyeOg4+HUWc#Zteg9qM2aeYeI!Ia5+7X6bGZ2SP4UY}
zoKT^T)8twGNLRnGFcon5Ad&xoDFXzmVpUa@l$74Bdnk{c05n{ky?|`cStG2q7;sNh
zN~dhFeN?aN5}U%E6~V+edyh|yz|wzhoG+H#vcS9`*1Fko{@A$i<1<Wx7%zVNa|bc!
zv^Q^dN;vrZC%fn0f+bvA%xDos_$>~Xkz%7t<GN(M`3<u9f%`lFPqO&-_5kG|Bpz@g
z(2*}6*`i(~1FP6h^ew0jS6KZ&AEWfxH$%wSJg#TKkWc7){QL-VE8{fWUXl>DglGbh
zTgeYS#|M_v3;KzN19kkt7CPeAVA7u<8kxbUzxv^BW0$gob>3Vn3zwOPsx~*soVfH|
zo;u0)B&IG8jiOH_n`&m-8O|S8DDbx8v(!{k7``g)%#(vV0C#u0JhPk61aX46!H0%)
zbEx&SEe$)0es&@LLNcEZae`p5w3p#escaK)VsGRwDKaA2*6|6##>r{XuwmOk<$jUA
z7N95!sG+CzQ@dU=o&VTQ=qCI)Qah>k_%-m_(jzT*ZwT%YpHF^nNF>X88`6ccod;Cx
zOjuU-sVoQ84>Oa)<{icno^R%3tIA!)#XHKK{jPH+hP_4Zij9xo4f2m|3=F}UzjPbz
zZm6GYS|<(w6$&Clp<%yZ=g;Sn%RS-?<w$WD(aAfY+TXrRCd0|;4)N?=#8jInJ0eoT
zmN4cZdE>KDz?5xP1e*W+$OY4kQ7%F>xWgN#66(xw{UtpIKFnVs|Lyh$A+@B70Ro_^
zyIg{&4!Do?hI$MY{?ZNRfl2{3bQ@4WctThi@Z&<kcoE>ICvsFT+IwpX2|a;WjJz&0
zzh|j6>vaqPm1M(jzGbixqU&(`25Du_E=b0);cmau4Ox@?*Z!vLRD3N!fm5Ekw*}N<
zv<<eG5*8#}`)SkJ|LO(Mxj28c1%RsCEI{_s?FVu-r^=w0in7x-su-0F#xERKfM~NV
zabQl*+xZfGG|4(CENGwJFDTAli^b$n8+JizKV#@oMs+Eve$5-+7BcHNl>S!`gh&YR
zyLC0OC+}>m`Bg+Q<RAR^_jHK2?E2I}YX*HKpDexkR(BOZU~qk^F@!_QhQPkY46b=9
z8zM-40DRlv`u*R(m#C<*U|KfwCx^ImGfptV<27n2oEH&~>ff_Xf^24BzLd*U|98*U
z2MRc!6`}v7UR?goZL%$h$d3Yrh?hiQydn6)8kx1|#*!6mWbgX@Z?5?e)m*aH1wR0N
zH#5#~<}85pq%Xz#!Iihm<`10Fb#Z`!ZJ*t4z7u4djiCVMnnvcT)y%sEv4)700Rlxq
zBB}}s3iGIuegKnx3}KW|ZVK&ZOF;*)<t~?JpUS}niuPsirC`3p<??Gb%x?hCI+x4;
zxc;~$SP9WtKr7AU9I0xC$JuIDVG109NEY(N{QlBvQU(YzOH?XdHj3znet+*PNSN4n
z)Qx^oEmAL#M=YFvlaZ5@PvI^jb-b?Vpq&zFds_aR>eZ1AC(w&sl&S>P2HfP5KG%{t
z)?jYlL7lCm3jbRDPFy+E`P{#@RO!)PK<H177|y8p@d1c&rk$OgyXFx+3JsTkL*eXL
zv+ybSe{yDfumEsRp(=H%YCer0k;JC_0}il))GNpu->nl%CJIP;Dmz{0pC$MD31!R>
zL6AaQvG47lC@$>9wxvS=YnW>RQ!yUxu9=V7gAI%IH*NPCOO2>0g{o`n9Z(L<tS31;
z`G3k6_lR2}EVa)~52Scb^A5NucvsyQ;G4+jtc14LS`JCrzzYo^WS>#O`uF#jMkT@Q
zG+;MYZ#;i)+f}6c(q{Sx=B#K@2)NGQY=0f}777`C!0l#2au!^+vxlE=@^s&f58uq6
zO-)TpREXQ#K<d89;q7c930C=<t>oR~m9i_78DBqt_$Yhq`bsK{2ypZE3U9q3_$<Yr
zI@6LqEI0FB4tGtRmt$5V0KfB^Y2iabS<mqTyZC*R=>&7{KuA!frDbGvaLP>_`(=p0
zVGA$pdHBt9yeZw<z`ZYhiOPF(Q!}$K#&Xm34H?#%ZI5~zGg9)|%hv}h^A30UaSPVH
zgK`j538^x>`Id_?<i9T;xeP)D>T}$k=jH}jx$_%FiY7-%>AugS$*~S1mvyT@Yr?<k
zn@4ACF*GC@p^J(02EyUZwVdF~Z_Z3l3az%bmOdOd0O3KuEL6gN=z+h7K;hFGZtAE8
z^&wQ;8?hupbEE#YWU{`@>k%nzP#8b&IbVa}p{a*wQkB(3s*t-Ww2la>YzD*4e164#
zIPV*<>SU)}h=+H0{t~xr{rq5ZC@4>NdW<>O$83jMy3b}ihL5aBUoXRPpTwrlwJrhs
zGFxdgSejS=+m?H%#CuE@8I+P{fl#|IyHENr<v8@VDcji?dadi+WJju9+ZT(bEHROs
z5j9dGhc6me$DMm3>*)sjd5CCe2u{9P#?7JP>LL!o)6GkcoT3g1QfWYEI|-D3W6cat
z@n{p8tj)-M_-LbC^g<4wShe_8R1Qi+sFOv~Bjxg1$gVGHy3M`%ti6UqdKo!Vs2&zx
z^_O`>_7=i<5mc@r?=>EO2I)Cz^?ASMIGP!^mZQJ~qV{A%UG5`)&)%yNHqA%3y0$*>
znC&qrqt2$V!~mgB`g3y3$o14F6}?|07y3&gCCZEndYblXzWftT;cA)T^@QE{ATy>(
z!puy!M&1`BBZKiCg?n3`5E+To+%l}0m)G1vk+e{LjZC2LR<~UwmkP>F7V5&_j*D;v
z-pyZOv&E=1^hw@KQC=}u?%?0&6KZA#edUU)<JE_KsT00re^({JO`rO1{bHHX^e9>>
zD7U?ChRK?y4*OYq@Z*%dWPjdLZLL%i;kBcl)<Jm*6qLj-{8fTnA4GN*ipi1g@h1lj
z758OzKHIDNc0a(F67nW^j(NmDe}8F=a>#4&c<GtJ_mgcR)F(3TV*>0tmfUozToG%I
z$&!L1$Tf^(am=UQMHS(9mj(B#8kuur9&^JHW#W00B?5&d;>=KpDiVBigT5=mS8{Nx
zySuyR&W)yYctiJ6H(-rasdnVe1hHrUG95R0tUg2A5YJz?H90$53V^K)$dB|hSTN`k
zjc+WA-dHxiM#2P)&+Q-QgY2`9KKQS$Q@vs4%utaifL-0H8bnP~6y>lOA~X^)ew^WM
z4Fer8Z2R(%Lg+@j|6Hg4Q201qu+D&*)3K7{=(6AF9D>axEWNR63MSekg~s|RH>W_f
z-gj%fUN#qUM7|<jCuLOD080i?J;l=kGP*0Y&mdyP?Zr>JdV7m8!=VI&**$WzU8+^k
z9Ryqfc818=v9B!i?6==6<<xt<SQFhL#iC`c-+lQ?iK%W7x)_y^(DPUytJq9hW%?r4
zxt?x7Vj3nB9q<~@wnu|ZM|+5M>C|XXz?Q7G8RVOG((aRbt&rB@wT9zTGb9qhNB}}o
z46q+;e)kQcF@<LB(Za(HNiEbFfBF6t_k~Au;?M7yga{dCO`UE`_Xn~*7%|&xJd9o+
zIl(Z8%F70FNZRy1OY0&M1CecHCi|W9LQoUhLFuOtmr74G`(%E<$zz9Bo$*Fi0jbE6
zW1b`xf;(TVTS4)QP@=Z`wAuI?s;&|M=dUllyRjhi98vm^3HA`@vC8gy<Iy3-q@X13
z)c@^C88X#mr-EEC*{Q9&R&dA%vg8+qMz*)NdvqxoC^|bk_h<Da3luV0v&>isfy4^^
zC;WE!BxonaBqV^um!wlU;?|%BH<=1!ooN-6i}$oZl$jhxx6_%>BR?JU1}K7If#Z3s
zS(3=@i-a>;-!xh&K_--xlpKP&OCbC@TfO~$GZcioKuiZ2ngp9~UekT9y{`)Nht0vW
zg8(*(9#KlL6GRrmEl>-Rv~cdfBmog<=$5L-b`1kh$K0u!ly*Q~C7Q>8JF_UYOg4na
zntlM+a*t$kTnEZnIO0PVIGSKkvO6}E)@YS2Oc{F5H@6=tuCOk5EM0s?d8eQi0uiZ9
z4B@8j$TP+0Eg;^F)Fc9Jf6Ni4Q~1`l<sc&dnFv{#RHKqFbl2g7&i&sA4qr0&8^@82
zgk-J~|8`%2&ogc870I2G#m2G$D2(NF)oq51CJW2$Xd3Wxp6gDT+XLZGiN=m-NSE(C
zI_3Gk6v)l}yvrvyr<}t;l!IZ6w~|x}tm^QgqfCx9GgCS+qw)wExK^sss=~ZJ;+Ws;
ziiFwrrel@40}T|aa<dMyxPM!Z0n2Tpu=e?pe>jfvDc9uY^(<Lh!kRK*C2U-X7-HRi
zTPPX1ajJ;qn3+)nmZR{VaKKQKt3nB$XfFJ6nVB2(+PLg#Uas4h-2n&fTHikRLOtGm
z7(HIQNWrZ(S<n6WKqHV0rYnAjOQaO7O;2zT?x*KsvKbKg5mCK@NvG=1%)x)}Krs2z
z()`Y|tMkbE-xd5M#jX|{F6G<VU!N)%G5xKB9KcT3(KXsB+5K7T+?pU_NDa_M#hV<|
zVd?7AE(6I8&bh;37;1(8MB|?zLWO{(G~!{dgYP)~U5+$>ZVbEUjk~+DrkN2CO2W(U
z`t*@tK%KT`35(>-*wN(IIAx;)j~#!iSrNK<UXHa%QX`i#U5EF5ZY``Lvd%pq*(GaT
zdK$_2BvBGS@4HGsk1T7l%ow?UiOS)ToO8{6i0Z0dLjf}_a&uSv#K#mwt-ig5>F(q|
zb@wy~JFFkwxM2wNt$uUk;t;;*(@g`Vq_|T#YeYY@SpSHZAGeEvTUWG4yJ$oq8USbl
zNl8g0gVv_xJ;LWfc2&yaZ76oBM5{^k+T%+^;v!gNWlhaP@EBuQQ_7?ya!rwf+*F0G
z=V;B}iMpL4sJs)KJgZZGUjSQ@7bG`ZB`4hHuhhIGG91#t@7@`6mOv2(v)vlQ=&h=e
zt;JE22vC;C8fVEOkqX&lJ?$~^&H(O|oB6)Rz9%We2z*?@BM;2W2xTr(nHPm8pg&0%
zjZD;%@lI?ANhmP=MBU7Tz+j(AX&^gnq>+FekBRqur2r1SgME;gw&a~!Nk`!R9E-?K
z(rMvS-+0d+9VYk7hYLuia?YD>nrY0NZCv|w+~npOzVL*UTuJI6C-puG<AGC;%ndl#
zY%TT^XFPWA&x{>=1sPS2Iq^x{ZmKJm>qtFH%?pSYvHnD}<w`VN!P(ZZcM}CzU(fy|
zYftm!tnL@O4izV{rwJiUj_^3<YWjO$|3W?IriUboD@eO~R7O!U5g+aP((8<nM+8T5
zF;;>`N~ws~&pQNEriWbiBwA<0#0NpU60y21RbC1N{I}GZcst{GCaYs*AENh9P$Erz
zHYUXC3z*>2nH=;L+A%%H+dvP>&}>mvQE_fs7n*fLLjxhMy0#xTV9S)?dY>0W#Qz2d
z&tQ7F&rPMUvNqD;>m$fO%+F933<e!QssD4VuLuyAVos`%lu}1ua0mmSRnoA#I({?M
zJQ8KmPY1d30Ktf#(5XgPB2Yuw;{#WhPpn-T&d4Vi$73=J`2U+)c6?0_0Dc*DVhw8>
z*%Z#+Q+)stb|_AFz%XSkYEId8eiw~ES{TmGo6fcx3I7wl=LY;s7--*YlVf$TLV2Z}
z1SHa!b0pOQb}xToJl@}gP~;U5pdz{Z`6)liZsw+V4S(Rh+>vOJnz8j6XtJH^Nh#x-
zdB+#nz9tc>7HP<UWAeZ@L+lE8om<1*`z3nE%3d953sgUd+*%wWF47V1ZiV0%2wFKl
z9QCn2V<1|3v^r(DX<&|uN_iVjo~$NTb~K)xywBmXcRrl#sFu;>q=U{Z>emi}qg6e-
z)iVL)m4<m5o<gx`soUi}{;FVW5do3C%se^DLE9f6X8M?OmcW-<gV+2l$yo=;O$Go!
ztw{IRvo8^w{GyRd!*0VGL)0Beizz8F>cg)|wFOu?U9NiE{PxsK<FarE2Zw5}es0v<
zKl$@Tu%+$IfvQN(h<^b&?&C(Jf+^nLx#ow}`Go@^gVeb)VN?k%5tHOfIxBpcq1kPZ
z@`{ka>}E#|zQmnNnO*TD#Cpz%B;b!CUSe?3*@$9hv44CD`fa3HGee2Gle&kkzTIEM
zIcD7Z)%J;)xl^f>^<{AECm%%Ayh8FDzh-4;r-GEtOx5L;Rz@qOqT!K_T`8D#BdO3Z
zNbeCC_fCW#p+4)CDD>J_DN_UZW;+XfWfu1Vj1}{-e)4sokS@|&a54V89l={L&eF?y
zPn1c2OKVlcoC72U-3IFT$+a%uCu-cuHcbLFMOY&kr031rAwRcU&$U}izjXNL2snim
z{pY8dp1e8pzN_e&D>az;XwWKb?W*J*kn`%@YC5DoHvB8#qe`QsR1D`D4q?u;|C7Q4
zgC}GF-Pj_KH!qX>j#J$ewvuvmp<qNYSs_-7cANINGN;FfEwxFdj9^iOyC)L%F$QwS
z`QV8nxf55sz-o-^UmI%WxH*#6Nzwx3dy&7%W4Oe}tt)n=_U)~aD<L$i#gEY2JN7*M
za!$^9uYqXbmLhm>&R%E6q-y-dH7ubHUS6rHidUq`+ayL=il1+)yx^cw@mKFnT#E70
z(L1>G9M1Jo70F$}_^`!nd;&&>w&0{$^ak|9a;_mT*CQ*eo%VpV96|s&TF^_7i`bZa
zz_rX>(LU7~>PF^ht<Lc5096jLGtqMfp}&N0wt}zCbHLr;yOcHQPLRqQmU2bc_5+e!
zCy7hbQCnu0Ydb_ECeIM}B5BXHNQd!*tD8gc(f}ie^v4gk@Uh;Gy&}zWtYO&aUEZF$
zK>3f@+0eX+J9(d3WsI;-CtV|^`B<edm*4*F^S<rNZ*TvReErsR?uRtWx)4v(M*$<P
zt!Wh1In$6g63VJU^vXk)$^nlLy}A`oG5e9+eEZnPw|5raUjBN|?OnV66lHN%XP@0~
zyI|nS_hGM*tA~@1UGASObu`F}iV~Ow7QdjU7pADBG>V?{E)c0V;Q(!`X9q5vJNGm#
z?ND@dv?=0HRxslQ`l)-67bI<(uVZ-v;^N{cshn9Ss8Wpzq2x$eRfTSr1Let+C)(Lg
zNx)@M_<=#q?CsmP&!0aZ1p-B(K^%UH_D(Z&uHE+Yvn-=6@HNb6d>9gv0G4Y$J2kpz
z?_T;3OUuI{2Sr7fxv*9(5J}e2ZcwzWe@&qsy7t+|Z)?~KtEPpm?d>~5N@7sb+$}yH
zSy_3p2Z^mlW?b6D?&I~A*P45m|MuJ2qet__8MRg!32ABPIahBf50`NdeDcI}u^;XH
z*|U*yz$w9U)Ldw`@e=m!ov?+6v9ZQdvOYcPT3YciHBCORiKj0T68wr#>J+nPNXvkH
z42_P8z54f2;>83MPUBD>H-(DmHe4cpL0fwdr6JQkb}Mzpb`|%b)8kBPo|?M)lZq7_
zb)D$+s>XlcEUuP7V(kGAc_5$r$kAcros&pe*?+8CH;QK~K75TD0dYO&=9UGeXIV+O
zj81<)Z%b0g$2(l<m)|cHHptZlAvJd#SV9>Z9#+40Eox?VwymoxJ}XNyOv3)uk|j$@
z?%m?%=1#M0JPgvLoCnq`W>mbXBveAZu&|J~-MIU7uT%Ewt<rI*pqe5lYxGu*4h^Zn
z>^J=J$96tGzWP2Vz_E)vcI?1mSZo~Q+PJa1$(dP$W}<|zUo}e}ACO1k@xs}&<t1*~
z*G<0^6{Xv?DUzv8PEKmj^-iDuN7AwT8hs$Al39E2uEt*pHmq6yV7(jfckeo4^jG$H
z)A&CS)0%qS5^9w7vZu@^1}hX5xnqDJgU{`3rtRH$=ih(-UB7W-3@k;_EVM5qJUkiZ
z!DF3>nS&_5K+nLJvNBc7qGqJ4O6*cTM^(0x{}zd7s<M+Ol!EqII=Majc%=d~aHKUr
zRnlYp6<ihxD|)wjj0vKYTSu7l`E%#CQ5sT>{lh)Sv`(Hp*>z8<zZ|30pe`C9w3MUZ
zQA9+FIKxFg<gn$zeZzx;<!h)D@+d6VjnwV)$dUD&>O==l6{Madvd3m7Cgp&9BSi*t
zY6E$!+^ue%+mxRi{J29r${b%oQ%p0f*%KWb%Ok20LX6YQ>;;O3(H}mX`1`NF(x#{l
zu+ylh69~+;D4jlh@ZhqAk%oam3f#sNHNWoKq>!D;!BJ2pc5mC}&GdJ=sH>X@>|hFY
zLHIE1A$M6R&w(p{(E<X*96HaA_BT!6<{L72Yjdo=z8>8@AqNj0ynEG|g43!pPsxda
z0-*;;Q3;7h%&Av%a$+K_(SDKDo9=G2B!fJC<b-MnwK_7{+xG2K*}Z!=x2(t8KmPcm
zwWZ}bU>XujPu`ihuM%%tS2Bu3ts|&&1w$hiUK~Fg%vrI-D-Rr?*tdIR*E<>-Mo$$P
zFv`oHz8CYH+R`#qYc#l#E23O#pjiCWX3f2|Mj|huQTkuG3he_#p4mOpwR{#-oP?!=
z2M_NN6x0zDtGWG8=hMJIotoIPx;_pI4J)gAPQ71a{{6S*b))6&yLT(^Z{uIj$2=JK
zs<98%1wJJjRM;RL&kwv(Z?Hqrc(};5yD@_$CM%e&@7`A@>HbFTYJp01QijSzaD(lR
z;^vd?g(W35X#7d@n6!iazxg=&{rh)rIWIbV_+en6Lsga>Ht*r1M=^M4!vvKn;bcB3
zr$pkX*zN3x?ozo84cfw5uTP-U)F8w$jN-T8Bu@FKz40SP-hLmhw_eHned>%-P0>N3
zO?P<(RPi;DUr<(7A4!W=mAlX38XO3u-77jwDn#7Xp|9Zr07NC)_c!qIS)1*0eJ<)^
zZe})8qhVl*YDq(ysd_?4$JFGcDZ**6wCigcjh5V#7!D6$-{aMlzk$`5SIz`gf30_#
z+;kPnj6+t&g6yVbI$hfE+BFjdTuGO~OLrPIlP<2}lhp-ktU{SB#ohkhjT;M-^s+Ca
za-wnU*s(i3Idu9Vv+7+<ZaqngDk?FlTq^N&x-JR!d0sh?OOsJm-k8x<lMoc68j(y1
z*tz@Q>o#U<%z^=#@o&$e{>3!)IPKh5|Bi<*BC;f;4#utt(syuQp3T!!#$|`YkJ~oq
z>z6zk78Yxo5t3uQw0N_0w5uwkzD3V;sdZIRuTx0yskWq#?e+E3=?-0T!6gYYPJ0as
z?zNf!``2zoW!}agva43D(tc~>eUI0i@xrV1bojk3h3YOYQCnS_-G>GT-(Imydez9&
z(L3iJrZF4}jBPoyMd99=$%cVN&Fy#Y+^GUHPe6H35w+xMz#J6GF19G!+uFX`xKB6k
z?-g7x5OEaGoT-jH=2M+wAoua($K6McOx!G3e(2DlP86E3myz(A*u=!di>jHgw3C7l
z8s6^DV|igA9XE4h>*;j*NZ5df0z9M!^k5_-Z?@#?-K5-HeQ>!t=xc`Q+BJBTjO5kX
z>2VD_OtPMv`^zp)4&ED0RvMVe&xV)Ke>T3;>?jQzMTyE3%`u@j`lGH~xiSLcu!o;t
zo(dt<Z}rDi<1+4oKc7B*%8jsab2aaai3yicJQzFTI<0XfwVtG8Vg<&=<_P^LHbm^x
z%c}nI2G3leJ@hW{E&MHT<3LsNLp+;9LKPI{_?VbeWTr!HMGqf7G<3+p|GUrtSP5B2
zKtx1jhk$@ddE_zf{g<v^xq4Ni^zHHs7cNM8OuSspD;b5;HUaQTx;duS!WuxZGy~mQ
zzB>2kn%Za2o{bE?U(*A9xf0E+>2Sd$t<+DQ-;$G(YQV<QZCg&YOt0B0t+{O3GWc=I
zv**v3jebA7Pxoyk8d4A=6d*9w!3`ZDj$X7T%`M7dk{3jcOC7pP{^j*n+InELYh5ht
zuIKXuy3qn_TidI?@|P|-UpOCn{=x-3*f;0+rPq6r`wq(5wal3=OPGz<RF%d^tq%3P
z?ayVi<J7!M4&(ZTTm~L5F26rz+WU1GkU&ke(ke-tODRyD5=TaAxDQ(d(sOb`nv6f*
zE_3YF%$0QRe+I2a<MQRTuQj8Fh7JmAk9gm{quN=e%GH&e-JjDPQRXKb`{T!UirU$;
z;~PiyQ(nACP4c?IPwDTs+BbRQf`Ngx(=r=d0O#OCrZkcPhKB5tw9^$Kz**LMqhPh<
zYO+q%^xwX|Lc4aI_V)JfZm3bs&CMMe8++vI>-)1k;VGV+86C}X`T0Xm{6>P>$&-I0
zj-A)lz0zyf5Pbjs?<f*05MCFuWs}q&s;a64Du@~t6Sy=2xz7ZW)~N!&5~C`dOo-zF
zcQ>~^+qWOzwryKCL+13{+#CTSu*SB&K4E9)@rz}(6V4j$M_W&Gs;pWS6d1w$daNcs
z!2e5;j#Jnolcs1T#~1Mu{FI@Qkp~&$eqp>eIf7&9PI*&Pl3{WNN$`yL*RR#k)TTzE
z#NXqSW0-M16PTQuGQ(4fZV<6Ho3|_kLV(KfJTP#*p^;HH`gv=i>6+>qXluU!>o|@o
z?B7wJQuU6@%Wqh__Gx6~(Itx)A95GA5T41Y?5VGrF_5)y58SioivXpq%^353-rdbz
zQ(Ya08^fR_U55g{eED+z;zg&A<9pOvIBQYhstw$&eh9svPSdQOmX@m)Q9Ani$w&md
z@Qix>fs>Y&Nx4(qLQs}8J{jV$makZ$ZekL*a1qbAM`va99ZBb)DUGuK>ILXeR0)$T
zeE<IagqD60(_3hErt!y-7DY$h0p@O?S_p+Nva)^-s8y}Hd*_alwoFe_{NSLC^5gx-
z5TMm9EG)9OC~588xN6IhvnXR%ldx4>d?rahM;i~#S&cal54C@Z@S2K``oyAc+xBkx
zvSk-+Y+hzL_Uxpxl3}(F{Z$2BU0r$nMZ{Q}y{hu2>ts|FB*Q<{nfo(iv$C`=Xw-G0
z{o_Sq;-SMUMTCW`AQa?{&B@BjlF$HUAqKHPug*6UcJ(4YGs|yP*sa*upl?kn13t%s
zR=UW`zZal@ZPiZv4Dc(mx94FVEe&-InVYK1IDgjW%(7)Z0%E!E_U_qpK2}9`Xf$U^
zU1%TcTz0UW6}~!j^;6y{xwqNb=H-#wB<%!LG0?@!-w#|(^bxqA5m8qfrbUZV;G6X~
zy6kZJ^kc}gu_K_aDRy?R9j6@rUbt{!N>)FQt4dp&ALsG!-|L*XtLn5~{jhIqO_T0B
zN&E3*YWLpaQ5E^!!Z}09_mAqVSiQOy1=!f>{%lo5^e>-3pFMTzCN#;WD=NCcAb@VQ
z7A<eyl&2c2`fZibhJIcN7NnOO@Zrj<UAuQv@W&0+mb>aS$a+S_-8qTL$tnu|TP+=9
zcsSl3OsIKr%x9@7vURYe;~x8+diQJ0E+{JAR#Q`h6DH16tS*%5O8)(OYRYZH#*JQm
zrCs+vFqsCBNXyYb>3~YXXC{*s<Hc%jzSIOQISUF7>@()>-IefwJwieW^*b%qT5LZ-
zrb_vG%$Lv;ptZoz&_o3B5QY|BsSvMYopS-D*o5xx?ljM7Crbuh3x9+^9r<c;7y#!*
zRMbdJ4X18GXQv73GHL|$fwQj?|A_*q+4Pv7<@vCd&PMTbmoLjC_vG%|JP6{x`_Dg5
z$ErrCS@xZ5S(kv6wQKt`U$?Qlk3L?xpi+8_zd2`%_D%tTH#&L{<IP8hay?tpt(jI~
zvQ0Cqs9tm5skMjVwBEXF!49}34r=LS80``fQQ%kDnc-QH6qk_jP+)cb0>KSiw<b&V
zzm>;M<0-#v*|~|Um~dLp_!$+ITGS(BP~Ua_JekU1FrW>EE?v1P5}%N^>o29HhqL!`
z>~DfxaR_?&>YwlUxc!eo0E)!}J7#ECI5nP}dJDN9PYDV8af>OgUAxvO=sCh-6RIxU
z2e18uLyakuo$fVveqK;#@9bQM+D2e-@CzXKDus_Oh=yWf7l<1Yj1MAY`Ss<y>fr0<
z2OPW4D~BFAheB|6zgEqJ5q3-wmgAFVdo>H^zk!$cd2p~<P4#&pF~`Kc8xvre-Q`ag
zy>9FSe?9e>o2zbsXY1sY-tOwo>iB@DN3p#@55>F|jyv+a{P<y{qBd#sp5436p6|)4
zZ}aO(c5rJN{Rbd9X83-{@>Q!q&oT~A>)(5(F1XO{n{JS^i;AkM#;H@M?)c?5R3Ti(
zL+h`6%rEXbbj^=XF1O?p{mAR0mIG+J_|0)aGt&cA^eKpdB!Ci)DOl8ab<1Ym^C~L$
zWM@azL1d^cU4!@biHmEXANzLHr^$FUI@Um53A}&bMm<}a2OQMCLj_b#1q@0}ORHXa
znvvWKSmg$NYZQuZQ9*$Ml6Nh{1MH_;RY)k5f<SORpda$2L7k_f%%vDYacW*%_@pHT
z$XkW+K@6Rm{OjWbQf}Nh-tbbf09}t=Spz+<7g!H4UNyWupy%o;ar@rAbLY;r#LFHy
z;Fmr4sX%a-U`J$Wcv4^<RlJYMJVMP7E?7O2E#$l-&FgoY@5ftAt7K(!H9U%ozN3eA
zbc1zOZI-ijXppOGjIGXFHRn_zB~|7Xr)8ag+S%I!eyWiaY+M>@I(qnG!i5~SQ8f#T
zM4*`Un>WWdH#bvDoBLj%Rn_3<p<poct7Db%>VB!+$O?|VUk?CtC8UiQ#NN@%awy!i
z<|m@pDFXDNs7DVD4m!E^8@;Nx_)=V4&1B6v!@&<XZpe*~i`zBYQ>Th2>ALGvW0Bix
zK$yV}m0|U~SKerOdWOBR&=R6_H;5he;q*V8xjo;ty7pqvNY|&90|rd1p<#RallH?Q
zeG#c=RA0o!oi6a_)6mxs$4e3#dq<z*gOt3vEQZ&#rOaUdqAY|%zZTsiA|MOMEy9{F
zkAj+7jUGmqg@%U*-FD?4undGFq}H?^&UYBBw&J-T|IE}VPgHsN>F4c4EkgSuvS}35
zr3;#xaRjM}h@^b%`U#oh)TT}FYE+=i@WdsK%PJ_VZ(f%`N!o<y65oC+HcNj2)fY<9
z1qX+;?fm=*jxmvuf8w5>3kqW8w-~3z06QaRV?Ae0P2+FhxkEOXSJp!_GcyyA=b3F|
za4<KlRY*og4|zGO?UBwQ&nbr;f`S=3#sq|>8I^D+;pw(|k2WgTrcH4O3op<K(_Qi0
z_uu#LA&Fcm)@)Dw?b}L_3rY7elDVS+!4QAq04&gHan8yr84=j8vBBanppp`vU#&zD
zh8^?)a0vyNnVCuMnqIo(?_Jw2YwF1H#<s&)PoHiPTmaNs5k;vA3~@hATfe^ArI}~L
z24TvcT}NVIJ<zQ*b92M;4xh=X-^W+2zuHa5e)407cJ?vLH$ALYk-dpO4@CtC7G2QM
z5yHOqc<trrd;a2uV#39wN_1o-hlYl3tkc=Mk)ZOC&I)cxhd6{rn<ndL(lRm|elLG}
zS>eo?Gk1Dau_Tnq%u(lUcx(Io02iQ^{S@$cx=F=$<#1`OH8LY-6ULo)BHe3)RLVjI
zyJ)=0_vX!Rh-NjPAMlI03^uQpNM3lf;7Yo-U)?|0#KYOiFJGPo9;yBEP{?o9Q9@i%
z#G*ua4BbL6?#Ug3%+R&hOx&T()}Zmhy$?oR-v(zwysSc%Qp~RP^vP4FB+r+dywa?W
zR^oE(syqpw7EYFP?hhp_3sV*n6;;O(97q=0x33l@qqIx!7LE=u4G4S%ZLUdb^o#J*
zM54rF!Di@~)CjGIz-7Iq5nJ-PuFjUxDEY^lHESFP7<$C6^SnyRnk`LC&cZFw6%dc$
zWd^eTX{kFN(*Ly5luD)Y7yJAB=Y40r$0vXa<%9`6cu>)8^bsr>#JCon6WEPQ5Ot1J
z53H8l36-?PQFdSQWPOKIwjPA0h8*{dpI$G!eASSTDhSBN^IT>qH3c8IdK2<Z+jm>f
zkh_i!`*=u#ec?WSDNI@`8JVJwnWsS*vy;K-_y8G~SJL6Ax3?dUOM1slQ_tEhTlU1;
zZnVDgqxqt>{fJ4+f|HtG^Rk%1bAB217D--q8v0(di@q*ew5YA44z$b|ZWCW!{i<$O
zlcS*&9X&)cBq|=dbWC0~rb0j~e=09V@&m8b#RUr%h({;G_mm@K_0W2(LDU&h(e!JD
ze?Un1Y?V<2VFPhmJI5`Z0K|f=?#0{M>-Fvc(s&}5xZ!!VV8<v5Eg~W!MkVY8Jnu~*
z-XvnPLB?&Q94Yk>^i>4L9wMp%sS1eQ%OC}g6rz@*&cDE2?#P2LUlemtnHFyES<1=T
zL$qqr#}VSRvT`e4Ec2|j^&0%T=ZSH61(d@yuUR)jA}1n`(_W{hJ`4*>LRZ8Ef~i2&
zA0mV2=O-{XWkbcliqC3m2cvFN^JB#0&c|ol<Zqhb6&?gk+_<4@eYHCT{T<3sOVq*0
z$xQVitU>!gV5<T>y`a<hmQNW<7i94$R6^nrv<O{8kt*fu*RKh(wF%&{zz2a)kC5Fu
z;a|d5>;CxT84)D$DD!Q9|5za|m5%_F2M?+}c<>+?brg%kA$VIy%E*WVgqYe*n>GK3
z8rYDYfcwS*?qw7jS@7N}T1r=IX}<<;UIX>e;pdMfqXTl~Z%ePdinC_UNjmJ(nVa+D
z-1O>Ioh)_Su0TgmFUiBR(Mi+c%o#pPC0g#qOe@ni`WxuDpVP@W9Nx~PJKf20UQ<yR
zU?}POT6f#Fr^5Xi?&GCGUQBB*m5n7~?y&^T2ZrZr!z9)Gx5}hUxS&P35)@j@r2Kx0
z>l=<_1xLqpq70!CoG496rD=EW-mS({8)8>Kt^E71$5OFzUH9^Vxu?)kG=iQGc`gzC
zz$Pnk6>*ehrA9yV`|A$JON#5r%lqJUGEV_sm2h=Qa!*rsMlQMph=8!=F-N|5Co>NR
z(_>%`*f52P>DA)BEWA}Djmq*O6%NbvkU#Dj0zZ((;}E~UmILymap4(kD<oIzg8M(1
z7VItokNIjKHIW#H=hv+)E;W{yM<V+n1uN1ZXgqxUIPUG+BM3_`bTX{KV9$g1A@&1o
zMM6tK^zXSHmUZC30djoA%9}40@Ra;egd84)qcVp!@ghFH0y+;-I1CL9H+xUx^U<+J
z>z=+0=EX~wA}fyJxsast^0O_)-=JQ^p>%;v){sZdLqMvBVbVN2J#kn9#1Dao|H`u2
zjG%ez)-57TQRh0SX5ix5xgME7a(89IOT%Rnb|^g}syF6Kl&eFi4g&Yyk!WG&kLq6z
z`R7*Ga{vpo;r3Bl6kazWo-Hfub^Uef_-;}ARDs>nGYwLoj?2GyfSv|Y1Qx+7l2Ft4
zX$32`G%S190o#`V5YZWosptK<(_TwBeLr7$uxF2oX~nZO*HXpeoCl6PUorXmosTbS
z4Q*2ea<a9kC8Q+l&7$seL0>-xlKt7kIz{TT$!!uR6-aaBxi4QDUMyMBZeAU?-fz<Z
zZIlR#%p3!te>sl2wzahZ9BhCbd@O&3ZtTsQOA*TApFKN{FZS%(b=J!(mpBdTxEYen
znxZ15bvw_LqneJ|(eJ*!+^g2CsiU&SsXb_Ushu9kOCe<*JZp*YL6e;slo!^^I`1z#
z6^Aa<)xJ{71y&iPC^dxwZ1t7(??L;BQSOqEsNY8I=V<=CboG`vbT?@5c?^0pJcgBP
z>gt5#<P6aHmx5OkY&doJNE}{J)rqb|<<w+7Xe(@J#hROzDn6Y7KqoUd(K5vt&y7X)
zF-LyV#uKfnrNyu%EJttnD(B+GCImwqwhV^R6dbs@%)`S21bHPX!oYq=v?ZQ_<?>_z
zA(lf_1fU5vce6l#AasUg!lVpF!(}hya$CKvb$G;P|15ettI^WTW-E<vN?fQq8X6ES
z_hQ5+x=HpnUDwBlwlIq3WENXCIHIH#WFCH1h~l$!Rm~&e5L8tp(RQY?@y-PluOQu*
zIXW(8WoxBcX`KAHd&7neW^*&ljJLV&<G(k#b~x+k%6d$k_MB|niG0(YUN@Ig?+7^z
z?<e}$IhVue02;GEvP5<+gWmc`NLJs(<S~}uF09JaxgT6L5P^GU#)DVJ?L0Px06pk+
zKu56vm^=|iUh6=x!S;A!Z6%7ni6KRMqP2;Tw2`2^nUiRXe*ozUFLYw3Sx`bZbiRmL
zeLGUcEdJW(gu>!U3{>}2-ZxmU4$><*5^fo{v@XYm*G184QnhG}bR;?rI(_A4o)TFI
zp%ZHMDF{u^@aU(d(=C%9ofXgBkmDXkL|E%S3Ji=QZ63UDdK%2q%|74XMo6UB$aB8k
zLPAQ&H02<a5MLc?{MT*R5RIILmyM{vlRv#Z_E<zf-~?JmB1p?fTStdEun=^E0Hm*!
zx0(HU!%!RvzCb+>KPvrRrn{)&4)Dl<$`nXfM71VbDOgt+>Y6pvgVU8XB~JR*ty_r#
zgqo(?_&rNoGkJ)xZQs8?MP`PA)R~p5Fy8;pZQ{FN;DcAkfa9xBY`m7CG}gm>pST&l
z6*4n}>%J5g*52c7gnFavl&$>W!8t67)PoFPrAZsUZ)%c9v&|yIYZm2<oNVZ+(f9A4
z-LYd8Cd&^+tg*3KLhF!2=jj~+E7hFFyxLl~oWK0tR`2aue!0A4v>s$EUB&&Zvs0h;
zo7DB9(g?Yfd0|bN%{ZzP6E<wwl86ozp&dI;f<ssZC`%Xrv3`9N+7pgtG#!?wPQ2$q
z%-9icyDl8Uhzf!sd0+%CUn0ko4_tz6)W%E9?pzYzx$_iW4EGm0dX!$S;INFNvgS}R
zcp6TfbM5<<seU*_nUsa`yz*{R>zyfa<iP7cNQ>FsHP7+Pz$nUz-RP+Ko;E~9MPa|u
zn4%hJhn<~WU{Fvj8ZElPct|XP&|jTtZ;VW_XU`r*OsMWq3RgS+K6mEKT~Zr_+-QQ1
z0F*K7@Nlz3d+}y`LSPhRA7HKSs#yOYKYpy|;_9GXRl3geePAF3MQRfiEnb5fqKJn#
z8w8?42`JVBsg9HfNz6)Bazw9Xo~jwAjrt7WQypF}_b0^-hu{bi>9OwG1|3(|OuUnh
z7fHGJYvlJHUD@)1Z<(o}DC@y~^7D_5c2>v(XLr}7L@1T&B_zv&QrPt|atsD?vrkV>
zmUNUMC?EHi9{OkLnyugZt^3PxV5F8@@$96e?T-_wJ=?c$A3;VZ#EyO|XTj*GW5a91
zB!PJII*wIar5jprb99{3(MhU`<zXG3oN@|?i0H}LdD_o7n~t{vl&HAo4&+XcpxJ8O
zQ1Az%RJ?;=DuCfWAk*^t<;(qO#mlR_zF<+YP%rluIWJ8G|Mt_gUz{^QXdy@Q)KA;B
z9BHZ0P`eTLNMWM-nN9{|<zMtC67m)~mOuZ@Kk`13UrAB1{KgV%v!<+-B@uGWuZE70
z8;Qs?g;r}w_eyxPllWm@ixuEL@cv4W89V~P_`%!@4IDk5*Ac2t)-B=qwixV^zzf34
zXYqEx0D&CDd6Mp-*Ev_oUQ4aFkaXz0znV`r5k~w1l@vH8(awQfbjN(^sOS=un9W0D
z;Rqgb;tnEE7XN4u#5--y{41r-VWNOWz;e=2XEa><t*J@OYsQ6W_pt7(;SF^!QEo!n
zGUm*aXK=>z*u&&kqRviEZ_tGj(aFNI6Nm7FCN#uyOV~wucqChfXCX2IDU*UC2~GxH
zlZFYY;ZtLM31}8nfH}7_S8`Yp1(QTPs1+43wH<5<XJ&3bN;+Be<QYGHToKl%Yocd(
z@)Iu|+n0d%DVRW@N7OOR2rjk*BbUKDh>)K5?{n{b>N_&g;%|Q#&x!5a-n`_x)1}mU
zQo8C6b7maamx@(@2W|BnIfs%)h-TRE&!M48!VjGLbqSFmy!3*u?q14^<m6gXq=Og(
zEgXnOiQw}SC;mEq{5bJFS9f=C5<C#1L})u4=Q%o6;=oCPr|?^=(P_{tGeEK_JWLRc
zUKo`nOF7nID4C(p{6jx~R)ECsq2P^`@ubQ`I@<xDEUE@H)YP7#WkUy@Z0ZoYFEk0F
z`XnJCL3&A0Dt1I~ff*iZ+4pnmm1)(^9CQW22p$MvIz4D<f3ow;<4bmicyjGJ7e|98
zQW+rIoC2CUMEBjfb*lo#t{(eDsUwbHMq+<>oXHpONPGqmw~L?u{nw<gRvEw;dG#uc
zx-U&ldUyt_8mbs-6N|x+qm#aHa-?~qL|E9^s6&)Fv`hdAl9-HVBr@teHY%7C0T`NB
z4luT`QTrNH5zq;shQNeg5)_<}OAhi%yPPF+D*Xy2iliXY&hz;3CQme`k-mzCsN9ee
z>@IOJ=*ot=8zJT}XZxL@)rh@12wt~&fe1(1QyT3Lnc0ylNvGb&w)F&J3x>NqE0MY*
zc@X%7*dh^R$m9@?usnHKkEo=3$|EE-Vut^#z3&W)GV9iDOpF1YZyc2%VrWHWM3kHz
zF*O1rA|O#gK{65~G--6i1gMROWN?ryh=>G1P!uF92#BOsq99pvy3cOJ^PO|5Zq@yB
zs&3Vx%0G7Z8}?p%g=ekxtiS(WMAFi*jw<6z)>EVyD7Vl9-bFMr4itg0Wk)_5O^~{A
z1v?v@pg2IAnRUM`Kva)f4NdY$fH2>8O`}Ln09aN?$a>sL37aQeFx}eG(T$4i_XK|+
z=|a#@yax~;Y8dqr1Zu;GAb@-K6r6*5RV$F3y&2BMY5{bODgUc~!9H-xv|d9=_>NE+
zXEik=n=$mhm_nSkX|{iYHWjh_Bm#Z|Wwn;z_sFHgFZv2Zg6IPzl1cdz7Qg`+ek9Nc
zT#HH`KzgBHl*K3!nYeig>RV9Mo{e+n(9isnfc_K1$sUiQqN2=Q6651{=Af2r0tMkg
zdUCiES!B&M+yHxwQb~uZBFc*2JveyNJ5|vPzzf=zz9LI>V&6Vb)Ceb`^f=BS*S!m2
zRKDA<Q_wP9L^NTlyh})2)iv2XE0@~IMzIlEludcJ1X_wzhmph_P^#afNb$&p3m3L&
z)(k9Kv}pUrjm)6thJ}=nV<>l+L9<FX?spR5uu8bxBLD$Pq!J1(N@tLBa_gBU$goA-
z-f3_d3F+r=-618Vjh3rRUMsk4^&H4CK{!;42pRt3LsHNKpGoM@vQRlUZB!K3z+KoX
zTK?5~;-wK1Y@|>KION;6M}Xl>N0$vef>uD6LnM3{o&cl`yMQ_E-~+h;GvXH~Hs%r(
z*($;3&lo5Wf?-A*0v|9%V{#xfUI5W|LxynzjW-dH6HyV7^Di$g)=o9{A^sTI%up~j
zT?|K!3=qX`#4#t&oQb@UiTwBm*Zx0=$6#4`+&o1~#PIbm#7JV(6^|XOMN2@MNyUaO
z-~72d8rH5`XK<lZ6|wT~JUlD`$N5vZx)RP~Ge>Rhhc@Q|lyvvm!v$#>g#1sV*Hsnq
z^{4NXY!uID3TU|&XF`%(!0_ZS@!gc>^SRexW^f(i)<9iwvhg0)(z?I)53LB$%_<YH
z4GI86UXl1tfD7ugf9K;%2IvG0sDjcy0!ER%|F=D04U8nt6Y%98H6ui#D2-Nu10PEC
zEZl9RZRFs8e+AmNbJs4*wG(0vuvf%s0_EM^51i`@3kwA#B~O6?rmH!3*5MWp5*Y-;
zhRFc-VySDDrcHvUPDEbDR2-s}zR-Y4CzO?aP)js(6xU`%VIN_I!)?$NzS_yz`70{7
zYe^Lssj9lVitpZsF_&o=tFKVaX2b3fK^O1~p?>LOu-KB9wY8c^GQCHIws#5MDGF)~
z)jCLP9RO$m*{&px*7wfiaI`qn)9`TR<r*-L1o0*?I>~pz??+q|otKwqofw&zs02i`
z3cS0YEbO%6QUoThtkYKwmvj5{O>!Z{bA)U~?Mo;@bf%=Bz=28|938EN@Lf=lcElHW
z0st4RlAlsVmW|Z=04@VkTmMZAhV(cRK--RmC4%g=;pyqoZlpT};7}m>v?~dY0!$@5
zX*x<&1yRITHs*PXK@1r2vQSOSh}t0Q0$k?>&O!jm>KYgrRM*tR1NueUu^OJTPg!%&
zGK9KlUXcY*5Xx}`XL&QuvJEdOWkK+u$N&aB4CHa=-o2Hr@vui9B=k?iZV~4I$P*En
z2M;#m1d&QRAX2uDj$!Hh0X0BC0N2+5#z6${fb|6aM^ra)*2S74A3uIvXYtt@1&#j@
z^M$+y<LcGJbeY8<i1H2Th+RTi{~E70j757tHi7yAbwk4k*VjO{1yC@943L_-`rW#R
zPq_fY!Qi}k{krqTWbX+WN&+PzNbUKUqJSu6wxIADzD(9p={{hMC6Wi$3F8#Yk+<X^
zK?Dg80NOL6(AT64KUq9;C!XOBYOG9w?FC6(ctAzjq&D0E<k5~pT|rZ}qKE+qF<{y_
z&18K|(HKag1p5O0JgUCFe!+xU)*npHnXy9oc;GUO52#b0X0po*Wujr#vz-7lKz`xk
ze*A{f0eQ-Kt*&`gh*g0L3^Pu{HjT>(DUN;mZYf;eAS0Ta5rAXNAf>d^eX}#kC>~Kx
zBI5@r6#KJ_vOQ@Cu|6bd?xv{$*FeqYdE0|yQt(8Hqv^A821ztc$fMHyPwdH0P#giD
z^qlK74o|Qa$BUvbZ*D@l?52^F2GPRAO(uUEnTwSO5_js6Pp%hKs|LWVqBW7AyAegS
z^w+@fEkv!x!gmQ50)<k#KX3W{?LtDTIN%IY*-L`c_O*X&fLvM?s+tjnh=e#k+cXK$
z5?1<}=Q^P}#6vy~C}+LyA0diwkzi|Jm@whp*CNy+V2QHBLl+?#PL`gAK;b6bt<$y#
z&zpg4faN(MJ*kh(vN%~~{CRq8!h^TBwLL`r2TctRjf16}^h<zVw$t%&w~@esVf53S
zF(gEL_}c|rz4A2da)ArVD7R@`iY9;w3@FVfv<8X0z4MoChkanb9Tc<^t^=(S8RC<e
zUSNChaea^W)AM%*hN12)=P&wb3v=;$&*bD}6c6tpo~VX~#=W97XW%X1xCBt)jh2j!
z&Mlk`6)$-Cyi_!-E=T*C1+>a$Wc7?ta^P5EnDTQOy2<VD0On6jw`$z^ywS=~lvf8r
ziMmZH4V4!J11iEXA%9PyI&FR>VIoSQsBzG6SFT*yrupFlG<qD$eHbE$!w|$4A=G01
zX&=s8E^!-87Cha^X#{D}3ZEfUZKOQPA_Jz_5LsKuC2DpzWgAoU@bTk!&`+_2MqpB;
z0I;^Ki!h7HBQoya-)?PfO{)Z)t&Zv>+k?Cn=*jXPMkVRFv_TWX%78N?$XK5D7n0-L
zy#GL+NO@V&8V4EWh57|Jq~z}~c_hvPk^3=l$)UFV$)X)iNlBTPpZ;NTyDAZ8BdS-g
z``gQKad8#6AWyPQV@`f;9LN_|11?3>N~;8dV~=VmKY=a@5Z_uPqI-VrYJs^0%a9^_
zkfQa6qM{;$3d7zBC=Uy^M_p618^(k|C0YTzLO*X_My_oy$@Gr`wOJq*>`q`{9MDD@
z(Krna4NCKi5=`VP)->-qebj2Tz*mVNUz5`_KbCbZSytpE{ou}0O@U1xGTHIzDuS+i
zyL6~D=u!lN$kVdc!!rT)L}DW}fzc8R7{$TCL8YX(X)ue#?NT!uwTH{_^R8XL;;uI&
zv&Adg%BrcRU>p9qDwgEQi6Sy8Ti-O2W>mP+NSVdcLZO5#W!ue+Kn{@H2IZ;Az}ho6
z*Ah<}pb?D`6B7f4^B*@MuSRMf$tzG#s;hUx@7Si<8iX9U!_1U3O%19WX5^=Ubdet<
z0#hlNAnU8~CF_1Jp##C45e_i(aOcyBAWKkAeC;3_nk$V$<)1rrk*pQKEb*E}*_O#)
zr}U7~H2zC!Zf%62tJZ5FLIJKqrJ*b1>_2H@_s}%}SBzE(?f>&WE6IJZKOewTM^l5<
zfJGteh-TFaE&i=r1GSd2E}1tBEwX;Som2g>Do=3~NxfEN{XZpX6|TTU^lzQG)TzD(
zHE%wm2xZ8h%^@O1TbP!7z?@;CZ&P6>iYh89=8sttFZnHc95%rs(|mXy)me!73D|Pj
z_svPH=00q(1j1>QV$v!R2Y4r)Xt<9OS*(_F;+)!KPGAvLHVO)sYjw1>1!`UIj$qOM
zz_cxZ9a{Rjpn%(aK@*QjYMr-dZB<ojT?OILfxQWYOsGVdQk=*7&#V@>zyX~Amw-Wi
zQ&?C6?CR%3k{4WrdeO6KVl!Fj{|(ah8Ind%M~blwVZ^|D8sj}z_anW5tUEOhrVc3u
z;q;CC7_IWBku1vXf9~2IeABNJU{#8-QofHQ@QsvS07uvTSr?)-aN?0^kKb$8!zloU
zc!!EYbELQ`e?AN!9)|jhQ+)w&2=x;uw#22-M+Ys8=6-H+;0c?=b!1#1!SrcP)X$MW
zT_VNQEXO`Ud3kx7kFXQEMT*u)@-HBqP9Vrt!~x(xlcMGAQzNLK9Sbi8d!Pchfc)k8
zlU0iPy1ofPU9i@5JJ{tzdEKxkaEgO;33n^L<OAUdIRbf0k#qSTd);3m27xR#on{td
z`mMd)*!CdndlYdmDB{iVhzcf9HXU94IT6SI8iY7yP7QiKLzY6~4zem-=2A|{yhDyG
zedL+Ix;_*=B2fOg6jBu)sC&n*sH8-=h6GkWG~7Ee@+yXSl|3Bm1a(yb-DNJ83~Xr7
z1#zyKrC0^$+~b}iBIC7(qc#REAV7Quw+R=@H2>puV6H+|u=Nd_N`-9_6RVk*^-tuq
zoDEu%jU=~l@~`}M&N7!@PwBBP%9JAaR$%WNj+~dj&DhMMzrDNwKOX&!tuO67cYh^|
zehhoEm|O`}XD%b(U=R}%)z6)Kh)sC`mB}k<dw|gj0269@-uF1qLqyC_DL+lpeJTTh
z1Xc86pY|VX|MTaYnK@#m&F~!YmvjS5y+7cv$PqOJUDy)#_W4tF5NjM6&1U`QpMT;1
zuVQ(Ol-YkA*{$S0H@v+?YV7;nQ?chq8PE>D^YU6+?O|P)VhKORW<M<KvSrH{RM5Te
z7Yi<Hy|G>s$^&n@RrI3VzpdrIda`xE@9$FM5JF+Gw`sOy2s-vDdYjB$FX<|-r-rPy
zP>r`~#uGdjanXL9?uZ_;Rb961$*&KY8(dxaLu83%<*$BvIlyzY7>eKfr^XQ_=gaKV
z%aUteHBct3r6l0iVs`f6AHuq?9w=IfIsk-}=*#I2XxkxP(}#~a2t?;qd4wgoZE<~l
zedm!~J9nxfp^DAfAf`2U;FE|yVHi*-Li)GhL*SL1<~6D=(P^)P%0Hh|QaB3xEyWsn
z0dJk$RSSWsw<yn}v=8<N6aZ(>E3pVSa>~3{HVHgO4bHz@P>5VoTU$GS>OXB@UGrl9
z+BIvS;aZ?W1Imp3eu`ceSq>mdR#3sg=2BsQAyOoN89O0?*8vd-zz=2Ee|fzWy-GOT
z>y^c|AmXg=u;j!RffOSOl~*uY_jP?*U_MwM^0MK^t%B^?zZr20rp}*yM4dn7j*l!j
zRQy-CGXyuJ#-SMKiT3yj!od$YUmeHJY{iM;R|yeH`K|D!6JfQ2vwNJpu<y#Id%$-s
z@Yr+bT!0)r-Ky~?i?V$>=bwQbmDWp_F13FCyy?2<=zo<EH?mdbW@$TRWONz!gRL}#
zc0sy#GP}CFNwh^%Bc;3-d(7r_P!EweHsua<)vSX;8wXfZ{qEhy>jNuTPnI`~>j`>F
zX3wIw#3&ISs=47-)+dynTjZDING;0#p&lc)K)iMLRsD<OKmLoXo-T#WdJ9UES*qvO
ztz)U>#&g%uJdn3MgR-FM5KQ5`+@%fjuVj6iAy&C6Z}He%49QSeQ!}&`ZO?4T5up8>
zlC!bH(g|)B5m9H5+!~PjKwTfndGV9t#+Dxx2^FtD{@m%uvhF(!S1IEfvH?bo?-%ML
zMTIEwzV45BViIevd*JXYr(-~$;JnZnuyt7LGd~xMy!OA87ZTulwZNW25-*Fku;}b5
z`wl~4XI#UfMOx>&|NRzN=7KDEH1jIiW{5oPgRk4nc~dTnOH0>lt>P5Qn^&k-8_3hz
zqf!a_gdx?mLLZL2eZPy7Q!Jil-npE6<5z|$oJQOvKaUDEKY=V~6q7h04@4uSg#YeU
zwUhNfayT@<<fpKaEsZkGtG`nqBGP!;=g0A|Cy~Z-Q||tYQ9hWtlTCuAc|#O@6ED39
z&kR#Tij5gm66PL09IW%}oR}z+jl7LAbEC%N3z|^N$;DGlJO{J33PLwap>#?d*-u$l
z`7ddw(Y|Ku#K0e(9czQG*aJX80)!z|zM=`aY@TjlX+NEWaUAD$uwrOs*-4Xfn|KO+
zuap!!Ez04<&cS_lUL3`VFJOqn$xzqVw=&$?lZ>=JU@b-zQZ_(npp;n;La`FWD!&oo
za4IP-R!WST7VT*QUj2C9>9~Ixuet~OTUn{LTENr;1$lenRI+{>u?1>NR|zeODZNM7
ztku*Q|M^$E=ME^KY6Uu&N_+Y8B?>l$=l%0wA~t*Qg?zvl=Thtbi^a-dy=B$$Ln0t$
zs?rknm{VEy>e!Q;10y3E$hvM7U6OtAIXtmXhmDP+Q#OcM-2Zuda*F>9cupF;?BQGc
zC79AtSY3mt2BlC@*L;1&(l^e%z8{gmx7Fg>m~fMjqN1W={3G7zd`uR)wTtJGC~t+3
z%lGs1J0P?UD|<;dlWaa0TXMA-$s*5t?RhdpYHZt9HUaZIsiC3BKsGiCh~K<{@pLAa
zwRiRtEkd^#*AFkOfA{>)fF^Nwqak44wf1DK1aM+%LfSxWz5fDo&_JuZPtE8WFRJh6
z&tW^+ireVSB(59{m7%Z!!kG`AVNirsBJ}I_n{N!~1MYTms6Vvr;Xn{XI#`OTpZWkD
z<L-vr81%!X_cS60MZL<W24-71G0GCzQvqpd9W^5-;<L`Hcp)NY4{$t)?bfIiBXu&v
zQ8d4K_i`hI>m2MWkGW<iQx63eKmFo;Tgd6oRAcYMuH5zWLxj9)l_oj?e=g-7vNgqc
zV`OVMyHQ>w95xq#GQ-Ilw(Z=)O_E8dMpPtiuIRHjonH-KBDKfSzk}e)ueE_U8<PhB
zEZn1_o{@Cm%_|S41i2FF6GNhr=EIxK@>Y^6z>c%ZNY=LX*ZEDiVeRX4nP{|t{fScL
zJ()q~0SK8ulc31Rx=2dde#Y8NEppNJZ#<9O0H_uqf;j`5=Qn3G-~Dd7l=&(|k|~R$
ztMN#CD{ZjfrZ)C?{=9dQH$Yzd8xq2C7gi<<k0UDqWMDT9jc2h^jw7N!P7tM#sBIyW
z9_FLu<XWIoKFRz_oaXvJ{IHeYSb#I--2Kr2F>vx3s2ET^c*yIg7#b0Ixu5ELUJH0k
z3P8c<^X^7o3Q>rEp#Mq3yd^m^)xN*CIoAL!-?ynqB6i=?vD)+V8Q?5t2q6SRuh>zu
zZ_Kp184vpC2EC3`&DUq1Z>@iFPUl6TZ_k)}5$_Sh{re3WBKv~wm78T9VUK=#<*vfl
zKhFPk(LpQKMN*d6>zVKs<7K;c+>U;?-&kW+H|5H2pKeDjy&n84|3yJxBk5;GbvJOK
zJ7<qhPM!H9>U_x^3{c5w3a`*U)&2f?^K`a+k7NCfT$ho)uSpjUCMS<}H6p#0!yiFh
ziU|3lr6DaUBcElFv_-Y6W{a7MCE)`hqz_bZd0DvkcDN@Mun-|{iv{`;g~Pr1gk(2g
zQNNjtJON+BR+?4La1H3{KV%S)!SfD)fQ}450Vlt+&_<$(tsjcpZ1Gj=Vb|&9x&6yU
z6Ir+BT$F9&+nsaCk!t<)8L&jcq(^%K&WAq?US$8cqbuu<E1i&XJC(1f(t{dy<BBmd
zk%NPUxok&h5z^w4ex5Uqtq;HxT};#5!lHCndV^XEOO4-vq3y`n{Pa8+XWVZiiwQs^
zNjDG$kv|^R@YnuVl2)i8*x|28JJh?Z3FTV3CR4)lL&BmgZ#qJYJSUH+nK}wi@3kJt
z(-90X*FA24$pFrfVm~Tj6cw%2J@Mm)kUj_I*<deUUtiF(JuayJo@Q1Z4`K)^jSymD
zs$ccoGJW|KB<7q&Zccno%)i102t`mT3y8RJgwT>Z=1DUq`Y(bG1MP~64D^1sf*LI0
z!a<^*-)sgsxe$0(!Um3LznnxBg9#J_jX{XVKoKX(f{n<NtFj;JTGrIm!$n32R3{ND
z?0M-o+_;l0{NijBad&YUcuWufc-i~pC+Nz2FT7y~TOWeeC*U5S79me2j5=a&7U-%f
zvp_#~;g*zx`2@NR8vrw|8B00tyRa0b?Rdj;VCtbAD&J;~9GFvu77G_5;l3(A5jEE_
z8@R8!EoKaorZsWJRE(O(B$D;~f+q{*q*-zkvNS-5+h{0ty^Ch_>ZcpX=MTu|oxJ_M
zC`|S8|E`xquRf_+J@;F~)pMUcOH0H5OtV5@FGO=)ugMCbqGlTnOja|`aUF^Yq{6`v
zQ^#amFp=@0vBS5<=i(rAq7fbx41RoQoJPL(dnkT>pl?lf+6pH?zI9;(U|G?a+obZv
zScKI%3!6fv9cvPON$0VNEAquPiNq=Ov_mk_b%5~RE8V{x^-RDCC!t<=EsH%i1hLMC
zJAMgH$cf(`85uc~>kjAg*iA2Fj*L+{mak9ZrMyhIN)BhiyeHeD{nYd$=FB1cbC;Ug
z`QT>d$Bz=<Sjcn8QLO4=v&O5Mvm~JQjNA&rsH88p-Zjy`j1?Yu-NTb0_{68o1|l|U
zC2KiCLqmTs9>*N(JHZ0aOb+38QTKuqQFXTi{BkrEi^rf}_ZptSFDV_H$ul=47+-yU
zFds#N^aA9zd1}nblZGUaXCPELfO)9-xtB(f!RqANZScZXE19^<Hc3R{aKK~}oS<BS
z!FhA$`Li`FUrFH@ZX3PU%?asEx<<d4xVQ;W;e1=x+t`r}P};Si2}SrFND0zZ5}Pc~
z_iL`gMe|f|KiL-=_vVoKCl=PL4%enjsn63fL!<xGX>!{Lz_F?(P!6$gmg%#=wY+!#
zf{=XW?T3(Xl!zWnVp>95Ac5`G7K9&w4jK<*%#w@GS}&i&ArI5|6{VGcgN@?dmGKQv
zmrYngVYFz9h>P=cg{{d(-WczS=`3BfWHCT<g&89Jngg&6^3V|{b81j6M=wByU|bhV
zs=DAvcfGTDW=f7Aw)<qxTKsNb4jzrSi??6nxv~O8ZZ2r%)FxL$h3+>lO}oX`nYx6!
zl0ik8xZ9<3_!Xr*jCHX`%trJgBXx7X^>tUH*AkC5^5KfDrpqx5MZhV6tAE6*B4U;R
z#(D2>P<XkM)?~AgEYB{oH&T7sLoNl?-(J>BGxeUFrBx@9cTr8o0$xRljUO6lwn?U3
zO4k+CY?Fe}qx$@08_QuIR>G2X{m^$3HmBcq{2<sQbCq`mqM+lp0l5|iBw(_|$>*#3
zfarsPI;`~o-AgS_q_q|w64reTm%)Y2zJIP}IWrLxxOWU(mpb93f@Mv9J(%u>#Zr)y
zC*Q}nb*io2zmsOBT_Tdr+UgiDyTyI4t}MUm>Q26c->1iCUP=eqCZX#g6QXG#1uec^
z&Jv%c<c<j3P~&REpn{~0uQC&(NGdTxto6WbBOWbkh&;H>C9JTzf`URh$p5HCFzECJ
zb&(OW<g>hfqQis1eqBKZO)7{bbFgL+dJ$RtShOa)=-o5Z$E+tj?eq)Alnm>?)OqAt
zyNm!bp1`s9^CsZsY%@d1NIDvPL<W={M-l5}eCLOxY51(khCP>ro6T6=!=?DJ@^&dX
zuKPUj_(o{H@8w!`GLLRMZ1sTkMElaP%)XejwgYDAC<7e$D~m!<viXgxe(p?UVOa`6
zz+ADj34t3qR3@?}eH=9)U~;V)B5tIR5RQ6gB+NX0aq?PLxLZxMiX+ORM!hS<2-jZk
zsh5Or5EUfh)vovrB_s%M+@OjBWo~|yPNy?gCpaE`Sb3_902+C5c&KuYHkKj!&O+K^
z<;YDSnqV@mFcX|Ro&7|;6ORCEX(C!%H1qsAOK+WqGa6ZrF2_9(DH;R0>W8|5;RCvJ
zM^F1{XCg}vF+VKI-iM4@02MVeM^HuXcF2y9dw~**#(>2?3%Wpslf;lv7Q&U3WwU0$
z=NR^G|A?_fWJtw%R?s1&i_4(2;hk_3csjVs-qSCO_ORl1vaN4XXh1LEum2+?2~$JB
z!_jJSl)&z%f5orDJx7&vM6_6FrO*`qLokDR>8|2Ks!7v>Gsp|d8VStlWL)S7#sI+U
z@yD2Q%Q9$wmF{u$7!{FOZB(?asK9UtwThBUzYE8=2mQo20I%cAvX45fz{lTaD@l8T
zDL0!~D%P_YcO$dKon>39EUO;7PfvuU%q`y7FHV0^i_o^@Qed*WUl8oUv5MrfOeD7c
zqRN}Tza8S9H~DJ$?(g1@n2YfNr`!VljYWa!=R3Q)iZPGjJ~a@d%HS02X-y*nq8f<j
zVU&hQ3WG$Lrm~ya_~hJCvVI|2IXXN2<JI`^CCc<tT%Cq<14$#CY4A@jyM)c14>M%t
zzH4RKn^t5v8Z@3rc}Y~sz8*Z!20{i_F@~fv-v>MNnA8eC`*D15xeYNq{TCe_JJ5PV
zV|epQiy=RptZhl|hQqShMHh39WP_tPwlNzq{?&KAB!fz*bE`lnRL!ztokwxs^Hx^9
zb>|<YrHR6ImuMS5COfQNF$R8NqmwK#ijZg9{jTufI=v6V?m0d`Li)6=%*7M%Ji_9Y
z6`+Mt65&Th)}IWRO3-JGDt5yI%DKRmB?B)j>QjskqtVBodg_I%GiY}I{Yz;2XcCsI
zUZy`MRc5e&-}k_xg!$CyyN|c6M+4m#=Q-v4X?6dC>F-um{qXQKWnvujCcVV7BfCV&
zHW0Ds>_BZrFPA)8DW^9y!$p9;y-N)gw?@Z9xSZ}o*7tzPbpFBBEQYF`Q?vtrs+)kz
z$Mn3!KyCizRdV-nnB{JHU}$v@S2GCH^0irSBM*RXf>0<QGiXHX%~^Cx%HHxXX_I5?
zgLA$2+c3E@QH`;kAU5vg8pSgrATvfaAB}+ml5;52%_oII6be5U=7tM<HX(gf6AbX9
zlJR6NzRjR`$=4&MkCKCpzd;5Z;fGtv05QUChsqCrB(Dt^QR+L}2xa@Oqa<||OaI4A
zk0NB-XkVp#D%ahO(VFhwola)|5TaNbp><}BZuo;J`3UH#<tsW-=~V&wXHc=>Z&?Nu
zdSv`HzufWsVsd`~lX*MHC~<IQ6eD7!`+@{v6X8OgSkJOuV&n@b%Mv96!EM39MQHRY
zI#9$-hzrvMiAMJ9#I0T-bvcAm;;b5$b?)cZV0nC$-|x<WXAu!U5TeD9CvxlU+vl==
z`~LiaIxbV3%c82P&Lx_Nw@yy$|6T%1{J(2K`TtHXVvnZ@QkcVcOK55n0~xwi+OaRx
z;_iBR75EucKV2K&_d49v<HhIFsn)&{wW2Hz3Enna$EhhcrY<zlq3r5}a250J5*h>C
zmOF|%Kdn7*QdInyirrAFcEC#^Wc8d-6WPq54z=%kyGB!2r{zY31m+=N{trVtfbGin
z!IDR^@*b#G_aCarjR2~U7dVu_au5GsgvlQOMrO@HN=R{es3nNdl^7U`AkM+zF21<q
zMsIJDMRQXSl}k8!2l(AnaTC>5E7yMv2bU*54gGod+S><6L|z^;uLS<MOw`K%?eG&N
zx6fKnuW%^t*!1U!a&RMPwnqmy9dEqk8NWTC1ksy=Gm5$@KTX(vs=sAC*t+A?le5km
ziB3mxM~&{^*tUn*l6bME#`09|%s4nkQ@?h0zWws%=up$>OmT#<gzK0pMgl}8>Wo_8
zi}!~r!u<r2wC<D$(v7&e6K^dzeHNN$iwdYHBFf?7TKNmG#YZM|JH41{+YORSOE+Jn
z;>a|gVlcdi=vYs$0{_V_(fD6IH*4q4TygC&AN?-f)YvE++hJY*Oq5iQstL}vPCe2Q
zHv!CkDy{1(=%MBB0=8LoDu*Hj2s`gu<wuE~9{udIZh3{k{{ujneoSqr{}ZQyCub5@
zaPCAaL~`8=Hyv#8;LY5z7oeGy_wn%|{+`C8i8E+<nE~JWtO1yk?!n`$$<n1PdytZ{
z34BVY?(NRq4ly~RQ$Nx`@6%)-Za`g3D{5{PGaEuNGGr6#jZxXj6$)>qy@{U7;FZ?K
z*nx)Ndh6MTI{~!YOt=;-+I|-^^7v^<*@I<VqQ!l&5!XP(_?}t>ruyB+pI^N(wOH>T
zbr(Ah7S42&nW-x@2HGC0A1hLBAhti}3_}}SvfJYX$}$;yjDnU-H=uNxiVAzU65_2f
z?p>i=7@NvagUAMf!(GAj{P|LlP6oqSKiS7a@f4;q7ccuqSFq#G;SEX*T?hYLRtkIk
zL0No<8YoLwc_U0-v*>!YIz`p3O=hAqcPlL^cc#GomYAqd*-1^kP%22)W3Hy{rEsob
z+pWq8gRP92HQQ7yicmynuD8<Wkm9MV@2@ausNVO63Ft<bvrrKZos*Og$*3T1LkR*6
zp1+B@Ob47s6$UtVhw?SnQC0wG*zS$Dk)VP0R-C#6$D3K2GZdO04GYfGJl0$XPRWXb
z8@#+d7)8N8RC$yKpi)C=cy7eAD+^TDA$+|}H8l-{awwiXAXm5Az0kMjOu}-?lA{Zc
zM*K?1bHUV4r3vPI=MYU=A_4Z{vbQMkECThj<KV;H(<j3V35l|fm9wTsVusX1Mzg{~
zRX0@hsxXv7r7ZL8f`U}O4-u#{>-lr`6_?l|)=sG}SO(AB#E*(rp0bP!W#qy-5Nbt;
ziF+YZ*{N=m&%onc8?lqpbzHM{fLeq;Ja(Cw$jB5ET|r09oNW{dUYPkS{qGb;`Dsn)
z^Qg}FjcEaL!Pzj7H1oiDwnfuazU1swovDR|5}ke+2tz1l9t2~Eh{;fJ?zl9Uy!-w~
zC5xHsYd<<|&X%pan$Ga+`I|BJ_fpCh-*Bb8YolEW&k^k<pn6Wdy~04N+txGH=9`wu
z3p4>yY2@;O(vJGvJcqaLAOk&}hk}!f>YLroeR;b}-A3JFp1r{WtIX`Z(AOlpFsur*
zG+PgFCL(0q;={T;xa8>!jC^hG@`#OO0u$R+bX_icHUBg0GA5lKWhClN+A{`i{Fm$t
zl^P!cF`?Vn-lPWi&u7P^&1=5R?ULF8@AHu4@llW2GmO?70vaDOxS19X4xwY$%;QA*
zB_Hz1#};m)<ipop$|c%$VDy!yL8IP42z-7ynk|Mgt<r&Or?o);^a1+Uhd!7GT>R8j
zSS2%ckLn6EO(@$M`lX?^;-pMRW=G978VHg#7$>MNDMD1m%!w|bg{V9I%l+2Hix*o5
z3i*^7y0%4a`KLH3D*yv>zQ>0zsYS*9;x!q+o!<WRsZ!UtQ#YD7YwJ>s=oN)$q@|@(
zU4~bWLp|=eE<Z|H5e^1RIgD*Y%Tf8|RWs(G2}}9>>92*8=#nr{0;BvrLA|D-n~D@p
znqr&nDLd)Y2c&FUZDzhz%wS@1(Xp8`1>UCd7Y^pB5?xCM4jAUGVNJ@mE(M*?2E>A#
ziAIG|a2RKgSJcApsm$1=L#{f;@Ui}dt-XXnx<c0D#D(6W@?d}I-Z2~KC*jv87v|OR
zoa|meS@Ps!&k34}>+dQV%R77rTKHRy_$C`=hDw7#L?i`F<0v`W%0VG00_!LiXhw&!
zV5~zR<#$zR_3jOP=Y6U!ExtK8aLZ#aI_*xQKe3LC6oH@>t<g+;%hT+1F#I+r<<M_)
zVJl^s;wpk%{7*Y#KJg9+H$uW7E&dfAGrK2vXWt%{yHn&B^C9Aap@dWaKDo&{dAG(P
zOq(_+|MI59EAl~=oz$osHKZid6t)KADRz%Y&kWp}4Zv)NNV8hf-v`X)lYOzBaha4n
z-xSj$Nt=Trq(;1OuaB~$Q_6_r_D0$1drWg+5J<2ewapd-jjQ`YY%^t|YGrEgmbBeS
z)|nmhio7pu^W#?LLLR<e+M1fR-TLWk9oX`Ykb)3S!^=xIs{~L0q{JSWU%n>@Y9me~
zM!O?Fc6nC6dv7@xMgz?5;^5fg(U^ZEqi;*(lCG-csbcm?*PIJ26BwT1SR2bt;XFm+
z2Ze77`NoFX_`18K!sKUKn!6pdAD&agz=>J4vL~GT;?%pBc~+g==`&?eOv$rB(u%So
zq1e>$tB6{`e_z7yk|V_emYLm2x$FBIfXpW9PamjBa@#5G`0U)5*EE==lA&mEy=v<u
zK-J)HUk&>qT9=u+3(5PES66P@wfS3@XFD%nqUS+ImtqY1y3{lI6BuB>EBVC)CXS=8
z>(t}Z#|2-EJt+;9?t7{qV8*>Px#erasF`FHH(}9spB}2q7<dCVb6*5;;sBomB~rH9
zwd>wSf)jk-=b#>6MVQCc&|~3$U}K0$JqP3Ptzlddz{6fET|0zzVdF#VjNd=l8Yn$_
zNklAcBK!pK*Sg1_uUIsU8EWgKpKE@RLn{uHaaL?E&D@A_Rz;N$ON1|e%Q#U!<dRH>
z7cksy@K(4&CKOFLtuS+eqTN%vt^*I9tAaJ6;^ItIn<rC@LP1^6Kh$7nUiB$!7d!~a
z+flostpOk)Z!=jRuhBBIkV2wo%J%!VnS)jX3oT!Zy~|9_uDrL~?dy;vSa2%_Wa$?>
zBCqyVD6DD>$MCafm@t6R?p_?bTlr=__o)dwj)u^o#mONi$_H>d0}~i>lIy!7umJFK
z*qu8gCk5o1YMyPew3Hr1*HjcFfQWxtX<({~WQR@MaJN&|Xb!hZu9q!b<QbX%yj*ox
zd#Hc#-X9BdZvRRxBD<H*A#T!?J<in`u&LT*l>^X~>--vR^ybN!8_g||3hZ{B?_OHe
z4XP2>3NxW%4eCApfZ*Oknrt^_1lYVObnN*=1~a$E#keCo(H_)|$qs#QK`18><`G{z
z{Vwuy=49_~%(T>;lo=T;(ZMWt9YQ|ehF%2Ka(R#P9l5i+bo9Z`KlsqzVY*eS(AUAZ
z=Ti*PZGln_F77?Ga<0XaJJsb{NBJn76&>1`ocb=yp`vZMK}PDa?c>_b6E(BJ;0r!C
zJ<%})<CE^Tzi4JiHW-u9`@#l;Q1)G9Fw+*kinBDQ2I_mL)?<2|?B|Lj>Njgsh~N0x
z^Rxce-8oP9I5tmT!MgwRrG=~o%t+Gn8BMgRS!=9NJ64GHf7^~w53$o3Iz?b34Uz7f
z;)CbeZ>nbw&^B|ZT)ul<?+Dt2QZUTN!ym`xMCp*>%!qYg*XrIEH5j92*AY6q1s)f#
zy_<sDEE@ssC=>^Dp3Io43dn=IMqnK5{(EC3`P{yLUN7qs^~DHQ0S<TW%)sFf=+v8O
zsjaFJ+q*FQvpet2D0K6THj3MyaY$=Y030~lFoZLtSYLM!piYwOz|HUqY~ZoUOjI(z
z4S4Zg^K0*hv2~Q>Ve}2R+|AXMh{Hr7%o9;q!`F?_024=KFK6k~C*QlyPF{W<!+OHM
zOy2|Wiy;RF4hvlvQ&I6;BBM={)kG@7?NpK`P)kz0d%BW+uSt0W(!rR1^&<T6hOc~_
z$RuqF(6x2WZAs9W7J^gN<>=a8dMdN$+3bCIxm4OSYwI!$0CJRseGDJ!CGX#P*n@t+
zDp4dP+2{^tJ0v%oeMHd*IPsrj{FgjG4xA$JH^dC!-Y<t!q7EMU0k(xhW;s{GoD5H*
zdq-m+NW5umB>Stf<~cYTNIQ}6fFvToTVky#GyDE9UlQlOiXaDf@cPIbBlk;Vpqi)y
z^AXIa_qWN?^~>T<%9kRit9fBYvEWwt?a}Ox0J^qNqQR^y9x`FNoIW*B2r;*=FxTND
zb5SC`#J9n=zJ=WmL)-y)8Uws*$h0*sdnefskXz1=Zztu?QYi6O;K&;HHyA%n>~hrI
z3oSuI<&HC@o*lD;yJtV$5>GTEXNR1jSSNjizSF+`C5>z3qv5vxJ@qMh`Vib6W5tJF
zcNIATg*z@`4B@O?IO^@nK$^imJ<SpD*_HQZJE*Ouef8*QjSEiSh(%LTQ;hp?8(|5C
zk)ClMG4YTCdjuaQ6U-`b^YBc{By0CY+Ga|lh0@^7L?7JAQ<;O&N+BNyTU0Yx(F}=B
zh;yOq@|9<zY?yg@TIVnzSV6~zfzPG)FX$ih5jGA9fQ#aDBRCvAin$?Ad<xCoFjA=~
zL~>nJWaK6Sm-fFpib0~^S6!>GzZH!VeS@+D4Oz96c7y5aSpg(tj>V}fD&$|U9me#;
zipxpn+xz4xHfBJ=Jf-Z1tF5Omyd7N4zrT2fge%|b>?YTP^}G~naxzpWZe1sGGf3aO
z?;FEpHnj^U;;<9M9!&enkGHn0=}ve^md5zi-`A&o(J;My&E;IbL$@~gtr2Kqz`v^N
zaKbr)l&atexOe0b`RQ$NURYCym{e1Lyj`x4Ai8kaCPV2hKH{2*7qFGi_4QW0j@g*B
zoctn-o@YLK@y!~6y%!9R=C;7XBstBz|Hm!277(g>HlKiI)7|6$On7`ug|db{to*ZD
z>@H+OLUb1Z@UC~Su1qhC*kd%2yiJMGR<b+ZiA=mPHLq{#_=xS4@5-_35LgqQ>oicw
z)~y4R3YmV_fvJ(*>@v9;UPBIl1VNgREL`lSv8C_h`ydQX8BMkM{DpuOSlp8DU&$!1
z@5o$sr1dORiLar9DY<w^AJ&JhgZJ$|JMizF&+AZknWgeC==oYzL2oo)`TH|po?6|3
zG4sjE-bS7AbxUL7o$H?5B%xTgm8F6+)Yu82)RYgA_wLOk%%J%0s=&g<%9k5gm;v%%
zbIYS$*n^=KHz+DFya?sWN3ycA4t*$0zlz6wK+e@rW`$dHMf3$IFfcfFu6qINB1o1y
zyKR?sPx?eEB$EWcuDz{ACzi}aFA-ApDJyGXcP$?q@aN+>pHypR8keo@`q<;p$_<GP
z#~dTFRCMhsBidEGdSQxVq7N>l?Bg}FNc>KpDgLt#*}k0{`g-M_m)$OI>qUC#<dpSB
zJ-vk<+0WM=detq#Kp)s9zS%K;3|W#N&5AUIFvtwPy*wU1J9MHh+qo*DUA}(woz1rn
zQB{hY*=2!5HJ&T0Yu9tlmL`l$vr|+^y3aEnyj=YPGZLfTV4xAu(1na8(cp`Z^ty?X
zd08*q=os~PFVP@3sX5vZlb-^HNW5^&Q~%NEyM05*9rr!eyrZFA7D`1@2yeQ2iEHAZ
zuF2^1N-^~fopc$P9ZymC>ddERTsJ0HA^#SBpF&Zu_U7^zSDL^moaBW|3_^l|lk7!!
z7Ut8Ilp`j2F;N&QFuXUl?i8XIqI?&xD?O%`gIfkW%MO3=BMPM|$UK5O-J(v0X%sA;
zt3EJxdZKUT6&xiEnHT3u+HVBUZY@FqK>3)9i3{0#9<K04HVQKYwEN!P8ya(EUt*f1
zL~KWTdb&w<{NX_au2-`};$xA9ot;Q-?ymAk!yKN%vjyHWf3;&B=moH~=pNmotUN9L
zhW_PxVP(xTqr7p1=tu*caPuvLcEK@i>}z1(Fcxs*(ZQ@N-;S5;vuVu(08j|>&z3BE
z7yh3{ia{SdDua}Ez!dFuCbcWfdLy%ES?NbMN^HEBmsg^$vr%pf@mpRD_{@gFZ5dhW
zhqx{b64r*K6B$S6NG1p64gt{VIWcnb+*xqeun)#~Ow6{<PKA5C#c?A(k=$rXYU-8*
zP5C<k22GWHnec{txHP+E)pv@vA|sNdHL2$~oh&o*F<;*5ADajKR@i+pOP4*!Dwqtm
z`rR6p%!N*@)pJzHbw)MJU5vdD07>EafJ_^TQTPen)0=*~wql;EkFHulb%Wd~$h{EL
zCqn4G<6mxxM+(FM1oGFs*m~1oYRa@sUj3&ciOn)k)(Zr3`Zi29c|`Jsaak8sWY(dR
zdvwF-R+jkJV=9X(i%ho{Nj3CNWRS|bs^tl^>E;_y(uOg9vv>R4`oee0C9735b|=ww
zh>hiA|405k&_*<DSLml>R&Y#fka?1{Pgg}#vA?{3;fBdGxpEo1bF;T%R!#VdB2lj|
zIDQLLS;qlt`YpR$e>eR4!9(ir>d{?YpPX(5efv;Np=owz6=q_kTHk#4?%jQukbbV=
z#bH8c3<<1ZdcL!HZ)6a`b)YTWE=xI<-j8muuab_}J>CYvUxE4qO5S5m04Q=3@0&?{
zA6hd+KroE=P1gKYRrD5h>bGyXwyw_na$`-*aCUcc?w&r^jtULeM5q>IUfadpTha$d
z(r4|FUN<e9)&}g+D(*4lxbvAD=i3`=tcx&5r5<DLzns#WRNAHIat2&l(kfhn=}(o4
z56cZ+2ZCjzYRB^Y{fWPn_n5#mn2prq4A|CqsGVf0^xznV5kGlsz=%uwqAp4s{4qk%
z<nyb8#3ly?egE=DJnAeu^!l=pVjE1dXq-?JvvS?@=uth!JV}$ngy?9%f_rz7_#D1_
zs{Pl&+MH$_A5k%}_D|-T#&2$zL#?y}k+UP927z(V!+i8a)kjQB++x?EfBWIl-K9A6
zJT^%<gRPM24qHS39cYk#Ti~-B>AQebkJ+ZXCH&fZN<-;?qWlWkJi$@`V=>d8R;C@i
zd9T&cV{{Z_idQ?2)TNnPHr|6TcJA}5t(jwASaHxp02l^0y-&myKpRd7AZKX%=q4Zo
zGijUC9&$s0$3~@RCUa+-y(~`%oIVfMk4*;S#Q_uRgAr_z35<`9AEc;Clf0Q~Ee4s+
zn1?6NR(ZI10Mga?$IMd^9N5-RoQI+T?^Cug9xfwz1m(Y{)9TK$Di<Om68|R>ZL{)&
X>rxcFgBRM6r=cBGIgqsf^rim;eGeT-

diff --git a/articles/Integrating-User-Defined-Functions-into-rxode2_files/figure-html/convertC-1.png b/articles/Integrating-User-Defined-Functions-into-rxode2_files/figure-html/convertC-1.png
index 2a5f6b9f6f1a35ef1eb76daad7fac6184ce64f7c..46a074fbccb1f95bb4be4210379e718179febfd0 100644
GIT binary patch
literal 40398
zcmeFaXINBQ)-_t@HlXcML`C7KC<+J)B9hIV0tpf&A0<gfKyor4BU&WMA_4*eDmh~$
z2Z@q%mYnk)tFYhiyFc&!^Zsd{=d^*Uz1LoAt~tjXW31f|<zyrnHtgO&p->p8f1Q`7
zP}bN{D66jiYc2job7h4Vep`D@O5!|z$FG@l>8|)~y~SUuRuszK2jssM?wuog6w03z
z>iIK@b^*T{?VY4;7js9ZiY?>WSFc}x_PLz;FD0YbYKcB?%j6{^Usu_LS1cH~RA@(3
z$fze&^!@W8;jL%g`Y=z=XCdBd+fGdWbJXvAl7EAvt7}L1qHfRay!_S)ClL_`=WO<=
z40C2`rqy8gl>nIt1^SrqKNUClg#PpAmhW=IE0_P`M!o#M^+)@idnq~Q9xralW0-@3
zgB#Yb=M&oO#;hp!Jnk`5sT6H~Vb%73%jLJ6&*9<Z=JqUB{*(Gq&AFrXD*vgi9JlA^
zMHa>@)=+8LuCG?EyuC^0Zu9F`EwnwG)7kQK+2wXT{?>9nSs<b!WPW}Rm6o3VAR}Xb
zlZ>vIL1}oQS#sgE4{seN{4VB39vd7ljT!4Mefu`=5Oo6sgO5LR4fo3J8E(`)A}JD^
zADnAy62spfYNmO{R8}_9-qa=bUmkw_nUkwX!DdpGDlaE@cnuF5RZvhhHcoHr)`=r_
z()$Gj6i%vVvE|p-YwviR{^_Ba`Q`_js;awaY&<;K!m)QbH;OOXoxXQ(jis$^L`&xF
zFRdwUR)tf~<{E@-em%Fe{K!V#w~zm;f5nQ$?Y~#e&G3J1rT%+kvIhT)TjjqC@ZSaa
z-%Hg0jlTaz-~Zp}qZj$7R;Tu$uDz5AI2H2b$tFF0eGYbZiIe`)w2py+7&kYHQl$I^
zexpjcfx1M-9Xk#Q2)HDzB6Ci$5PsFN=F(TQ<o@B-69WytQGvp?M$K8y99&!(vE9RC
zV+!rxpE7RS7Mf&QA7xk(j&hoP;o*wq?{Xq=83VqwmOa}m?D;a~isKZEoR-$2w5Yb3
zS!|M7V^oZKGRMAsXKvlP_2}u-sykc6-kvnAv$)Aa|GL{lfFwpm8=E9=Zq2BB8=3k#
z@;z!q4*mJ(>8HE-<r~v(GP1D5awtU|K6VV3)v^={EGR8i!RJeY#p4trWO{R1T@{Ut
zqDIHYqyi0UKkBI^nfRVPd$uw{mUW`PMm8iQq;o?4_gL>7w!5{V>+`*hexi;^JGeEj
zME&E%W@~GkXx6B<SIAl}-DW^B*{rcYtF0Q7Q5JHpF3UOnq<Pc3f%;^?87W#xkZ82)
z;#jcXah>~v*-I`A%qK7X`0+z;vX1^V6uBgFhyB9B>hb!;<xJX6H>xxm-@ZLeMxvQ+
zqa-I+CLNgD*r<swDt@`oz;8GFx1OG!A)Zt1!?k-hLrq5`b-BL!@f#=T7w>MgDiw7Y
z&1bKAL-p_%TAnaU0m~Ymzu&jDT|0W^{ppHuX$8!Mabt0@^i;EJHj917R(`|s^E)1Q
zdIyLfKK$2>8#f+3eypIT=9e`;Ds@yltKsIvux+z*#8B3pOrX%s@(Nk;n7q6^b=q<$
zP-@(7GaW0N9H^g|tT*q^dvPF^|M9oigN^CN9eFRJ@C>P451%}#e9V*eHX_JD5|FN@
zJJ@xQBhbUcBhj$pkmz{Xx!rvFPj272Q^?P}JS(v_&R>LAc={Im3mA|0{0J*(M67Te
zKYrOfW8`A0PrAy6O`9rltHf*HHay)cth{2y3Yz3a0WPhKE6#J1RR+P%;jdrs|H8YB
zrrkL0R(vsO&T5=EwSWIv*ZFqVn^TRp%rAS4va+&vAHQ~&#bx?xae$!0!-o$!I5_0B
zw1ULuhtD0=OjG2^p8r^vV5noQ==mWdLu9Y8&7}yL!17e9Zci_-F#HQ66B96No>3wP
z!c8ek@$y7pl~THm>AkymHKwGO=#P_A9MZC8Q=7GrrhBmSWbFo(L9HYBbg9_fJ91~|
z;ZN(-vRyLSdVhY<Av57H-V@Rt<oN6Z!=X-Vp_?6fTTSbeq=jt<D;F2$yn=#U8cW;z
z`=jqOFmoI_bP>Bn`O>8)@09zG{<VpIKkm?$HN*b*9z5tbcb)SmJ1To&z+A{-{4Lv&
zBS)zICr!%{ClO7Vj?EvLKZpGKIbdVjoTb@c6Q=@*i->isDE0O6iNMod&waL+gl<Mg
zMrEv~aHH$e0`{a=T%7qXwsrKIsl0AkW6@ZbWEyQh)@4`{An13})X(Ik|Ht_FiSNwj
z*Q=tG90n3=6yx=!Lv`AAmzS3(=;r?sqm`-V>+4%1@*RL*IYuo?B`(OQDk`k3Oin3Q
zLsd*nY;<_|c}1^hvA%+`vUf?KaK!kyHJSbgoA;6h>bKvpaidp2z`KqPU3*8zXel3V
z61Lbx)BBcVNbl}#rW}Tm@$t&x)-P3w#wrUFajsR~N7dP=^+~1`Ik#5v+x)sBI+}N&
zNH>OpFDJSzI1~qqYhzER88u^EmqRZisb#7q8igH__G>7sN<!Nq;m%TpU{BO5++o_7
zrZh7%Lwj@k_R1KIRC!L7SVcuePqL+M_E$>>1qBH#%+d#w;(?r(a-_Ajb#|_KDOM?3
zxp!fzSp`pqr@;1Ob6%LJnT=bT<+*V2;_kg;&pa{x7as30Y|5|~vg#^Qnu}3Ol&^`^
zWMpQJMs8~upJ`}l*uHC5WM1B13_FjVM$%SKGWiju6rFJO%UVW8Mk;MHi^yB?rJ2&%
z4iDKACO?lMqqwVP(udRV=LRi`KrtQ>PeUHD$$3gMMm7E-Ey19ag@HxnO6bL><?pY4
z=@=X|vnUrnCnNLhk8Rud&6`B@3=DkgqLpL(W+w-t5kZ_>T)(`#-mAs*RYrOU-Ryhw
z>z4_ci;X+D%0}9A4^j<EgR0VpvZ8(Y42+Nv9C|_?q`YP#!*J{W>KK7dUf{*qACufv
z{`KKDr9>kos{G~4FTyYRzP|e9elIfWTK0_0&aN&}5i%X8sw*QEq`cCL{rD?{hcZGt
z3w=m}NX}?jkVU<4=y)M9-sz)l)b{nEQT01t<DvPPF^9Q<Wd6T%DA*`6G7<3_@$}sO
zxBm7FN`VG;bHd-BR!;O*<R*I4Xs`F3yMG;z5}Vu?YO;OD4x`CmwSA50$$$LuhjnlH
z!5S^>2Y%;S>*ANk&uOMvzj*NAgL#96l@;nr;0gS@Gm`J4hYxQI1d7<p^JI?Gf}BRF
zVQ=2-eEA1`=EqPsEu>bjUVZ)Z-SzZXK*UpNjb=3oh7~7}+fg@4TkdvsTbxQTspTbA
zfYoXEOeCj`j*bq$?VyHqkccX>qQJ4vL{+XG$e$OH(JPvl7HuzFxR5$&GuRNB%i?&+
zJafYP<HwJiF0CiGJ^fAM)5aEWveI3p!EwkU>DAbXC!D4&r2S4%5yNZ4`Pd^WB1yh2
zdv(U4{i%TU<S*5lCef*eP(iD%p3JEx1#9bsTSr`(k=zmg#$=}-yL!7{-yAf-;~huk
zH8cXSxeh(}a@7J{41RGOLX;&31P2Fe*>^HZ2MQ_i8&`WOT2Ayfq+0P?w48QE7GBGK
z!rjwzJ4xasA7WG*{a9RuvuAIU9L#U>vn=PXz;{$D=h^WeZQmaATldJ-9Dexj9arYa
zcjn!q4hhfp2wt<0T)rg>ZrK@*{Uv$nQswh~=Q_HSM?2X?MK#-V-D{%bvZ7Fam;>Io
z73AfW-Ce)!gl^uykz0!sS+;HsnQC&Bt%#6SDfB+daOuiPCue7Vmw7wqk#9^!^+_?t
z&fO(}4&y&0yq-V*)4y2%@3YHy+(GVGP+zZJDmoRQU+kx3G}wn~T$%0airJL4NpH-#
z%P55nL>}kcw{OV6`H3uh_Pq1ac2UJ>R7_0XaqW0~>qw#2`SW+h7N>O<I(b|#TUfkj
zi+ua`ul4KK?-sFpk2-H^@%xTzamV{9{jupM94Adg=YGYHPELBWVT<%3r_{83x|8TU
zXLD!U?n3=|!@-7Bvg5{U4T4i+)h0$q%eTpjgdCLe?rD9%>YbjR{)uY77c07s)k#j<
zWhM-}x?fG)vg4m1ySDY2ON(<w7KBctG>ja!Te0&yaEQC;<jWmgS6==6d8KBy^~(c(
zhw<x~(`}E1>_^|+Si$gja($xHj3pk_u>1|RprpiGQQk%?=(4bW(axFK+1&vN^Q+gc
z?VG64uE9>mLLH?Z6u9ayU>>w~^PVe4x@+jOYP*{}WAx?F`%R8RqNG5^u9T%!$7#ns
zV0F2&ja{ZvR(wJC@6YO`trpi2F1t@zyjq;^7Uw*0;7ejn$hmdSi!<FMJb~*rT{3)E
z$QOXRp=NY{?B#Ktai8q@3qIVMyr-5cMJE8v>3cuOx3Jc`#a%M`H&g*Gk!$o-M7W2`
z1V#k}?75@hee}VXkVNAeBeF#%s#I={_gt>g0@CHb*{4Vbi67j$v5)iLJUl#4S;e(b
zjN}()Cp7yXGw~=pIb~!mOscP{pl2VsMa-w-VdZG0a}PFG(GE&&XJYD3ljq~%ktGN%
zSj^e;`SX_Dntv-}TdQR`rM9)VPZV&*9X@)rIFSi?ABaqaTpy`;(<Rj#nnA51SO|MN
zyT9m=69`KyE_2e4TPx%G@3<*gw}!`U!;w>`uEb@{Mv*0)8%!@v4D#>@y?b{xV2c#y
zRnfsA?beYvZB!=2a+sL&Y)`4{q80Mbf&Kf(yZp@w2m{#n{&wv+_uKV$4h}|@Z!h}^
z+r|#2^>f$Es%SfhktL#1sQ}(<53MH4*-|}r3?%0~?Vo@C1t8+8ASw9>3Dq+ofGe`&
zQB$e-k_M0$DIjY%?R*8SpAtJA^A}js$)V;rAbcYX?qjV<*yMo&2PVF5lkK^+j^*w1
z8O}6^i9W-wB40m5kJwDdHuNYQoSahI4qcLym#2}W0}RBxcN-aloBHkBOm0U|2<6CQ
z0{`om1Vkdg7PeaCA&3C2>)v0z^ZE1V%4iiHQoga!Cxt%p^70a39dqTw>&k!KFH}po
z9zyafkjR^#nN-lDyLom0xu&BtL4pdn!h_AU_Tz<-k&)ekmT&HF+O=O+R`!ku9fuOA
zzv87N(rA;*oGGBRq_)$@J<`QAG-y!eG4o_no<D!p({sHM>q{o2uO?2LyL~g};u(u*
zl(hfJUVrnf9{|^JLf<PY6t;*AyRBu|t!VU)Qg9LbfM2g*J0QD;2fM5?P(HxL#7Hiy
zF~MLnojz3`^`t^|mdBz`#v*TLm*TQC*W4c&5meq+6`crVHqe~ioEYrq7gbrQNUB+s
zT4I!1vRMprdz5Otew2E0Y(rzCxzHUna36GX31C2bD7CaWm3<RyVa%c)A+s0dR073X
zxwF8_e70fQe#herNZ*%9@w6Y+J$UdSm4-||4rGGsnUO6=HmVHa@+mpRfUkcOXifkf
zstP`tU{u9H@CmA{<zB1Bd3%B%qgCR(*;Hec0hyxA8q<J%ya6L)@jD0l7#A0pk7_Xl
zMSf|OyLvUm(#mQAt;-QUK80Pob_v~RyNkN<Iw9c%<M!?ICP@IFy$CBgG;;{(17?28
zC1^+pf_<wJr$r#d&H3pYRQfF4Xurt7DyDH~-4<^jAD>TD`i@Sy0X$*G{Nlxn-2!F-
z-=FP`#?CfGSk2aCFNUEj$#1ntc@|%5B1xrQmXwtA4FbeGEF`2h)nMg5m^r1zZ`Sa#
zHjVK#g2?OT%eSby*JsCj(NQb9xMUGL^zAW|aYM?DzPd!tr5G|(l<@N?>q$Bvu54#y
z3`5f>I@94%P+VLUtC>FjWsCTAbmZi2%F5N!!R+Wb5bi3!8}pYhcdl8p#_JR3Y1MeO
z7}5keO#X`8A~t>6xcc1-wm>mwwYcoXw62oC=!jq!H5AECNn>(Pw@CEgDFeU^k$Awq
zvXDD(X#OFZ8FrU*-C1k$J=hTiy9kwlmNqpt75nyM{lLd7T%<_F#l@{$wMr?~Qt#gV
z`*^ZTTeogSd~xW#%D;R6ejh-fJksPLnSdU@dZ(706;GK3WR9vQRU&CuzP_;W_dXh~
zCnIRT!nI?~x^<k$XUC5pznFV##kFhKP8e5n#b~B0v%1cID6gnENENpERR$bi87|HI
zbm#H2fPmp&>3xgh^-X(LQyN|XMvu4L*<#fC`JMx^zfzbaV@(rSixa3<qKnh*2z|}S
zT_xnnDT-`ZVmD#Gh=>TJ<3#ITdC{f$u??Fy+YA7<$r89nP%&2Su&{7_&KvrdG@jlD
z+NCx%HIX{*&7&QQCjT;66^ntehzR4+j{Lsr82>|;UcN$3$?LV={c;ESp4%TqE&Qp1
zR{A}8FHYb#QlnE_EMiJuj%-2Vkk4LPC<K3z-IePyKXM+k-S>@28?iXWVDO$qSZzdQ
z<pN?n1s~eJ^9uFFi;2nv!zV^H&z^tXBt_|J>Ff-fp0=A9YS!+rj;R11OEhm5qaM^A
zqS0uco}O=phb<<4)hZw!UhUt%f0GOY8B>bF<zc^oMl+u;SF?>8TP-xK798AX_|t8N
zDrpC`3wpUkgL(PJGI;q)`{#z1Vupv+P?!H%xpJj>p9V!P2x)_~VK)crlpLo9tANZW
z00V$hs5B4@VJJce+1TEmxwE#fAvIpecJSToM8B2JNpk1YsUa=j80KdCmFhACRcVzU
z^_Vf>%U+!83IR#bl?e(83fT?2Wi8HfqHUKucYl+>spFJ}m+0HxQauhiwBBP%aF{Sc
zFkJ!wi(NqIlHNk{ufOC_Ty6rp^$eyDDPuAicjj$LCikQyZ(b9Z589>R`5241OQq7_
z=DkAls0^HmLtS05uU<(SRfK0=yQXG5HMsZU*VS&kj5~L}#jJkI&E-6LR2u22x<=bI
z2FNr0Sw8UQ>kE$)uK(B%*v@~w>@<d=IM!7Rre>SH-EVU1)i1eW@6bX%b8Uy-Y~4f$
zuR4E`nZ2DI9dA|RcY-JR5;j>9kS}-kvjo!BQ-^W8x0Ow<B`wmj;*q@-ORc*zhZpBu
zk4`AJp1E@$6M0`nuP{Hq8nN%?m!@rRpSod7T-fHlE)la6p|K;90cr_78>#XpzZO)y
z(wFj4PS9k|ebb@jnCw4rfHcQXckx|(u!Xf+pxG%Dwa0L2aUoU*wTQHq$bJM1(6(>i
z{$@A-+3h=&HKqN7yc!2%QRR`9L$#K(`PajCg=nOu#|M&|d&YaMFPyKu8Pna}eT1{0
zomvjIY@+Pk!p%<U{POb6g{|+ST^4$xgT~bO?BeD#D|*Yk1RS!frJJXQnxi`lnSA_0
z9mlWxJ!6qp@ZsJPrIY(pK36;AWp6=!tmXr4=SC2d=}ul|ch)kH+QNZcPr71k=LFOG
zlX|+ko<d>(2&$^8l2pMr6aCp$z`b06zq|R3-r5Y*ZP>C!1tpPG5mZA?Hn#I%bY5V)
zpNac7i3qtDm$5#WJ^I7svTpI>u&bz$m83-}Muu->Dmj9jck9-f*aZM^Dca{R#*2M|
zpFbP*$-2%)hfB|0>Fs5ocp#p^;kx9eJ^vC6Le$OvbVla-3n*3`+)2iusMJ2{-l~q7
zlZ;VsK5A}0#5o#7V^fMzi>VZ%oJOWmc4rZ-43*eIhYK##t?LGA<0Z&$3OV-x+=F5t
zSsr9QutbaTrOa3xumx`|I}4&TQa>X1YG%(bEOZVH@ra9y&rDC})zqjEu4QE81+oW7
z9^FJWG!iya+QO8a09<O%<_EuQ=9IEgij_1poHYAY7AsmUJcR~FhvG!E^NWM`hOuau
zSa<4lFKNbDcD5JJ&(BxJX^R7K6#nCSk3!k`=yviRr8uo~gj(QkwAEIrnf-JplJ-U`
zbGENSHllZWB-E(<O)6Wy8BcYj0((b$JEyq#QnOQRq0TOs!oUXq6Na%tqD!vhcOE==
z0Hhbm>MHU^)X}E*>%-8n??M5hju+*rz8cC&Cbe%WBH}bn6r*>%=1}7Bs!ip1gf-c{
zcg1PJe{I~TaOKL&G}|F=Dj2@<AE6RAhg;4pOt-sBN=a4VuU<jBPk-t9j~5__`@xry
z_)@Ex0&S4probpajN>w|TW?~5aDDr>6b&5kmOw$Prk3+vfGDHqZlE(*wa#W&KzN&l
zP$0+QycgCpR{gg>JuFZ;b0(<&hEboHDwoDrxB7EWcPA=F%6l=-UfIxOi9LSgR7w;w
zX7u~_r?s>Ws$>r~&fP)e9NGW+8P^Vhuy7U&#tfgrN+hzzj~Y`=^)r6`KYlzVlc}fg
zy{LaRo>7v9R<)4p%a2e)Noi4w;>@>i?{jdM9F4b9+HP;}^W+&NhXO7t_Qhru!fpa*
zl5`6?jTCvZy+QbS1_qkSsZl6J2SgVV(Aw)+Khd`M(L)Ib0m(V-svg>y8mtZ~=cLd|
zIemoG*-yqhkP&LWKkc`kwNC4+e5k0U6+1d=S@l5&?W_R07S4lu!H6LAw%FNH97hDr
zz1-Z4-7SU@Y;U5QB%Nl~-T;b7=GsPSSdTbE)ZYB{^El!7*kuCn)PQLKa>s=B1L(AN
zP4*9It){ftod2X=(pP2t>M3&#GPq(@<hDqVFMrYYALyrYq;6F=>RX|wx9~>Ni^F)R
zi8hn2C%Cy)d}035cjKik@`*<KXp62k7OFjUi4TNsrVaTyq6O2^N^WlKOg61IuwG5+
z%+rrQ!q2a?Zu1^F^eM>sZwN@-v7>3+N<F5UvukKJ58Xc82iP@>*>NFSqoW@SDk?al
ztKgPkp<_-9AH*tV0ja)5<O?V3Qbe%uptyLKnj`o)?J;$CPN{6Ygdc(7Z{MCish+RC
zFEhibv-Q)wvuDP(-p}`Fwtj2YDl5y$4c`=(uwR<VD(@8U7w~H>aL=x;N$Ovioi=u?
z8QXKvn1jEj+Kk6Oz{e-7t!;npfQ`NV8c#n*<7B2nGh0u;Fvl4|?$Y|3rvjF28#;A<
z)l56*OTXd~&#E44=E!uIK=a7heW+}+#0OqJsaT;zHRtU+cV4sDKUN!W;13EH*a)!8
zF|=KACu|bFz6V`dlhra&<%}#|kKSy_*%E8}HR<||WPFZ6X0<1S`4F>fgvc6m!Ho<@
zeZK$}Y;kD0H7I0v>;g?O(I{VF4i=5SXwO*1Ib2;`^>p79mX>N~Wx3q%$n@UL|M?&l
z1d9@oI?|BYtl*tH(U|d<#hsiXD;k2pCMQ~T%MhAce17<jyn+JK(Y(L77~mfS8+cSM
zCmL1ZE=s_JK56-Ca&pfBJaiMAT*OzO`}V@$@p5hLL96yVx6gL=JI)n<%bgH*oJ>H2
zxsUqjpSN*oH$xfVRB8Ksul-}KWYayS-D=J6<=<F;m>&~RjjtaFGkb3{<MSq(Wo~BX
z$ccn&UK%!zFGfctzc<?3&v5+d78hL8KWd`%_}k~~Y|ftp-wFyjeS4kQUCdI`mzHX{
zU6=k$O;<2oF@NN+0RpZn>Rt~pPC;Q|Ra0g<<JQYH6KkU6!wMjisYn)xL3WdOXeyXV
z7kG^@hSVW%W8<A<I_D!~^MqgM56aFGG^|5SnGhyCkoH~!uxV!)`mxQ-&2e&YNZ#zL
zybO@22o<90ovH$$=3x)kpcNFkt-#|8pVR-ff|4`wx`qF=$D!hhVgX}N+4_vzJ3uJ;
zrA=bHz78Fdmn?8W=^Ir}Uds^cx)_^mn_tgvnsgTX)koTb>GABu<C=(InToWV+DKHf
zMpcidYeOG{LN{+V155nzi511|!=p!!NZBQ{Ykq!yn%zjZxy4oxUj&nxHU7)qL5LBc
z0T^E;)8T2s@#Cbzw%DBeWbBrgH#^(qpEwIaiH%A}T+MGU>!350SemU#roDOpez;#u
zVxo_Kyxsjlf^p5}g+bdXubKKdl85|H+8Wm<9rZij&YCtixEb7C=Iz@Y`NhSWASzQ+
zeZU_^RllR6_dN=dDR~WOPga-dGhi%cXZz#40t20k+ld?q*hEymqiP9qzyz2V=!c=1
zj@dq6GgQb>NuSD}(<zwYqsj8kqSj00%k12`yxi{D2V4T?O_Kfs?3{x3nKX}@bhB4{
z*UO$Wr|q9sgR~P`R@SGMY8i=@dv{d*r1n?}(%D<vAs0PM8Okw9ef@CKB!o%%^i7Yn
zs|=*4rXGRH19CD>$hzlM+Rc9X7J3VBCb&Us6$luP6A5Vq6=pgx#QKS&n`Hy#00kh@
zu%b)A`k7JH+3~O2bU-h!MQT-Cvu4IpDC}dV>W2$JdkP4FMAs!Pl$o-J+j2dFgClZs
z{;Y~omo4^BRzwgTylhE179r|5c^Fs^=s3OakfE8G86iZ|YzCrGNFwFKg!|Xh34KcY
z<>B2O;1U+@X+G9+o_N|kF7oC6)7jcg5Tf9fL&VdUiuicu+pHumR@VeCj@M`xqZ5hQ
zaa1k(!sF^}2QJFXTc8!bKV{|zwK>p~ITy7+w9vsub<&BTA!qpOUu5o2ugB6+7Tv*{
zquoDY)kO%Ica=fO$;p_Y@RXEOi28p<Hj>Y}ZMcOF%pw<JH<TQ&aB2T=&?DY;8yOg?
zAOck-n`^IGyLPuH9pSkdgM6Sgd_dvKn(1Pe_T{|*iR5S^H~>-74kDqg>uCLyEJYv1
zx3A4tR-8VlV$s#naTwIyy?gg8PwgSi?&v6uG`lNT5(UNjl4nR`x~(w?(Ykm8*~p3C
ztvU#v3MGM0bdK!RYiNWPxfaWcJo1~Dw|6*NhkmHrAgkp-$OXuB^SC%6v7J444)rJ;
zYRh3>UZ^FPL85SX#yt!`Jx3^qRL5$n8;RmavZ527XrzjqH0cvzNcv_|!9ANHNL<Ha
z60I7<at!h}vr85jbqLG7;E%(qGcz-*5)2g*O={&q?hrBmh!f|l&!5ja&-O-OF}+b8
zj75Qub-q8jTpA>rdW{|ce+v-~gcHSob}D-*26O`_?04eu1ID>L!F)t1$3*4ioCXg}
z2y>8Um@Sle$n-%FE3DCj8}IlES^{dKKR|mpDmw>seL6#uJ|uRtr|b;+=;XGrHX$~4
z_I@Ht0L>T|=jKX$d9a00-<_SEdVhm!#4ocwYj`LXRziW8b9d!;R|s<pe?zxlF7!eq
zyTNiz+r}`kXNMsQqj<>w{dbGGi_{`^p7m%4b+4f1d7?D}13>OjiJC6X&;LQ_O|Yku
zz|5pc1^*aU9iwh_@48?i8ayI%A)R@v4gCmx5M<xA3-osJD_#0%JQ)zN;a-sS7&(=D
z3VE^>)zs>g=CGF_7ySU;7elNPki15RQ&M>x7Z(?ee!dUcebVGD$UZ5KU2AUil<gye
zBWY+Lh6C%fQHikuOecORhF)YKVi9;IB@2sT>CPM;R4pQch%b(8>Fn-KuMQ0johX;$
zi3EwaHk_{AlXi3Eu|Ph9lIzeJ)iUfxln!hoMITD9=y1*|D7{LUl%u8XU0vZIB`-)w
z^nb`@QGiW@SRB9)sX<G&EOr!io=qSO6`UTC#x=2o=y?75wV8dEaPg-*>n6mP#>9mz
zTHK6_(QY7?D%@7AN_n)K%!`FO!v62w0DNUdzI=(6oszW!lk$$p`AXR?3zaeI0$<{u
zO~6kA?URrP=!LH%B8g>Va<GvzQdede@*(+NqG`RdyZB-XisG1m?)jEFqO6+Moq*tO
zfG~@lm>bI0cAfshx-*x)ASp*7$-W(y2^NdNunoUGVhnpEj8+i32E(3Ha!8I*&(|29
zKJC`l)&_Du0{yPh;`|JlJ4?-~^qc+4WHA7Ns|dZ6<vh1xMu;N~Zin8d0!>8u)HDVz
zap?3?sr;ZLvC4za5n33Nd%g!L|8^Ij{`F6{*Klxi+r+5HDAP?4m^O%(hJSzV-v&Mp
z+<_)kjs&<)-hTaho-k~5d6p;(<h}~BqG5;``SHcI_V)I(J)!O=To;{c{Jveg3=$1F
zlwDg}`?uU+0c-5aw-(s7M8*XbE6-;WqZ~auH57-Civ*jn_E3hP7l#sE9wHiOZ+Y14
zc!ew>^QZ?e45i#?uZ&XS6vb9Hs!NFcWXKr{0;8u`pjq+a#rwuFC<KsORtj9E2Uf#z
zqHgR&y=4|KtArQDy07v`O`5;IKjFt-9FX8(WBZzOSFf<JPzqEzNbX=#^Li^H(jPT}
zN;94wZY9<qVl_dQ+sT}?GQXsx8mSN&iMPfN7h-XpzwvaBpuBQ)RgU%Y4&%WNWBiP%
zp8%E9r@b_*5;+0mK)<g+&}GC7gQQKrJ<26Fw+ygvIW(7XNE-@pA<*7{<vEN@MZ6<D
zKSBu!c&B2wF|-?p?m%1Qr8PBb5cU_xBgBsg3DxB+Z|R+PpeTf3PKlC$T?M%=FOdZ+
zCqCc01ykyEtGB?X%18;T;V?BA1A9a-9!Lqh2i*>8-W#L`!=ADbq98aqIX!u%9!q=<
zMA?BA!!`zkCF06%A?vpyBGWc>9WI40br<Z}fgsqjYhInXa{>{rlFAI?Nkz;40-OS=
zp#_bc=<Kh5{bwuLB`nV4msRY2gM;IcOrx~3T|h8A#QmZBiK1sE>Pn8=3K(oq7mc76
zf(c&T$xSA-<=OY<M=Lzjjf>^H-arVtK0iGIo`ZHKpTqmrt3MY&_}8F*(gg-c541NB
zGy4$`mC&AI;lNQQX&bQVlVM>Kbi#>hb#ue1!NypAR#sL*>|)_9t=bJd>5GR%bXY(p
zqJsb2#OM=M>LVD2h*}GldR2vz27D6^um})+oao<h!g#WUg@yG()l@<qn<(NdeS4$*
z`&#y+%z(IlLN}uj=2Dtr4cK~_^P`1PTA69Ldg(WMdj+BC`FJA}cTb>gQ>xMZJ|c)>
z6M+A!$g%D!OSz!~8j?f<DW%A?haz<=zrhP(-RwMN`0M9~D20e;AI_LVRw#SK$O%E1
z)oXUeX%HS0pEvTng%2gU)2o1fuZB$oV0u>cV(UbYL&*d`RG-KKTO9UwV}jnB%a|UQ
zsq|*Tq^qS_M?z`r<2VP&Ck_Gm7VS0>kv@O<!XPtOPg)<3L(-%k85cti$MiY$MMiWC
z4VnKgcpTj1c1393=yMGnkv`X=V?}%p)2$DDLmak=h=>qj`sPrRX0YpGs@U8wK2%I7
z6W3=(JCWMtQOhSlibR79yO_J63a-?TSMO04ltTG_!o+Xn$#@0ya)(CGXMb-d9*O`w
zV0oAnv9W+VN~^9<G>)jza+JP$_3GO2^vvnmSz`R?YtGIVTNqa$4jZ^tY|=HMvsYjS
zE5oFiL??ghq@gB1dDj2Yb;(6KSWFX{gDa5*Fb5l&?-~7eZq5Hdsr8kVm92pa7Kz3p
ztr~RJ#Qa!kv}~}Lr%m!-5S%6&tV#`sTRst(vV%v9%fA?&1iEaCB!FT|vLir|NnVi-
z{v8i>9ehWam^d5pWDyvFpw>%ND->P!qUm4wwDHxYQ+qa(fwMY~zu2&0gGN8#COtr(
zKmYftibg@KZ#W9A`))C(6o_D<*j60k;@ULyCfvBI)Td9M`fB4Nh{OU?yLt2yTHOj#
zxa>x*0cA&324!bwldUR>JYG~w7r|TB61Zka2paOVkkitt!Qj!O42+*fqOlup+n3rM
zRPoQ&1B6*7@-?0;fQ^V;@dl;UziQ(dp%tT~H98Ju8bLhrExu3!q=L<!0JoF#(!vyK
zW7s5|OYbCD?9Z=SyGic!>C?Ue%bKb1R!GLoAR=BNHTTDtig`4e*~ZxRmR}%R2Z7_4
z)<wR#?m9n`%TJsX4JntgvAFnI_JRL4G&Cg6F$|d)L7^rib*h1ESHM<rO&l*SEUduN
z5*m(hnDxo#HTY|sy$SSDs4G>Oj<y%+Eg?m2Ew*gKbGRw`v1Kn4#E0r^<j@=Dvz?jQ
zSmT(62k4eG>)zMgnrVmO^}y3qV)d<IoU1urTU#qBBclvsn{hE*dM8dOqRbI+E+QhC
zWeSTAk&MZ5pw^B*?v6!<Bb^PbP0`2-8b9dD$LU?f<)%q^vxnQ*jeue3CPx@)`g<dS
z^`AwpzkBbVAxYN!MxI2eg)5D?Tk0|$lTA*lN9vfK^d~M`NY5HIf?(Gn(FxPnr1oQ2
z^iFzer|3P-zOjli2NK%TXT#BWmNx_>0GWZcvy73raG}>rMe8!QHN&L`%t}yMt0EO7
zkU6}H^(!zE5Hc(FiaA}8c(&&Uk>N(#E(ive7`tQ;;0;C_Qx>U}=|Es5l(L<r^euH-
zkZw(EZ*O-HxzTo>U>eBr8AS8Bbd%q|IGhv+$dF-8kA}LtBS4fIx8&RcPa$w@U%?|3
zH#iS*HOZLz0UKa5G(n~O#<g@e<on9*<@G!4igP+=0HYxIzk}C@T8{7=2jq=JIU*Sj
zVXS{$3^yZguJ2EF!SVLQxDoC&^X4r0AvdWQHtmCW1X|=ENQwNO!}RRhB98sBHWE0E
zq!r?A1rMkMqYbI9z)ZMUwCmEM!><py+yD4OJ~oxiD4Z|XQRT2pG$J#Bp2+r50G3YG
zr}xy9oJ0VnwRk>KTq_ZSKw8qo7y|-aFTVB!9B?EL5a%rMDs&b5!~NMn+XJW*85s!(
z{Q+i1_(C(}=isdR0YhgM?xJq#LLV}vVdn;|S1bRuVM94dq!O&oitzU!@6x`sK1VP6
zpY{T>C`lI><vYNi2J71X-!{;n@6SM1QYNiKE_VFc0~eo^5vxF90gkbeUk#_o4c2em
zSmC-f<!Vg4889xB?gt)4YB>b^6R-%7wiemhs3tZLi%0bC3+K<z_dP8Em?6;%97Ia>
z?Bd+eO)P%&Gb=dUut$jb7#^-)-Wf9VyA5aK$lS~73Fo1unoW7U@A$aB94h?Roj?8m
z`TN4E<3B3jB~7H9pc<E-jPMyqjrHiN45vEEl)sVM-QSy9*VRY0ykzXZQ|AxkKcr6_
z<`~?1{-mnz`O|9+t~_u~@19yGK6v)L-ma8x78dvZ0<WQlsfOk`=h8KNqXDH0GiU4n
zZ5Of^*yy_0Ty1F=z(p~P;xAFCcrZiUP1)GI4Val!*p_K~o_$PAOzdsWc16eL1+PF@
zVj>uZUsNe6DcDUsySlm(l9IAIt)A^XzMplzW!+fF7U#nYb5r8&aa-}EZWZjqXN9Z5
zXf{|^NT<Sg2LJ35jY(T0ptDe=bE3vw(_a1X_VS-U?HU+U&BNe5o|u?O9WCIrs!~$X
zD|&f6=_;A9!+a}wJa2&)+q8@KFhr`NxP`iP3fkfaku{6pZHtFc^A5c{YTu<0xl@Df
zKso{FpbUVAcM02=KqY>TtsQ}4XE5VMu2|;1TI=vRet!PczNlzR+>FtjbNhzcEN4eJ
z0e!KKWGvtwWKq~FU~GxDBHLc*&s$7F{}dDyD5|JLq@|@*hLUSpcq@84kUvl;GEZ~p
z+=*QHYI2N{;>LS$`8cBIkz4dm|J9}S8RQ2PJxTB-_;pxeCEad*+V>ATdS)r`{O=|H
zI~V_ri~sHsk{<rQ<O?AmdKT*WC&|<}A3S2{yDJ}rzC3U*I-1--o1LANcVBaQ-2sx1
zPBTN9b%y6*$q9rMjZHPnNe{%pNW+cXy}rkFE<TJ{_3FTf3xtzM2thjUjt^d_xYi(g
z4aTUMUk2j-gsr>IOA*>n&uVpLQ(04)8A#hy+O1XVy85-391h6_1+dkkk6auaT8#B<
zUs3kyUnKLDT?8xW<2vKxdc<Y^9*@&LfRAGBr3vkbs3;C9ukBz1m+Hqi=eMma`|<np
zEHL7lK$xKNf;7DpEGC-Y>Xi{Z{gwSu@KlVsb2Qu7in5X4p9O)D)ixp_KKnywFdgNl
z0%)EsKf`{kyt2|-*<jTx9eS31#qdru{UqF_GPKI->aLH4+`?UpQBhGDOAAA?wCx-2
z&3CYi3-<G^{Lh8gqB&6z$(pP)R#2Vka<$a9ZM^xdJ$T`lfuX6XC};eNvK#cgUbYs-
zFLjZu*}iD^IYEq<s;Q|-dlM3}kAZ=~K<^|)h?_oRG73NjS}rr4Z22Hx6|}Wo8%m==
zgE%(NHMQlsi_P|^`fhQ1CPiw5Tg5o~^zH6(m|!5TjLz<?xyB)G7IE}`29{iuk`?p{
z5)cl*S(=>_{EFEF-9RHkF!NAph=KiRdkw5{?~}{-K4)xf3_E);8}XBKr*_bi7v@o0
zHEw)k{N^5v=9o{2m9p2LUWOuCt5md$!PR=OfmD|TxM?0a4&sxGMH~id0~1_TZ278R
zZm-$+s31_(;6Gk~Tp(`YklagwLi{kk%FsmM`8u+FgMse<^&MDy9_1>8UlI(-wH|5`
zMV~H7OGrr2MLYWc`Wz0DJ<dfFAkvzP9yim@Nq37j<&^63gTF?HV*tY-m4=c$-dELs
zmsJ>AUfYM9jX9C@^={b?b5nwy6K}Mwhb|5ffTWY6eTjA(_<~RTo77!=0%2z-a`p3F
z4+j&NoFBQP56ee4RA+gDQnK~;okCTHwXA<!U|*OH<teZk*UF+;yrx(BQ)^;kV(c;Z
z9&F$;i`Ot|Sd@3;%BM7!zj^!n_wRTxHfmpge+3vN53$@I^+H@~32%$_SG6?kBfB!^
z$f=JG0#q=NwvydF03R2h?)G9+&318y<W+tseuc&0@3CT0U)wpsn-w%_57*xn5wY3V
zrWA|)zlS2Rt_gq@KYP{`ex_wji!MDjEjB2}YWTAC!ajb(|DFPFX(6=y1uC+EGy;<L
zCN>se-&(NKO4H)gnqDkBO&4>8RWs#1K<Qq)ywc_9J09gCxCLXZFwaLwV9<WST?q<X
zG-hmth1pj6l(J<aBYbN|?-M*Z^K5k!n<}2eFnEib)dPBLu~S>tyvkgDBsL5&foXZJ
z?w$>Ke*gY`HY%iAzQ1$aD#W^BIRgrdW6MYHZw2bFLX7=?hVvN{7H>P$#76cST0WV%
z0H}Do1k5yOy0`)e#$v<ZZEjX~mT!6$t@O9{joc}(ic1_Zg*)8QrgWYZ=J`*^v$tMa
z({zB-&O;a6j;W2-|LmL3FBT9!K>emRO$dYy8I*v-%QNyBoST}6Ss`Nv!cYp3(Swas
z`Q2Y3a@XUnR{c&57PmLtJ9~T>?7MJPRh6mDAhz6-y}~D{AHi1vyv7!nFr3b%!^NES
z<RHn%|Ld9R#^)tgnp@g2@=!9iE#I^3M?GIc=VH3V&j(Cy9>2dQRB>ynhcN^5$bOcd
zJGQ}u?bkMfBY*Yt^9LlJT|#Kn*wE--LK5&7Uq{*dczIfSdU6~ODG%FVJwqkjWoRVh
zB5#;3?x9$7Ew9U~1L`NN2kQ9I5*Ty{{=(kLUtYY`I3p+u`>m9v9p}A*0=_ZXK9rlb
z5iC||o)x9+20c4ltkiNGNNgJc=~D7@Y>#tDA{3o~58W$X>HNM-K<~zPdyhlvB_a)C
z$j)H-#%}S9cIY9rH1=hnI{}s0{zR0C(kZ+AgzxnVz1v36c9!&5?akaWgy_VdR>(E|
z4sMAgxl;<ynjX~LtESiq)32DoClS}eh#UORU)MFE?e($vgHk6<pWZrOmxbA|@bE*_
z(Z1-oE>K$sgZ||<NeAt2Ha%c0FVXg#viK9G-y;LZC<*Gz*F87cx~Ap)axvPxcdwnw
zyDOJJQz_R=>0`P60QP2ektERH$@}SbSlAk8B_go>sGHX{RaaO4A-nptBir(i6jfDG
zi@&o`)zs8Jf~Ebxf7W@{_&<SA7a2O$dNYKd@=St0A2t|cOW^VlWXLv-d?zw5YkP~y
zt)0FtUj@Q9tA6p>9$&UvaJEV2&02N0DA#S5N0@6;LCL9K{@~ti|7V7Dn6&J!j$FZ|
zeU=N|v#OdJCigX36c=B501aFryHz*uIU7luB&O-({r#A>y8id^c<(?ue666A1uYM+
zge(^%Jq<EN=F0=<ezrNI(1(Yom5q`huso?&%EK#%m4E-($jFG-q9uoodhXmgqA#{R
zq+I8r|LpK8&`VOJdcT8bkK9I!ejB;9-`!cYMUtdfQ6zr<!?SlS(lmsP7g_M-{Z!zF
zYOxF><km(-NAuYYw=6#jI79lQ{7GLH-%}sAn1{&ajdm7Fi7-$GZqc7!ukH0?ZVs0E
z@D*<7>8tSZI4Vio2<nW~sr8hBd%r(L;zqBt)wO^J)rJg_WqCx006usjTCh=ZTn&&+
z1CMo>pB{+?c{tzqI9GXO<?;Y_2?`BE2ur!^@ju_bh&jVaG7d6l>*#al`2u25I23p)
zdCX3vPZGZJ%i!mUqO$T1l@aKp0yqW3F>;>bNn0MDK9u%v?LJx#k+`~?(ekByk0^w1
z$4a|dEkgo5eN<3-P|bqm3P(2hx^0CV$L|m0K`Lu%*r|SgerW5jkmq1revUF&tu;hk
zz_pSo30scxT|#)8EAL;C`$;ZeVjn{F_arK6Y6haFv*p8_00gH<uC`|RYOrNjJj&(C
z7(Tn6eZ?!q<p@wf<%5cnMXvUHKyY(}lU_c*HK)F=P7PN$RsZ|LV1`Se&(YYVrKRn0
z)K*etE5+dVGn#--`_>LMCW%~Y!}7>J6mqA(vzrGR^F-iCqO_6@|0E7<DJm%)q-xo<
z-unf%9@AlQlD<pizL(&zr3M&RDh-^19)wG>4)G{_<eLJ}o!Ul#P9-rR|GD$vBhJ{v
zET#SeJU|o&1J?eZPb2J~Fbi-~KOzazcM|TCw7ip66#4N-+s*&farZL&H>HGwK1V`s
z(XnY!cJp}$+{t)L4*mXDxQg|la#3kj(JI}5cSjKc0Sxp<w5FhRrUJb_W?jOuFrL;E
zlz^klKlyi!hL!Zlw1N1Xs+zjG)F}pv!E#02dmeQc)8)aY3g;A#yC{=MuAonH&*|9A
z$-1@?cN`x7LuO@8?Q#sS8-)la?&bAhDRrsr+!j5g4l*x%%k#4N4d_Q-SomqWIQW9|
zuQCtWSGHVQ4&FO6Pn)NowvC|U5M@$d)swS)pTD`Y=SRT~u~9#o)PBTy_7t)T;mfo0
z2?5`ApV{oN%REj|rQfZfSR7i;Vk=-}xEUE0B}o&3fI!AizZJb6y;=n>(I7T^ah8pW
zVmX*GR)Rc`<NNy&Cj7H!{OLfl<I_6UBG_EXb6+=Jo-H|aO+2w}N4Ia7iEy2bAPGbr
z>6AiwwQBAwyVwgBQ3p1`45lG~W)gO$zn}bFTAENWnh#8War3<NW(ew^aZ#z|i(0@>
zZgqbPYY>~^_3L@r3-#GvcKa!HkLa6suLer{Xl<!W()5Tx*2q(m1n+KJQMQ~oPpu=J
z6YJve0}nPhH1oP~!Wk+@#8-tC7HRY;JSE}O#TT59v!66JST&5Y<~w;J$mYf;e&a+5
zoESdfIws)i0lk%-N*Xh#e)VRK5k*SCW_oay>;o7qHBD|*o}V_>o>OD7Yq|A@%<7!p
z<qE%NtUr6PKXs<d|1qoU$Ybpy9JbNG1D;Z%KVZNev4vjQZWLS@dJ|R~yr7DAKP>(s
zyT;TU9{SpB*X)Jx-uJ&7-Iia`@7@*pWQ~XI5wSaF)m<WuyiERj;^fILh2U954$heM
zq4T6n<D7y<kjrc@0^<~p_o>0X6Uasu5AOo-Ms4jJE#xVMO5lt1^@r@Hh-GMDr3mhJ
ztj2YrW~PlP`z4H55mcJF%b33^PU2(@-{W}$h;}gO-!hi~o(CcJHMlM<fJ2<SA#Q--
zVv+trl_GOs-Cuna(w?apv@SwQ8rWaEF6^?GoT%x3kIC$QtUZ6tWiHt@-L4+G>Wk+2
z;^rb8^pHBWD<j_9+k1X!X(4;;g}W0Q;<d)V%?vw6K}E#}_}~r1fZT>;jD$|xF3$Fk
zF}aQ~)gnLE?%7-SgpqXJ7SG63xz3p{9R<28!I`_xU-ij6;7f~K7vh%qu`1a?;`3j`
z6a0K^2+2Fuyc9a$F21M(@bM^FMNv@~e6~HhQy*w6x{xv;+8=IeE-$A680;g*<@wOM
zlXbXvb|2brcV2^%fca0X^Jf+dU1xZb4u`y>@5R%!jymJGV(|N8Go`L#s6Udl`QZoR
zMFE0VqxrHfyij<Zk?ob+jtJK93z(O{s}y|ye$As~CG6oltn+`HrFMEJH=2duf&atZ
z>ft;{DpXL*q0A|L5+~3fvSzE|x)9~|WSw+>zQ1bF-kD=K)=zA3uim21S;A`r{F)aU
zo29!;fDIE95;}&K7Kg@Kx452ap6?QfhwtDI+5FVEa)&#9NCfBYwRIYJnc`Th=hK<q
zykNH2y)+-pp6Fu+yUl2?tgBEnPV(`A8*dvRq%`#W0M>;b3D5(kg+3I6yYxM}CloUG
z%tG_h0<>09lHx`tA*6N$FU<sxk(5Y6Yz)V$SjHypPATkWX10QS-xcKO@^aO>&4;J}
zKcgzH6Dl(gT&EvYR#gR#rasuTi&k%*E$WKk@kh~R%ZDp{ltjP`5_m8=@1%TVH@<H>
zh#r@71BbcxFOi{hU3vc}dCW10!0c4GwLS5?dJxSb5h$pNmiY&LbPGI>_@26P4RA<G
zL~!o~WI-3a1f(1G(o&#@$I04O#r&xQN%iH)=FKM^V^MB6$eUQO1$jCjvIqS|6PX|W
zqWV#nx;V<S#EbsjB)<E0620<M%#C5=o2jeG^7i6X>B8F`T<}hh*0!kc=>A|O)5q>4
zYf9UrwN(&U>XP4yBd{NqAaXPl9(hz(R(fX7MrRjz$cm|E#2^@ED#RBmG@K`@(3?a@
zM~~UMPT9_U+T!wDcrfL_+|*Dhw)&`W_L%VU{96_tUF515qz7fr`pCGr(r&S-)Oq)%
z5%EzJ_zoQZEQPcg2pwzhm_74JlRWbM;V)a*sZgVRrFb$=fCec_Z<>g0t8ZzwLvxn}
z_Nbe6X@u2XHBlI)eSRi*$sSfA*Nef~Kp-TtVfLbxIuBZr)PQ7FlQo$r0-M3iXQr&@
zIEfPUV^@pr@{p(`JE3>;LjOuOB*|J(=U(FF7JB)#gZ4N^?Vft73b})9yrMv10WUAF
z)bVnuv0Gc5{v5I|bu5Q(?~klM>ipMi+|lt&c#uHW$H@~&;lVRS2BSo&L9sV5^+to<
z4O?HqUfYHjq(V?3&^R3zXtL8;9Nw~YOrXKyPE|(2MtX|2C_snJYuS<aWRD<kji#+8
zO&1O`)ii5EGC1-*DBSX}hX;2I_Z8t~EZijZnq@CcY9MnF1xP_%UA@5P%tIWv6<L~E
zT5K-z7Z`e3Z-7S?pYagSN2?{ptAgXCqp79N++>bPwlZl<;|CBg#$l+Sbu4y&CO6vn
zXWAbTq*tBm3Zf3<dE>nmd~j&7<s$*}x-QO(fFHBRdo4Z!d9e}23MZ{&o#yJBOXixU
zvKkNok<rndojvRwpx|oZkt4(;+-irY&<;JvN`!gvt{^EeUZs7K6_)T&hKNj5D$ESp
zW~*noHE#vGme0EIgH;c2x(M4f=mP}Ci3l%J$r|+%!OKyoG`y%lknohFxSc@oY-L1U
zad0DU5Dq#JWJU?n+()2O(mbOi|3$QP(3Rsu&EjxV$T~YabEmZ2h660JQpald>ZWbO
zLrX3OTF#04kR^iO7?f(|HrRFtE(O$RYX^(ZhL!^F2jx1?V;k(n+nJ=y(g#Wr1Ih8n
zM-2sE)L16~k{Ha4hN*tIriu>GA7_CgY=^QUqtpMyxbufw574GRK~cAe#YtIXf~Pb<
zjF9a;iZfwd55(uM!kg&}!LJMbgguV!lT6hW6&+hJG+#30>Cgf2D-PZx4}CwecoL5F
z>C>mEL<+FQOeDF-;JplSTAAh0;fR2S!nNzz-=~d?jY;3QX|K=;$g+RopjJtFw4pf8
z%8_92gmJPEFkwD?!QT-gq`Fz^K>~wdTabT!#61ozy7o+g$|%e?-O<_$MkjF^n2lgY
zFt#4pJDoU*T!dO!(pSwuvH;XVlJX1y#YxoO_%{Er2c8A)tm!uLxW4?x%4@f<IIWH7
zy!3B7*&J+2`NMmJ2A%JnZ9O_9>|1ImE2H0Z;jvbc>wNB1MV;EIE@beNXtJd3Tkm!A
zTKCK#$lb$K;+hPuX&~hmkaZ<>p(sdH7@n&_$j*T<p^48;4oJ~ar#-AtRzEH<bu|uU
z?Pt<BHu|pUCYuN>M)AJlt}ejS-pKs{<*7j36{xpB17M2=kBXbaf3{21!J0sCsN%<P
zd_ND7EL_yrj8^*u&a{K+B9vtO^&hXfG*X6BZe;*o3N%_Pm}QK7AOAH|^0-*iu=*XL
z87?lODv4r?s*IeqfF0D1$N`&JL{33kzo}mu6!@4s?QOA2?v4%j2+u$BjmPB$+1+?u
zO&+FCcE$;hm`@TWSZJZ-W*0}IxdIfQBy9r<NWFz+$@W`53MUOpWXCEMWZ9_UGMOi`
zM>$pAq@^8#mq!x$5U#P0mYs#ZeBB{{Qd>qpxI4f8x(zZ!Sx|uM;tV|&9kZt#Qz2K6
zH4eEX?<~Y?0D@31B4W*SH?#cm!-^-)MCFm93O{`TdaKc2@ufS0=LV1-*rE6K0@K5%
z;zt~VNXl?37a(5{u8lC5M~@zbW|4cG$$w4{>=!KiC48kK3K=m7Gmz7$hYGx)Ji*%~
zHD)?zEvw=ASfh`K079AJg&ZX?H^gIKV~&{(O-zdV<FZkbCo=|8VN>&Scj`Dm8m#)p
zM~%r9uwu<aA2B0c20b+a?)al-AEJ`lkNnz+#4L#|!D~O-p#kO@=<5>UZqNb-4jZGr
zARG>k3{-I;ol1Fg<)bipB1Dp%E@7wfqHv-cH-L>SGA2eE>Wx*CeK(;42)RvoQ&bo(
z^&~E^2SQz?i|}@S)Xk?687Q^w5huMykPK`OH>ZR4iISbSrdH2SnlEWZN?f;vTVCNw
zdiI2Rw*Tc&?o(Zh1qCv&JDGE5j;rwM6(oS%ezl#vHG$5j(1DB&>)e$=I7b@b@%iRf
zRrdZ6c9Cm9U0t0M48|piHCp~~JO-DKfy5yXgg4um;6*_BfbG*qSDE1L9%LS#u!@Tj
z6bn+5Bo2>9&w)8XKNi9sQ{K=J0IJ>O=lgwRvmpSzfxC}lnddgwk*FZ)M8{p!n+;tb
z$M+_@UG3DwgpN#aV(=5pdlwE?my!h|*JwyNPF|Y^ql4mnbH2$6iv+wMpsVdMPYI)n
z#tX!e0p4d~DzZ?EqY%w>tSdJ1P_`pNe8G@(EtuDLc1!rmqi7v>Wd#Kf!YCrndQoqY
zy=brz>=Cj)OsX1*d%!R3+01syQS=~}{7)XmD|Jedxa3t-Y~`~z2|Cn2!2$U<k&#F8
zTWu3qRlOQ;R}BELnJVpts!K1AwZ1bu^GP`Rx^uerL9fUIj!ldC!SnXGlS7d+OfA~^
zxC^ggxhDd!hSjQ>+1#WEZ&+}L{v!qK>H5ZQZe+5h@%>fj#pb2C?lxt-4a+Mx<6c|1
zznJ%N9qvmsv<D4ANOnXG)J8AnilrHW6u|{<wxBGe+n}pRb(d(`qN&zpV@M4z&>3*|
zS7&-xSDyuT3WDlk(2(-DX}tO=@{`j{p=%J^%$WGn=#A3p&l^iSF>g1_Qd$N0Y<^99
zI7!FM%w;dtqc|RoH^>%<*U#)W)sKh~8Tl%!0LMs-`jm!Etx_1vx>FbYPiDS2D5dom
zZ4;}@nZ7=5u95S@lc+d;-rir9UNlF@nnp#<dfV-<yAvM%olD(jnrRw^rUWOWq(ru|
zUmNR6tj%5;mQ1Uw`w{uhM4NME;U^|^4MYQ^(ny0_00vG<<Z^Bz)(R#U@jTnCvA`1o
z0@=NbV5cw+JzriuNhmuc90k@y0Kvi;bK=AaUAW`F4XkJ~>eM0c%`vS@m@kN%m((wM
z`2r+N;7H6L=<uJD03~G;IfU#!y13Jtl<XHDYjz%lao|RtJ^uL~5eDFFdT|m^U98I|
zF)2wO+z8Qz@N%DE*#7lkm;AQWoMfu5o{=#=P%i)sQ3{Jaug$NYWW(s>J{uV3JOhtx
zFy6z$1=F6j7I>+)$pfZHpWMF3uAasA@g}$yXaGJ-)A;-Ga))(sw!|K1cSgs8m*!KV
z>i|?<lE_DOO~uQL6qJ;z4wVM($@)`AK|ulj@^}n|02&x@iclA<;3uUcMHB)5%%*3r
z&$;k_f9n4n!5;qM#<`16b}D5!mj*S~W-Sa0M8eS8d4P51=^p2ixcm22P<RV^e&jB$
z>r)^2<T!2`6LW7*V1c_&vCmO;8tc+@md<3fW_<WQ<LVO~qC@uDu_@Ok{aJ`SG(VT!
zj7q-yS*r6ypOR{(gA(2-<E?HsQ|X^2m{q}&78CPes&=7MP_}~~*1fRln*qmly6VNY
zpkN2Q!~SzZn^P(+(CCRR*8u}^F%c@K1WueFQf~81i7kx{aJv`g%i||-+)9{;AH;Yx
z<GvG!CN`z*k%oPzm#>(!L)uN@bf$~qLm6Yd5H_AbmSvsJS%=qpYhc!kp=<I1s%8MU
z=l+Uwm_FMb{n2SA3dK*>V^7xG`WKH=swyi9&486RzOnHeVG#i+h*9Z=QqMyZi5nth
zAXrw^SS*1@UteF_y$-^XLpj^D=hQLSrHm?uRb&X$;qLrPXXa<Fm^V<UJ~US7{`Mo+
zoN*w<0c+QMWA4!b?S7ONLJyLb2z?-lAHJl(3pnCL$`>|qm;2JRJ83$|2FV{Q71!1n
zXhJ(S<UHg!1E}i*w<lx!0wmDqNahBFD~OGiK@B5se<bLUz-@HsMAjhDQkQT91nf7)
zHef~6OOX(5Zo(swPNX!ipGRdc@9;Ag&0vI#J{B`~brS<~H9K{7f*H<Me}af`+->`i
zcG~#xmdAi=neA5K0lAj?T^DYG26HXA&(J^STj~^>k~S4H-y<QOmtg1v&_4AB^thni
zupAJCC!WZ0EbqV;y}E?LrtvRe^oC0P_!$5eRLs(P+joiBUFYDSg@+4_HV!SSpES7=
z47Xc)YHzrE@NiAHwRXjQhB%CAiq)dpbs8+6#o#HdxFtyCF3=s)V1ZmF<AXb2NU-Fi
zB-s>oGj6!o=JI~XItwq^&;x=aTmTYCedC&9pcA!CCWHkfyC(xSe0y-_W2ieUwo6q9
znD~T8t7j{I06J`9KFI~au@Lo~j-Ub802PbSK~>;zo_VSGdX%&hoQC9cnV%MDBpeYD
z*U_*E;w^IMSV^OU0u_!^dZ^s8)a+`sJe-RQ#=t$m+~8!obVu60eJd5j;ZL9(o1&!c
z25RTdeGzso7YskQwh1pCVW+|@{t~T!xZNCpKaCADb8`lyS#Q){cmV&T|9=Er1X%D`
zzwHp6X?Tf-#b>ZFx;!-Zp?1tj-j>_jW^Nt3`nkiYWz|yZ#5n?(?MEx$X=QT2qEVD~
z)A&)n5L-T+zwv*6;-Rqt0J>USdJ-G6H$c|8Z_8L!+@c<wni${7i16^{^`*{2agmX|
zPuU*au~4@k&Cea3?eBMY)CLtZI~8Xb-_4;S;i6#ElC#LA%9{-pCc<cGammS%yznF5
zamqjL&rJjQA?Mavw&b*mCreAKL$J?vkvhn(-<-9iL9K3RcnOW)+hVs2-Yc5^IC__W
zY2A;(*UExFlT585B~C>^-|ZyEO1ziH4+j!-ar_C=x2kfSRv?Ts`bZYv7+(-`ApniM
zx;nu;2B20`dtN{0HET%unD$)7GG*_jfK$gwLutNHZB7^foNXy3B?4Z@wx}{(H)EiB
zvqV2~5YFOx?mKs%uv~&LSy(|+q&pvv{p~ii#ej-lx<J3R0f)y%f4IAxAhc8kU=m&`
z%K^`_6S0C6``33l>Y&%ff!kNG#OR~9fG?K{#8?5Sl`;Hzqdks7c<ia2jBUU9BIKL|
zQ2Wvm(ImtG90TFu;jYm3J&~QljcdNRs(N$T9|4^ay!F6Eyik^xQ#nIpQ-;Lr&b#Y9
zN`7jnXw3|^H;LK&N&&Dtrk=ECj_GwstJjiicz*8J#sg`#aRpCy`E=uyN)gUZ!S>T>
zDn5Q=&%<q|Gs=%~T2KGmBOiCi=2p`KGNxmb&yeuXp~=*V%%#-jfR|dWRq*7WT>lv@
z<)CL8y@4Ju2h?W=`o$aMJ>R#BSNAWm@sKTE4xaI`IBU04w;4(Tqq(@{8NA0JNe@M}
z7JG;AVZdiZ?FO~*7uJbTTpu4dB=ErMpLerH(+AuXHxwFzzeju+2<L~zX|jyI2e7|m
zN|!~@QWu@=<7^W+Ii}PXEw5rqzvZMGl*VqZy@2dxSKvPI{FVR}{*3^<S()wWQ)XUE
zUCj}3k+_s=HegxcZ#@YmR2r-s{`Tm2V%B1j<}d8kcJtOUXjg3Z6-ry1?b?S~tCQ0y
z>Y9x`B6Ghy3{9LepN2abNN{UR$vA%df7<)5u&B>>+c8)Y3!K;pO8Fa6EC?z}Z)yw<
z2&jmNG?ivUdKo}KjS@>hhbGc&Ku{3{0qLS5RX_nzy7VGldOPc#(P!_geRrOHb8=;v
znP2(JTh?0d`w{3D{2aCuyD9aBa~pheSyk0-j8Zg2;yKIG;T}3~RL`g62NI>+uU$YT
zJ|1sCW&t3O)B;z}CnNcRUD=EMbD3A@Hf$_0i;*7d{evmdM3{pGo^y={mSkAdsw=3w
zCPfMxw-~85;)4L#2y+C=u%R}{tS6(ErDN$Q-<zE6@ad&n_8e16-OVZ<E)$T8mx#;C
zlwcmvbKb#z&41kBSz`;uf9EC+S?04}PxL0WB=KZ@jrF`GIK$T&uH-GF828Y%%lf%m
zW8Y`P?OTq%{8GP8@N?OVO@?mW-h<&I>Af#T9DCf2Qyu!9O2-Bxpbg#w<0p0U!}SBc
zJfZhcCD|6B7r2%=mrq#*1=p_LYN{u@KbNt?6!>~AKRwdRa_AZ!t`uyjd1|g<e<)6Q
zvgZ-@nPq$F%}AlYm&98cnA#QwcuZY(&imvT^)7hO0lUAg`_r@40Hh01XBYrLtpBf(
z(Ab9$3&@uud7;Q2f~WOqx{v4a2^XP`X8^^e=CHD4B8)vVtd_gM<u6S>Asa703Z$Y0
ztGnLr*^0Js%t14fr$VT<27$<g4}V;U$XmjAbAPPhSj!t=I@gjHt=J|WWIX<)%(Z8@
zJ&$*G^u5<}>#C|=PxMLX4;*-sIbH=`NRjy=Yvzlrocm~=Xg=bK4}hJJ?X1F5cdx$X
znj~~3S#4);zhDa6Vx&hb*=Xn8Qr7}|w71>HM1DoPJQ7l$X!j?pK?=B4waJ1XI2fhI
zr*ibJq0p4lRQh1*=)_(DSpu!I0@B<@9v;*klhW|{Y+pGNKk&@C_v67s!qd*GhzWG`
zKgYaT5)e@wwn=gQB?Qiv;jFcQP-kw6MYs<L>HREBNHfJqMG-10&UGwK@{a7f1vF_M
zF*aCUVdwL;OHj#?fasZUZ2~1IU|V!2RwzZa7BTI&s0l)C@!e&T@lIyu=0NQafLsP(
zPDm8|-=QO!7vWNc6RKP>d6XXGzP>4W=wnt(<I}j|8y5cZWGUHdlO{YTsP9(d{Z{)b
zvQg9oRUSZ36_C#!*!4QNG@{5*=ejSwA%r@?CZHQ18#l0kDvm|!9*ko0IXv_Ts3Ozk
zZjf<7GsB~6grxhN&@dNP2BeY>c&dOk8p{!n?e8e>=<|H@t=k|B@s&AnypGD=zPWEL
z?Ph9qw%9a$5!t+GQJmR_2iG7xjLS@wXKr=ryU}Q+He;kR;B?*OW4S}8Tfn!jN=cXw
zrbglLWFvn`^VN+O5d$eO5)CeZG5rXag7$yJR`21VLx(KjlPC&A8)#pa0-Qry0#FVj
z*GLmcAfMZa=k&dEa1yM}nCJnZIxQPb#?%Rr+8CF<RnO3Nw?u0Dz2ge9A@B}tLiOL@
z+s#7#5#>2Dmtybhf>_GCku9Jw{ObCO?HfovMP7QSD+m=gk<i7E5dtto5Df@E$}9+B
zS!tbqlRPQDua7=+lmcNmg+^E;-bX>_8q!Ku7uO@;EJyeZI$1#0=BTYn1ym6h07GDl
z`s9{*@gey_b3jY-&*?6<D5XVun-uy{j(|q6g65DzP?m#eq*==8W)ggEGkP87X?qO#
zB~MPs09_$&pOQaFOyDHy+^7C^CM*eq6!q-c7UI`{vJ3))G{d}qMkoWlO*o<lqP!7&
z9z_Op6H?tNtNz6b{=6$){9f~mg(!@)v4%fK<3iy`-JZFapfLav!3ao{BCYE|@!Z-X
zmNnCJGiwjlmIiiFwME~I2^@eZ;Ga!-n4O8J!QSd*8Q>=e*(Tu`zw^v3k4hQ)2RUnG
zl2d7jaI(#p3(4+v&K{b}#u81^R)GwwhfF}x9}c@^a!@V3WsLAwBxL3T%`w)4rx4Tr
zhc*TJ8v8+Gs{o)Zw?+@v9m!xDkT~Zk?Ff%oZiEk#LjbzU0?;!6TC?$1LmG;DAz0Wz
z+m{MSG$c|)N%o7nWGUpr(zV6Mh(w@d_OZSnq<;i40;b+Aeb#;A(qUa)T^eGvKk*+d
z=?wyW5}BH6a_UsE%H1}h%|~Av)ad~cKtuz=!vX;;EeT`KsiDUvS<1*MtkX^n{!Oqu
zpc5n-Bti`7>BV!?oI%?{%qFp1%$?)G4v4tv7$z2a{}SqLQTZgPR?8ky0SiGDhbxm4
ztPsPKKnfy`htGpq*E&5BMmrnzQDflW1LrKLtNDjQYk+ml8NKJM!4WxQ;>tW_oM25z
z=W~69<8gyZp_IVoxt*owlb@V*Xf2*%0XPg171-W+!L&z2rsp02B-s*0DM{v5v%FG2
z4<A@BOG;;7*zB-)m-<k%zO_8TGH&l}2ixVFT({JZzEx>y4e?aqnfEd3-tV8&Svt@O
zdurUZS;ud<&&2SO`<onoZGzHo06i6xsh_MSw2n~a@^60<clBVOtizGGlg)XlSk>l$
zc*Ea?IZxY5q@A&d(w+$bfC`Q&>;u9FEMN<GY=ngKq=0yHm?7(vh-;A<XZpEe<~5If
zh1RKPkxb1EXd_c!gnh}EjjwX1DEG_dm58j*r(<IwI2Rj4V_LJR@tO3uMF6M<AYmcM
zU&BuJl#u=<bZ#kK_}k|HV0UkN!tSmavpGVkNWWD(n#~swY=)q%vHCrjPm;(vyK9X9
z3j0?49j7$}iwO&+P3=$HOS89ak5%r7%T_pdj%+jnZ}JS9BiM>*zgeUAOeloJ{K<}=
zXq%#g&E#n8A)#KhtN2YUYkDXSWJw*XGOaEEaGS|Hj7FiWNgw9DmGA*RkG_k^H9<*8
z>T_<ezMbkq;r9a|j2J0R>|Tl|@mz%25H&+21z|P-==IS-V@tIIGTx^zF69y-2nF*-
zVo@pGM#}UcGgDbt*E%Dz4Ne5$zyN39&`yM4ec+@r=jyNq2f%c0{tmTk&n$3S@us$^
ztMvRxo^ra{q}^;K*NT%R;Tc-4BGsb7ieqdrSxZ=rSJW{0z7(L7>Icbe*oUhPGjSOT
zz6ojKUvp-Hx38-HOLfmyW?rVG>SizTBQO4wsru%g>YNQhk&Y}Q)r^dc?%0t|tBFU+
zLzzc*$7($KclYV9bI$BOv40=W`Jd0LE@v-XwQ6>}_t$DsskRZd+huQp{}Pu%*O|KB
ztjxjC-m&nKmu9V@C9LAX$*9w)G=#y-3$cDez+~j|<dI&L|A2k8zt-PTRsuG*qzJ*v
zK#t^e%PI<VWFfNNZ#&%E@1d<{G&GHMvRP%SL&Xve*u69AUY@^b_ilUEIq`bxgei6f
z#(~Q+yZ2=;KV5`wT5E$$rh95!yVR<me?qd850pEIXg{Lr29UO!!y2(L(45r!zyay)
z+gr%-J&H@RigA64^x!(kbYcZy$Lv4qTM}v#7WM5b4Aiq`Wj2BYtd8YDUb35YSYUI(
z`N}rW^dd|f1dZH=KH|j?t=i!LgSKR__l2i)k_n(DaRH3t`{P%x0T)ki66(z(D8;^5
zP7IRk^_c=T>AaKTpsTKZXSwY2#78J~0VGE#g6-q)?hY*@NDx3e4+6(AbVSN6ogdr#
z-l6JbbPEx8c%LvlZp6N`RVG0aE~h!L(Mu3_YjKdl$^L84Ad~=U@;S}f8Y2C!7sB^F
zXQb?Mg+xkpw5KSo3q&j*xo;2eU~khcT|K=v9|h%bV6mtS&noVUi`#RzG<4OjVj|!u
zr9XJ^`Q#-NBO|}Gp7D!VgOXC#bu~ooBHSrLx)(z83`D7VZoSU#Vx^JEz2WYMj6GO8
z0>$<HArk;}5eB0r0r<8+-%_)XI7=6A($Y-<&7ee<l3R`uj*8l)V(tpHhWYrn^{8Mp
zK|EIF_|X`f(bl4cDq_INP~t-13UGxB<hI~EtOgDfnT(kMmsc7J!c#!m-Q`bT;V>Y!
ziTQI|BM;%21*-k4s4#~V*}I0M2ncOj$J!z=DoOz)KpdR`@K}+8=2L?xQQ64h+|U;Z
zTRW-nXcBbAQIn7>=m4LxettOfIC`(7pti&wk&&?N@Ynu-kuN~#03`fVjm_&tL_>j^
z*$RpcDP95vP)CaH7<0j2&x4h{fdA^y%BziHVuiYXAFaW;Jl1jdY+gJRZa-*I?7j_L
zR!7~GzoAXGd87}IU^E~!i|h+>Otw^sV8G4Q+LL9%rQx`nKXCP|e`6o5ydkWl<sQ4e
zpgz^{(cnPgL}e=^5Dth|^cw#VM5=0{ZQAriG~a91{b~nRieC<iCrIqUWK$(65*${D
zTVDtv9>9&;2s<z+`a2fV5~5B>uY&be0BQ|3&50_1f(!@h7O$zZ#!4k|75$Kq6NKUf
zby92!7qI5DNhWfmW`7+mCPHee2q2^zK*lL8h3@CXoTEdD{FnIghBYgY$J&;k?sJkx
z6!{JfO9m%Lb5+hhv#^F*Q5gayj+iLgAlexr=~#lwL;@DT(p(g^+oShPD`4n7h+}o)
z&me-(e|mHWEiT0T7BJrwzF$?jKaV6;z%nU>X|y;A@v%;4V2p&)MJn`M7e>pA!XUKK
zk%CGbNXLsDIqa5_?UjapXiOjn0m1qm1<94j;l4vX;*S0j_W=r}K&fB{otHV5CkIi5
zIBZ5(9UYAF$@jG6s!ZEX=gy3Z1S_)au9U4M3JIE`dK3orD42y+QG7*IE{H<?R}K>u
zz;m#|f~C7vhE-2tEfGNpi}4i#V8m7(kbxrNuIwnV_lcmM%M96LAk-kuDxi{{grE5(
zjcYo;W{U~{_XHi8N~9CPh4Z2#r6@4tD6F7K85;=73PoHc#$2&tg<Z8(a?5LxX&aKf
ziB!&YvpNPc1i}+#Kd}1CashS@l8TAqYYleC3CO#mAP68NJLZ0=5Bg(}T!}UKPG%78
zbRZ^^3s!JGTZeWgtHV_QQc>Vdg~C#{jc=Tc41xxvfSiF;GAX7#DCz!X0X(ZNtO6Xf
zh#73-7+5XOw#NxR)_jpo``0ZU@Qaufg_g?GF>kIdyv#0=hp?9w_l0opz26|i)e%S1
z+?+(gTOq_azAinWly4^#%WnkpeYRL-W)coyXH4O*$YMaEN#r9Y7<^~#A%!Y;PvGek
z_BTTST-1vixK`pm5q4B>&xOu}Bhh1_!v3nMWvK0+*gPdI(3hUB%)M+HN#bCJ4|Hc{
zh$vHLrY9)dk4l`XdIXU;klek&Ti1{j3PH{R6O*{omKRhzg}lv~L>uls_7R1Eq7L8{
z5qwI5;w^}n?uf9cc`^LC*&Yfn&Lp6R&azyOu4IpmGlQPfC9r0Rg{H5rpfvz)?u}lA
z>72m?)?iTM)EaCFb%KX82~tqcRiXCrq&-2CI5#6A?p*h2dPuNxoKgweIi&Zg@tEj_
zS3T*MTLtftITUN%Qe8joWV<?%=-y79hyrirZ>y{im7ZKrx>&E}6ckB%^j{-A1fM`J
zw#t)~*4ycz6$;VE+~4?l8h%K+Ibtjaay=*2KELRX1!vkq<I53%_)fR`63e#4^*ZI}
zB{P#HgE;(QX)aObS)<po^3bJ7`<sE66L*k#`*Czo+*o1OWFg4=SXn1{`!+&u=PCEi
zl5$*sg@6VUNQQ%rh=4B4o-K;*D}^K!Y$M627#Lsj5xEZUu!GuG){p)<4b|7!jYu_6
znV0ElsfsGZgRfMWv#eA(dMS=H4}C)($l}A~8mDS2h#Cz(AVORmb@74Q$l0-md2#vh
zaK*^U!U2~Di;h7q87>y?@;g*TERqX^qCuBhs8J@4*O|<MkO~UB0ZH&D;y#s?m2FZH
zW7q#FU$+sRZIHQHXO3BLq_yz(n|#6`rGV0E%n#Clpv?l$x4C}%+a0GFl&DNM@ePLc
zx{HCmBT1{U+t+8LqLOR6E@cPd02Sm@GXTJmBwXS!^iX$ES<kK~dLW<3tU1@%dJA;=
z3!#Y|$h2mXg&`pjL)xwtpY=5rTgMLe^N4AX)}y2(X;O+P_nqf!R;VA-5=LnD6>J(a
z9<2t*n^?R31GjDHSfM+XQ9q$%ZMH9oGm18g-6$O+y(sbkb#|!jTLF%S4s1$*tbkw`
zsP>k!h-n8BZIiN_)>MMcmsh(!mQQISddPFD8w>|Zm`JZCzj@PhCGO)sAPqmM6JW-)
z*=F$XV;_Pt{Q=Vi;j(E6>YGVVgjQv@5UUY6%1Kj^Iw0Buwn`8)<&x-ev9I$*DIJvr
zGH5Y+M=_Ph^0k*2))Ik+gh3PP4E4~3T88jO5`fnYao<RvpSrcOsegc2FtURd7$`FW
zG|W6xCIaHz*B~xgXS<c*qiY_+=@upw4Tu3I)bkeBo*0wXIPISTBd@KT<?ZFiQ9~0V
z#w7UK9g1tHz|X0N&i88i{*5;z*+HHv%7o|RNOIADVU5<Dmt0Hml6e$5C60S+tfcfE
zX}<BRKN!#FaK~#l#BOm*D{I5>&OI-hd=`uSbY~N~&Ol>(%>a-JX`e-za5B_vA_{Z)
zCWdih0cTORzk+d|h6q<4&W33RLvM-FI^0^-0Ift&qs%`Hv)6iwAH;sBQ`N~7gytHd
zKYJxTAp7Vg@h~Mt&ID~uubFs$H5DNgMEw>%jpBaVnJ*Tx7qlDO4D0|4i`%3TPraS<
z)DAmvM>Lr+>F<Iz^!3|Q45J$yzn6Lkr7!D&9zse0=9CCcy5lg;o}xRp&r=!Q(LWj$
za#r2Z(eW25FyhQ_X%6+CfJjl&Pn2Ru$XE=jR2k1#a__c<8XOVfBFfQEBjO2^{gu&;
z_DGm%qdnMXB&y}lnlANLKKN=I%AUBtTSe*@*oPp|V0p(=<jd(5#J+!sOh@}#Zm;Le
zgvPDeUOwL+@i@3Mjy6jadY^YV(94*9^ZkOqqOIjYj`E$HE=5l?T^Stnd*|9Hpo>F_
z9=&e$2CqaJ4okWBnMFkYWBmlSXmqHZI-z;=o7LPYUf=?IS@!AEEm1Pb&}1b4_YY;C
zrW|3Dn_jc`Csnlr-0CT?RVKEq%eQZ@D3oG^{=pf)eXS^(4ng}mF(#H*I7%b@F+f7x
z0W$A-`L$M*6D81la;JA|bi*837Wcug<9w5~;7PP0|4>%stlrr_ZYMw@$No{6G|ON7
zP{zYqc<W5(@uwqo-0|Q3q=NvN04ugAxVvC!t0PxXIq{zRbWu1ECjdnyYnUYc3biX(
zzeWpp(BZz2{AM>9dTe71n*UYKGAbwsae6g(zOk2|ANLRYz|50n5oHp^hI%0yr={G>
zUq)^N!ohj%=d9_oz^8EM9}l_<kmn}h2$)YA!u8zuT_|)od=;};t}S%&+F5t}xfij(
zULrY%C=*ve>AfMQ{m0XA1ZY4u#F!+>gAb*fAN>A1WB2}?#(tz;M)dr*xsO`}+1%A!
z%P_}7G(@Mkqb7w>BjVy1DxWktYuDTi5-XAra5fo5ZuIQx+^?_3Hd-eflCi*xdrfp8
z^Sbzb#sKAKxjb%|qsqO^J+91z$;PumAd&ipYTr1^&;JUNJ`3E1$tENOS=WF6j3)u*
zxPWMBa4gY|343&_txdK9&{`Ur1WqiY8teiI<h5LjzMpPzSFJc@#GKlHk_3KVFqYg6
zwDl<_8sP0(qUCO(LYKx$ihCh}!_QCb?{~ood7K-fu@&9W1~&u_OgEI`+ARTDmy7iH
z^SHP3qa_cG!EMnFqH5t7FgFf#`Wb`K-ppHf)C%Qv(uaBO?BR??SPf{iYq^1zjE18>
zz9Zp+BVtTwS4w~X6J}4)hcmG9&$v<}Q6^waKR|>5*gIiYzJG=aTQe*<CL7?`=c9|f
zc5$C+iGmC3GX4~>`4>jm&F|kui>IZ&s1})A%ia3NaW5Ynl!NCCJ9u~By5;FekCAe}
z)>!$omlh7jqYtQG7^yf2lQun|-(jxE@E!<){b$I>0LX28)*`?-@y~^0{B(|=kTUa+
z9S14HkYC@``%|A2XAzb%u`FB^xNr7=SN#1x6uu(a4nT?7kB7Sp+XD4R1(OYRN+j|f
z+Jcr`PoUKf8fd++FEp8dF&XVe+>8FA@5%3VpSvow+Fy>RVMx4Qb2qJ+111ew8)wqK
z|A{KdNeKyR%yR}bNgP&M`(y64@+s0Gz3%a035FWi1ZlN4Ec!@=r8F&0+=n;gHTHBy
z%)nV6Fi}_^_xC~Wu84AcwKV0A*-|bTUqSBV_`E`?)zw_m-lQpMVSI}llv@0Ccv*&w
zQBBVEWamS}B`KT*WT1OkmK5X29Q?l4AB~OwpyABmEis`(3T5XD>GlbyE4~H_Sk-OI
zXkWp3Xq%<Tu!x6x&@|upzXsc8DRe@>Y)d&cmBgtyu|j#ede>rq^bK===`Vjg_%e7A
zF^UPKpjvWbEo7wfa#n?4Bp8CFzSZ{|YSUTv9PcE+WxA^;w=mj|e7{1K+uXPz=UK9c
zxG$m(+W`}8J=i5;*aze-^IaUDY9)1y#t30D(y;5Kv!>4vIWNIo)VOO)Ee0&Jz_t_n
z*whXJk?ol>`uUN2vwqC0sm;;*iCF$r>h|tQ8YFJ%D1TzI!O=bdrDx@2#YmOlKK4a0
zKg)ehAe5C(>&fow)w;F$?z<-oR4ci&X^L%J9oW@(1dFDLyJ$=|1E+7aH7IV-MW+Ns
z?u#wteViTg+$wtl!9Tz6ruU4O?>Ua`<rX9sY8Y`pQId=TudshriQLH&x|Exu1VJP2
z(i*+Z2o>aB=sK@>St()yP&yHPLr~CA0Qu&t5yOXb>=DD~JuY}H7jUxDmo6d9E}Fq=
zjk+>Izdy+Js)PmvU5&q(@Ok-pb9S)L`|LT^(B}MNVvV>o$5dk*!$$o3-AMce>nfTL
zUhaUQ%7_c8t|%*Do1q$wG|?euvm2VZM$i;xw?;K5tcbC~qOvvVYWSRngj4y;8Ahlf
z_lkl9jwl*~$FnsdEq#tD#yMA=V-#|aq2p;z(~|WrZSO+#6vH6~{>A*KhGy=3H({vL
z09`kw{)y1Xc5d%_L>76iXkW4I$EAIQLJA6|Bp>-EHyD8wrx;i?c-zSX1mTH`FvM^m
z9~Qo#HyBA0MKamwtw#J>uKPb!yX>L}KP_E2_htSr?(@)ziaAZl2!#iyOnx}L+O@$9
z*%XdFY>-XV%{JJ-|JYo-12^o*?P$0#^3vj@+G)b*fdsJ4DER?Y5@NygW1yOlIki)1
zACu}$K(2`loil-(Z|nz6U)}ndAuIR&`hZ6HE&I=czQi_A>6}o8Oed!aIaDG`nxVzZ
z-6gZ;S^GHkMV(Yb>P}dn=tN~bt-W;axctN`;lgL^^komNDVW?f9Dfd<=#)QKwUTRY
zme>G?j*PdPDeM9asf?lpt6suS_<c)Ii0oAOUtcT*r%k$0ZZ+}YXN^|91XDHk@8&lu
zKl_={KF+=C1<Lsw`?C}54HB@y8V~Rizk(hj)i`t7D1@`}v`Ue^6?4Bt7a?GYX@}k#
z6!9COIZh(S@i)xIS*FxPnHldXR0jY`Xt)L$EjpP1rcHE{W~54UCN}gAC4HDDPSJD=
zBKb2z>EPH2#bL5B=^|SP`58XkRbTZoJHZz1gJu#d?5Spf<GcwBjfV9QVlVP~qQO}N
zRt?~z#3jl$V_-oHX6DGDGyc*3QJ~<RPtPzMj5u>}f(I3^1S|Q8A|mwmPhoP@8Cc+)
z&PGSVXL(6;&4V-n_x!$=MYs0XoQXw^AQhkVCxMyX5)fd_=CNs<+)l8DG!;vca-T`;
zKPUBLE8^sCREzMlMbRm_5Lrbw+kEz&(=if`Sk{bobxz47Pz6gOHNAyh*HJw#66wR~
zQ)qCMl(K?h^>!h}4<~kk637P#(cf4Js8j=CCYz+N+7Nnl9@x!uX8||tUYXVs7y<?F
zpf|puBbbKzgF&jU0`0SA?_5<q(<{_T^Mon>AIL*}(8eb&&M5z~dg~KLa-Tr?<v&uS
zDkj?RfQ#TMi>tqJKyB8YYw>?@l2O%Q&aD^>|Mi?nWc>d>aG<pRweg7QJTQ;pb7(jF
z>eBg<#j`^HhJ%@!3HNKYn7Pw2N%}H<eaD7JY>T^tvZK!W*?xZUw(VJ#@~_U%)CB~z
z9uo4M$}*npKKbr^MC{e8e!cbIvQ_w}=Vp7%H(s{{MIq`sWtrs_L+=#oS^6<ejNA^W
z9Sh<-hRoQ__pWV#*!i!;Vi#6-y}vJ^<XRpeKR2uhPUvb@A;|FCkc4Nxxgxt_YJAap
z$G`m4b#*ln7#nr|W^Oh7)e_h=IF<5Nwnm*nr)=1ty1GNmsRSJ2tf8OS`HyJ+CI0E?
z#nI=Bv#bB6{eo7CoE^^C7sWAI|KlGvAW^zozRD|k<n2l*c426wPV75(rHqjaMN=5u
z-7YrN*2*!Dx(pDe=bmjV7Gib9{==M<;HfI<tgt2#0d@%qf@7LWnZ%U2jUZisvW<Op
zzCHhrACHp$j)LI->6sp;me2C~2(_3JAjNkdA{Vh>Q6LLS`u6qnMrzvzClZ9+7a;$;
zjX&UI2tT{@<C|N=O{O_jZg}>ED-J@zPz)s45|W6h@-4G`PJDUALT#RiP=9Fk*@>xz
z0(KVj*8G%u53{dltRKp6e9KR!-Y^*GDijGii0)qOjqunpKBN#^{_9EpOQf(tT=3_L
z70b=v-@8OG7`jmTg`7<t`uE9G)7)W{Ig+@q2%+fK6QvE1vnlCBg@zTV53cN}`Z2n@
zAOT!$GWse)q(AhTsTg5q{oeBnUhd+XS_eC&^tBuJFHuu*bsis9O61X*C5R$b;{qPO
z$-ckaP<@oUEd8bJ=~Gof)}9k9Y0_VZqakRe#rU@c@|%|l-MxFvcDi3AtNOTm`M=Kq
zomSTkPb2Pf&BQ6W5B8lnYg-Ne&c518L{syhUzRx-S@i7`I|$kiJ%}wk%TLC{iG%`%
ziHeWkid3R-W1M&~4`0!qU9ay=cIx@Ieb9(o=V;hqMXOs^&nN_;WAu|J)(D)@gCrEx
zS(}&&YqND`%GQzEk$>C={AQG}BqDtl*!-bqGIbu5ZCjB1+|oHy13|u)6%S<y7^!h-
z>TNBO@{R=8MUYCI=ad78x!-*i2TD-HcScJo(Ilqn!)1I7M#CR{K2}+V%PGQ{LJB(~
zQgP0?yDE61X99_SD_o;>3-|_2D=SUv!YUv&je63oU~l2uH&>p05gC@XV4aZ8%$+6j
zE&jqO)8iqaM*jNNQ!5q_pv&BVBU1dXEQxLgH&ZK`@25SD_Hdi(@6MJFDwJ^3j?&Kh
zPjr}*D_+*`APE?VnT<`JHip4KZATzvf_KaJ&#DWQx#u-J0DN1NiM9pSMSuT#|A21d
z3`++%9B52EXpw(q4wno+I4!XK)EVH_QG3H&wJ_V;6BBEF#xoIps6*ejR$F=FOoxh5
zFa@W}<c7*UP~+!|iP?aPw4%UGA=SMry%y{A?G!-2V+ZzBmk!EY65gEjTCyLnQKsD+
zA&OE0J?vkGzxpb;b|$;LlJD%$b=#)3qE{+a1s{;0a4VCl*^^!*qHZ!rJ52m=-vRHn
z7kwTyT-xf;_u=!$UoQ-NEPoy+Yo82uTRn6x<IceiTP=!4ssGO-1EnTtL`9kCyYcy0
zOnVIpxDW!ZW!iptW4%Jf<2V~yw%tO33*>Kace%ST)^N;~J>~h*r{(@e)i&@<DC~*?
z1g%ZV$!ij>&fq1DI}bCgyTBK}Bbprb2@n{@LCpVqTj-z<P^{mtw5;@Mat04+4k;kC
zNFpr&9W~EQHK37+Im%^FAX%xOIaz^s%}0lQ4WR(7<KQotY~^#EylJz>&JW$zi}Of$
zZMOq3)y`ljnyqc;b)Alz+DDe|I8@DJtuk3qDD12GGV-l#q7Ah%R4k+s%G=N`IQbK+
zFz{WljP+R(Bz<|U){671Cq@$A&hn`}89h{jLakhIS84;Gw!3ds<8Z)mXsB<)j(BX?
z;Qm|i6o}-TqL3#(e;r;HTRqL(Pr0+b{ot`{N7q?~pW7j@e(CMO*CKTh!2F;#lPtQi
z$}&)XvCVZ@Io>hoFEzmd?}od$yL6U3&XP)PyM6izNT8+<NP;0X1*D;L^2A|()JSBf
z4_C!2_Pyk@Z~Dh+{XCy~j3yYbhs&Els}?`hwZ?Z7KHOI;SuA>A>J%(WbDjo&PyKWE
ztSLVvM-EEH%$>pOmd;0wlK^V!FN2AdgFHR+fZ9(!@1d9<1S+bWKYec`Hd(qAx}<z~
zI2R}qJd}r@rSY+TnqL}xT|nwd&S3)f96RM;TG;qXT!Ksgt20&K4)r%ZW#l%}R6}^>
zpPHJP(&Xcdy1{XD3^K=S&?uvSy-ECu_ph*<JkdNZLtQ9H>(0YdGxlvfU41B*`DM9<
zXPNF;{^hOvz_FF?QYm<KX%K}1f4R5s64BayX?xF|P|&9>-^xnzyj9#!!x!BKBVhjJ
z)mWs@92-L|l?=~k3}Xx_X1@gopuGjRvWv-9=oxjnyiiUGHC)UML_R^Us$ie9Ob_D3
zx8QwNM=sKcw!PC{|GXm>GYOp6%v0bDYi{ZWmCvR5LhGE{N<7hk<a*&EnVLo;wf3E`
z5q8sK{A<U2!|u*1cn=%wCko7KqwV55f;e4xZT_^Q3k+^qryNu>Om<Pk%mMaJwcYoZ
zDD)+Ov1~;KmF!x4e!#im<v4HQ93f4O*<qiuTY6BbN1Ypf?Umih@9;wGKBL9#-R%ly
zqiNYo+molNX10v4KDV#`W%O@+XAUiZ0$@Xr=`-7=%ZEzPdSXA@Y;yDpXM0BdEFS|m
zI8P+>=GdSKnJl|9&EyTvcNc6}IzL#+?GT%+)*3Ay8+U0dSIFTfOjP-VPAxgy@JZ-!
zn%nEkL#PdQd(?)KoavUw#j>VH7i)aH39~gbd5tX!aJ_JOvb1zhK1e@5%2!XR&ArGQ
zx#@ydyJ||>(QFL}NzBl_M0%y<j$j-iWf1Pt_zbCt#KiDaZY!Ur=0rT|l2Q7VObZwR
zkVS!ig8Hqc4Sh&}Z=z^V6d{V7+3JuCx0STK+^4!=g`i9c@tD$*)ZKgM`cQ|C-sI0y
z$yv}o8fQElT_24yR<wo?n;EuN(PXE083hH3i-nbdQ(*(!MB@dStnrGboVFThB=ViJ
zCPq%4I#qyuc7~xIbO^Rf+>nRDwk|5Mab!=4?2`X*J56NbQ*B56%yplua|Q3iwKZE$
zx6TOmhaxdUXAqU1`mLoWDXO974xVa?pv0umxuzx<S|nceIS$*=kGljqx5(ExfR|0t
zi(_T&|B-S@3&eX76&<lWOJeD5`#VqMUE<5|S!w6q8Zat>y&faIM>2*5EEg7`7mFei
zHLK7S^1hrf_wl}U@cZ%~1cU<T$U2=HFGt<<^y%f29&)IOE~rX;M5`CZ5t`^y+Tu;Q
z+UmLi`s1o?6)?RZ{6x&&-wT3%yl`eZ?Tj^fW7zwfF*+q=U+>$@()t4%b09{0YDE0-
zW=X^_@5r`5gn~#hghaF0yoC+fyEKISoM7q4yXr3H^2w;#j`rOg`xZSTn_w};acOk$
zkco;cz>1R3mfC|?4yP(g5nYBx0sh;6{DYHJGw$r=WX<%+|I<I{2GO}+rJ2%a4I|J~
zN;2`ctga_Bw=>UaZ`$!Y1MQ{g?x`<!u7V8dlzIBT3)|2Ycnu|a{#OK~dbD@FZrt^j
z<jKf=BmcMp?F$FI_rC&UZn=~#&K%P4pUdzn=AivXH676-C1@fxb^ymo-4Qw%NH@oS
zxzYEIS^^I|v1Fyplal)B?w}*r>m4gSQM4W%weN)1uGf(E_Z<k_WN;gn%A(=r4)CmY
z#d4OYu<ZKSm}%cSDUZInexk9~{(&gY)ZO%!svQuYPW@{qbmqIuv-@Pfa^*@CFNVVS
z90z>qSC`R+vW)Uj`(lM7hB%SB7J4u_2!4#|!=760d@>VYhFTE3C68waYMC~Lr;m4T
z)K7g5iA)B`(AP3aeFVQv9d6J-%Myyp##>JoF!uF`zsIXtVQFUkMYa0a^hn@WQbyrj
zlVsO&70Bi9+KT;$(W~P=pw2$)n&g($w5$#rU)Hu1pB7i_`26I|p=1)^*-pX-0Rs`+
zYTvp9%80J3kpqyW3|tg(m%Q}K^YO3gfUs##)9M`U_X9v76B$2{C$ixT{9dq>=gnZL
zh!aD@=x~8%D-~q$#=pfBYRXKtK|0y@cjatA51a??1{@m+@tWlGb<TWAxGIXEO$xEU
z6*hl<Z;(FP;@-4!MEF3@bkq+IyltC!0c7l9y9#OpD-CWz^YwnUaw{tK1_R<IhmIB2
z);~w^2#aL`O_8Kf#Tb}RI$e-a^%;KA;mz@HUeLrrd0oW2$Kfk<2hrYP{ccYwvc;iL
zDuNZ;BUr65C{jwOs73^gRtY902u^&~Te8Xd!(eurW{NOW<v3`BgBPNE5+Cic{RJe%
z?%*v=)63hGMryvIR(lz9IBt!?xp4(Fi&5|s1^D6&jmR}VInSb7{Sd(#RLe|etwerD
zK`V*;AWf$Wpe%dxhCOeQY7bTh%=j?6W1w^3yRjBY&p`czBY~Ba#yw!_7+eP{t&N7n
zLz~{tTlm|pt|t>=Bu!(+pQ;EMgKotb@sUR!OUhz$7n+9g9oT<NYXPG)k(C4>MX2Gv
zNXAB})iqIYOYpnL<O8VoayAuLBhE!r2!o72Pp|yju{`QH9bPc>v0lT-s2)CuLj8e)
zM|^J6YRYzPS=O$Hf>8>L_jNm>)ca0GFX^|92)YVQrqgp=V9M;;!XoluO84yEZ4S4M
zZkMH$cNv7HAzXcYxW8(@rzeBa!pW?=0Ty?3)TU`)>qjb-=sJx;uPTKPTNL2;kWJX<
zeVUF>jzksigG`USRi@Z`0<uptb5desHPEkh9F;B6Iiu)GEmKjkebZZQ*jA9)wZ|vX
zzWDTZOqw21X<km!wvEk)a<%0m)gCngd#ew9eDKffC-ncU{OELgs0IQ3b9rMuy~^WA
zQKOcdf)unGnudx{?o1{^)TG}GQlEc$+Q!3hSoV=~xsTD>T~bDnOv%gU_ji>=w=8Tf
zDEJuUEj#s1>WGj^Rii8BE7phl?6n*VdT`?GS^cHGRf$p{P|b(x1sDy7$^704)749!
zex4R8&3N?iRJYrSj58?uvvN#NNwL`}fTg(9r++bAH;@O_phL<f>+q9n9se%+P@s8Q
zLk?zDp3M85>&tu`)lUnw*L$iyYVz!x=gGzjb+lOlv-pmXymd=rHw`@n?y*?(q)Bu{
zoyzv@n}S#_oiQdt6;s%NVd$uS2m-0nk1#0mLY!;v4MnT&O>ZGm4ybrkkE-7xBa3`*
z$dbv=xAy1Gbl>RzoDflfPA7K#Ia#Pv8(&sArw(i+)?c{pqq;mazd)m9Ad}ke#~r*a
z=P}w(5DL0(38aaq9VbKTC)L?gAxT2(SToJq)7t_9S|-kCERb~917Gi#Tvz3iHwqF+
z3L4(Ukv}oo-3vwpM4<UG!}Nv~6g+%gXS{A(G>&1#sary~9-gBvj0g|{u~Piy=QC8J
ziF@(dx*tM(wA_d(-y*-4R^~|U%%)z~O7mQYEKGVc{bG8m<U}+Mj|=l&vjSe@I)Ljf
zXo4hS*;6BjL?9H=Zsapt?HJR<5L)!E9m`Nl>)eu985jwu0AkVriihhnGe`tWI&suU
zfH}AcMdKfweC!i;UKDufrRqYYe`9!=T5=#!JRVmmhVvLW0P+9cABJXl^jpLeA55&q
S(drBabGOzXPj()?@P7ej7R|E&

literal 34032
zcmeFZcT|&U_bwbn#fHp`1rZSy1Qi7lkWQ>%KtMo1YDVcD={+i=jH05UNmr`WD7}Nj
zDAGGfOHg`mfl!mPZwKD@J>S3Qx4yN`SubnN!U;*9`?>F3u6<p5KTmI|D6sB2ybFWD
zu;Ol9Q^#PoxnVHOcXw=uR}8ka7{agZ_mmW_!Eg9kzFI_pUpwt@=s039Tz=?3rhvg&
z6Ab1U26yf9ZMVpYUMuU~WpBzd+iCKX_#0#L2B)49C4%JH730RA%R6R8CEk!R5U(@k
zIwb#GIrq}y<s$PebKiUHOpCaWm-GfcebE!Od4O5)<+ZiRt51B%CCyX!{C$(bRfG2t
z$wk+2t-(Rp)MZvLwS~P?2v3!7B@8WW+3|FEtwFM^>{2gzEoPb&d;3iA7PQ(+-*2l>
zfBp-zWV{vK3g*G1-`tqcKXxhhZ9~6LoqEWQe!qSDfBuFPJ`bjHx2sou3tViD;V!s4
zI3DJV;B&iP9%sKWHrP`r7Q`DS=WUvjY6UOOH0@!dRF90<6XSDpj|cKC5A#eHdM#gh
z;y3N>T^rdMQmxn#u>bR9x46hf7jc|Ao)Q#)?;bmKo3N<UIh?wVPFP`K)xP5Jdxc&F
z*oq2QB5!)S;{}W6_??DLfpNFb47R6_<A~MOxa3r;$i4%zKEn+$D#^DMO$Os@BPGpZ
zLzx3!5Z+bn8|un=NL`##B04#pJ>d4;@kdXAMuGkJo5c8!_XEq7?)H>=^Z5E|6L~d~
z>JPYa3=dQp6}U{2J1VK8A{siQ4Y{%T?(^#M@>AC{lcFsZAD6D)(Jyk5!>Ma(^33>L
zNmAizXiE(ZL?8dfjRFe`F8h*_I$rwfA?@qy^#V^<Q8BTcEkgKHrJwhrgSd5TixB#K
z>GIBE^c%D7<jl`A+#giF8~y)(-|*jkKz8t7d-$(kAUpVPeE7dHPTp?2CWLYOpq?B=
zr_&>4J%!EcBdZD=hnp#sn1>G^>X&$kiQD#kZEU<{ODT<NYHH%eUB7-kB{fw;PcP!h
zlP5Q$WhBp?I~R7*Mpsu)PsDMkHajQBgub?HVPm6gk}y5{Gem@y?^b;PA3h@Kz`IJD
zZpbi@cAV*>M`8)WJUl$ys|8a%1%VObc0+T6HQa9c1_lNP1@+(5*VmslES9JVIal-J
zv%iM6HVvoB4Bz*l@@C)Tofj@#AYxauu<&QS=9ZQ~Y1b*Qv`X#Ux8qAo<#KXzk~1<?
zV5wied_nJozZ#pHk1%r+Ru?Ca$he(nVLN)3q6O(2u9Bvkb0$@rR39#8b-|`9Th+-2
zt=emGtKpktlvOfq&X2I<$5v97oP3&#i_5~!uA!&EQ3Q4>J2%&~E88MW%Gq*kd|X3K
zZMLy8;Khr<p}Gi;Lx;W&)`YgeJ~|H94^l^4B=7&Y=0l&48D3c!pZ2_IXJ_Z(={Y}4
zpbz7S+jbu`u6V-Q`Qh%>_H?~aShT|Y&n#P>#5eQTR!P^Xp8N(lh&LG-8FTOd;1DtO
z`n`Ppd~k5Eg{7sUhK9z$vv+1nSCtafw{PFx+}#~<)o<@$-zVSE(b1~U{_K@p9UUE^
zrKP27%P|c1GXerx4ujRjGk@*w%(l=tC}g;n+<xQ!9yT^Jd;9Znsj9DEZ@?0W#nQBb
zj0{0mR@TAPDu8PbDzElE<}WW>(Po?>m9C5#dDlIA_UvZF#dAW%Ggl1VzCWIu9IlV5
ziI(-u^rTHq^c0Bjt3;lnh+n)IBI~(m+z>6hyqH%O1<N1!%#N>B5-JH|<-GWYqIau<
zgm0({<aZovO$IBx2q3B9@osj9(dGpG0*8|w83r{oGcyVv8DnE(qT=Fa@HT*O9?0aZ
z+S6*oMEK{28&o|#J^56l*EnK+MaOr4NAb+o_i1U7BPWCEqGhGC%o^%RYt<xP+{)@|
zmiOwSldG#)cdiYb)G2Py*&m<1radV&v{G70YisMk@^DN;jZj%+sG&#I1ew3Fx5(9T
zZE2RexHvTPy+YE1Y7Urq@g3)!_YpAR0;f@RL4(4e3O_dN%$w1cq@=hw9)ETrLAx1o
zJ907q7!9G_{g2Ccd%bOMZ}(bA#n)+<t+dTOC+Cqp1ni1u<vlz+)XxMHdpvhDz<jdY
z=Z9|JzHR*Z*#Y<E!O;0?0h00ezaEowu}tVXC7v`>N~t$G-gj61<Q7eRz~U`zZB>kn
zj1F^hg0F<8q@>&o7fmkyNST?jN6!21-MgOr`#jnt!}aox0`3QSHr-mK>F?ehuL%`u
zs16pGpU7%39jp#cCXpC_4<itW-@ip@=C!7lmMNTrPN9>9&#<cxzjlUxRfJt>SY3p8
zQc;og)vH%=Q*|`&;V9R+dg<wb0QyM41u&xGxoUx*OGjVZ9FN=Q#TSE3P(Pid5>=gH
zQ1sJ4$R*Nd4bkh3m73bxNhKw+yAPiIy0$#u(%v4@nPp}+R2vp1Z1jcyWg1+hz-{)P
zpnm=<<V|qk$eUJHRvdo({q*_EmoKG#z;x}()?LUxtCPI6()N(EIz%z6cH`bcXPqol
z<->;$%Qfl3J@w7s?As>y@w7(bKu(8YjTNaR<ooyU6W>1akjm&Dxz?RT2!wO@lT>5s
zT}$VqyoRMIIPkBF;Co;Sr;(wDhllgc?0J?7;2zw%^P`c1%T!NTl(cK9oo_K-)O~Ki
z;VnMX7$+qyEkdBLHSp$x_o3UHoSd94vU740<yVb9lJ|-7$(O}(yghwR7*1`lyKFvc
zIzBV=f`C@q*C(u;b3Kj?38|?9nws=oDC#k!w;z29m!5lKQ^r^=i}7V)Go2r*b6;QR
za2)vhdhR}Qp#UFh+}el&R?)=7q_D6s)1oEOx+h;)DdeoM-vV=_w5wouo}G|F?{0QM
z8R|E1-m|*d?`5Xm!gb)gJWt4*+}6}IX^OqBlWnf%-ae149`<IUw<xc;isD&@I}ff5
zu44kmw@hQuDM=7);TCG<8?Z4iIK@rL#qRr@rC_<|-)%MQd!73J{nt#BnoR4?%(ba$
zx@~>5Y{WUeTw}OcXMv;H^5inydS{WV-QrYlgdJhEJM4mq;_|LKhziE-X*yst`^F*G
zcIH@Vqdn%5cJb=_`nbBfT4ughM5oBbMvG(^WftR|drMnE6<Paem)+VIyLdPnun}~L
zTsum19>vGAXFCCmXfUh+*9_K$mfHX2g-)4I$#My-*6Z$$;&L<F`bf#;m5#D#5@CIu
z<H(WPT$}E>FPwG*FT}}**e5ge3)D*s#c@ORQF_s3g&qrIwGioIRlbX$00+@^qBASW
zYuh-a&!Zk*kb95~I*+%TAR8?4SQxgHw}DsT;}P(dAuIf<u401vOKxnmO?PerxDq6(
zBYCg8!$i#HdZ&HFDeF_N8k(BH!2;SLq$y_;30P>18Mx#fF;UTh8a&;hbZt<`%-Z^F
zNl8g(p|e%4ZErXn&frvU(V&rPj9gSplIrj(h2cYthwva{`#VGv9WYD?qI_ew<$@u>
zACa&>1{qluejorjz@D&q-83^B&VQ>HFW`ht2NQ8}a*Bs&Mp^nIenCbCqVr-rtW_`1
zmP>rQF&J}^%OrB>xk)>MbX{Ftf5x5|$c<a0cfxXIbhp@jTCZd*@h>ko1PsTl1rnma
ze5o4jW(=s(ZD6~j2rH@tEGsKqc7l7osjRBJ{1UwryNXrUTdm^hgwVB@t1E^~W|njw
zGq&wL47pC};JL9I+8tS8*fQCwzZ=#Esx)+TbjZ~4Ji=<8h=jx&@Wi3XJWBoIWH&P2
z`;{~rZI~asjPscbj0xf{`~B0SxhHp3vFlD!Qc}%bT`SjW!<C7zU%!r&az0b}lB=dE
zPQh_$<~zjtmk~m%#||7gFgh_&3+SMl8?a7*MVo3VIHL8p563w#T7SjTD;W~(LWam%
zGU2!(tqgbNSPdGH42nXy8B3MRvjK$20rG*t@eWc$q4RjA%VZZI2bER}*ddhQ5B7kc
z0lh%;&I9xyJF^(zS?kxCXXjeVSf5V3eVEgAwY)rG&mR5imt5V=?H7gcj1_Sa_c``q
z%d2u>fWnH`=4w!^4ea6Uv}Y*RLV7@1ii6XWW@Q8RQ5nG9Fxy`_uT5Cc9_)_s7)!2=
z5EnuzQ%_H?tE=k*!fk+MzC7muJmFz%*U8PrRR>3bat*v~7=p5-<IqhNuqlls6%_XR
z+t<#6QR$R=xof5CYLXAx%nej=uSiTm#FjVooXja+?6TtE<g9`G)ZE(}H8nLg(Va(<
zp-JLU5a3?xV_E*h=JRTRyhdJi^w&<;4PIPy;o*WUIko3usOkh9SZeX|Pzmcx2m}yY
z@U=mwHAH||c@6;U%L`6PuchDNK)DB5uuH1TL*k6#orc2KMNF!LyaomERobO<!MwDK
z?NauAm%y?!2@E=deE=*7lz!XZ!eB_%!>~teUStR-+1c6g1qG6U{HiE}f`6z(28Cac
zK~D$rDJv%5uKxDzj=X|`2zUU5P&0G$Q;^teAqP4xjJ4^P`K%koXhYBii^_BwZ9c1)
zdltp<7FBFi${7MLb{9akL1ntnY90Y)%+3-|*X8ON{92&|Y%BOdAUA!+4@XQ)Oia7;
z@dWxPK;bO-rkhbx&L>n5M<y2(e#@S+(vFRM)gkMiNjl<67|)}Jp4I;*JVghB3DSSn
z$^s{(@$pPMO2*SWd@&*m3k!oB*))S9mvag)k5s+pQE;G+G_9=-V^#5Z3d@s@j~_qo
z*|X<)ci&{z1wc|~_4CjBA3Ue1pr8QRtEH*wH5husT229q^AQ(qsuPtXh`azH>guDU
zvmh0t#0l2L?Yg!w-Z9^<XE&IhM?OPo@9GLSi+1OWii#QxfMACJK)`}N+(jSz&f<Wk
z^{_m~S_~s5IXRijmBLN`&WPd^w>2KDHk;2Otq1f^71@9HqW6}q62Mtj*2YLaHa*^D
z627#F&X#31#&=4Hg3#JHVO^-Q?!<Z>WzAGdI>w#SW!sY<ElzfmEh%BSAS(JMH`l#0
zdo);Po%(|%5_g)hjAiH$2imj=qu%#v?0`AUCcB2^W)3aIhet$E*u4Wizta#<M@JGR
z9STV@zN`!zDuk~GB#+K{4=by8N{1~#ZY3op*oqK9vJhIuAl}^sPePGBH#fJnz1?SN
z-|U^7?Cj@77uFyapbWR%U;dDp<LJ?&v^Rh-s5F`xB*DRMD5VfcKpPM6APh_J0rVG&
zwyZknv})}Ub9C~Qt>jAc(dGl4nS*9f3s`q`vxnG~Ia{3qV604>pp*&WSG!2VB@sj}
ztdFRXk?TCQMTty{WDEbaN^Vd}_!y+<-Y`6k?|8^8yG&v%c=ivCEG}@V;wiJ>ECg|(
z(xtKdev&@NDH5$<_@eiSd&M<#K$DJ+jv+#LdZUrNyu1S}chBCvLr|Z<>QvwrR98Xa
z2r1&WmR4d)${8^+u|eUzFJ8Q`u(lRWT5yalC~)tY7~UxSSoR({6|hCpXU%Pj%*8O9
zgdzh$=+Zh&U_4YA5Q3`Gbh1z=0%BCeb(*Y_EE=AeSf%sf_hSnSfd))_{)v`y@~YSp
z3BBC4vk<E}*xjJ2h~k9;R9+BY`UXO5QgU*E7lXdOO60Y%al}%^07gb&mwX9fSb9IU
z8glf!XBkbX`P<SbPO<e+##~IMZSTDMLK`U_z;IlgYZie~>cv<kye?ag;D+Qqj7rbw
zg=ie{W{jLs&*Ux9wzf6|R1$B7U1(DJz7sgs-~^>mgp!vR@QelCD#%-KZ(2uL9Iz9H
zrC!Sq8}38hIk)`*+ngP}juC;%rhh7D3kwT`FTu^K9p7<sa<()!z5-m^+R-tboD_YA
z0LB!8wY0QU*U;$a+>7zD1_+&?p1h`9yBb4bEc^PbeU9m9Z(n(G5_9wXLz*yTDHq~E
zl<Oh_2!nHzerz;89K_@4VO0h-z;VtPua$Num=_VgP{3AesJ11&dpFBd*1n#%oe-3_
zgLLQP8T|TKQ5-iE5Sf;3;YXxhc&CfA_>)u?UT}^#QV7SZO5Uu!cXBOz4gN1TQ_3Nt
zD`KNt!{eFG|9P-xeP7AeKAsdPl@%zo!u$R52&AK`PU8Lvfc|F!wk?0)Xn>nQa1~;P
zBczQ6hh<Da757c*mNyuPSj#1pRRx{me8<>O^&NS(z1{q=v9T9I1q~wl+V;c#)Xg;n
z9O<0q+%v6qCfH$h@dYn!@iUYM)mup;vz(kowE6)ETt{T`%^qnpb2m(M&Nnw-nA|6;
z)#v>jc2dN7jEZt;lyshkp6B#fM|5<Fg;hl8;^MaC_wR4oO@DIeE8*7APP?-lmOh>k
zT1qYN4n3<A?n>?qeE4uy`#mcg1?Mo$?HHg4{zF|n>ec7%&yEdjYWk$GsH^&nMz7d<
zDiMm=BtX?bWlR@M0}_!W(7mAUbc7;Kzt~L}6}Sjy0CMjE<hDg?4<?`;SQdIWVRc+o
z99ziX^itC=;+yf&4Hb-v896EQJp)7(zFJ5_g8C97O*Ib!5)w?p%k}mf7^Kdvj1vOk
zG)zA|y3TR<aBXAb?}lEpZBRg(&i*(l>10lDrw`BeD+KoVN5%wK1#-*$^&PBB(ZQ~M
zz_XWfOeGTIrk#BIjNO#$;zTDZz}W@$PZ7{Uq1xl1<!WJ;WXvrr!beVG%GtxqSKl~0
zISnmzF!&T+Qqe^rWZotkH^sh$lzVS{Qud8iM@C|TiK2(cR2{`TsQ(wAri$T<a&E;H
z&pFa@<S4GSQ)_cV3`;|w<r=nPQeq+>MA3kWoap`CdDC#Wn3(dB5p$JjnXvxKfZ<^*
z!yTCsm`bFK`vqU$fW_YJDA*JfIXmYhrxwxrJ<VpcpoYxN<<rz`gRG*Wsi_GDtf473
zTE-|Id&p_6b^G>n|GI#=d3^Eo-GJo}-B=}VlRti?0qtR8Jm5lv7$h9X=S`=p3>k3K
zX!i(=FsMAH$K%p%E}s#=uInXCGw5q?bG(|d;&LMmDXMeJo*m0~8`k|R<t9E(Gm3*C
zmj)?L)7_?r6Xx_@SAO|IUu{t8f62)y5FAiOSY7X*?kUiTF45ELzC*TOrq__Mi<d>*
zC{_@kmRT5m(^zWkuV0FsMz0Z#o8!kIwt)R=+yn<6`0~P28|&(vuB!Z^EOZK@L``{l
z$(LK2nt>r9o%@PoEA41gQ_o-YDmyznYi0DDZ>o71kdQGISjH*e=Ur=BUrb}MR&5}!
zMU~A90B*WPG+v>4HaXGPw&93@{Hx91zP$mIm-z0T@WKM?zDzPpm~7EGoWtxr0wU%B
zAgZoM0`6n9HmHfxNPldP*7x=G9i5!4L*NeBCLV>C1530099&%M-x@T58B9$4B6{(n
zF<ijHCMpby?P9yY9H5lu<gFlf#(dKGROz_xLD<eRT{jaU<!ozd6%$DaJU{C)?IX~U
z(Yw0xEqio2zEIcv^fRT<l?xXf?5JI7I{x1M<H7~*^W+_+DJkJMqt}+Z%?9qj5gqM)
z-t_I;?W5jbOUk81orH1f1_rdpYll5uv9Z-c1IzQDPtM4hg299d$BD5EEdqKwFDtv+
zOfLN(bEp)|lutQ4=&IkMSy#4a(|lHB*(wf^TFT0G1wQM#P(g;X<Gq|?L$`8ULcN3V
z07QF-<+(vbkD@pR<>Un@JfJ2radVTXsHlLQIEPO_>L3{5Z7@toa1hxn%+0?JR0X#5
z^hCm*7$fQsu~@pga(&O-qMab;)j?3mvZItzsZ?XgVWZ>Yt6u}`tgUN{-RBXjg4k7n
zi(z6`TBdd3!N4VfNARw%(=G;r+!si&2aRs##*X5BQ0r+fcQ*J$PECnBt}dpZ)mxkz
zl7UhN3P7E`TvD$>LXvUB`uYGYAunT!>?-W%Hx2o3w&`qf+KIA7D~1N0s>+Zc%!^|c
zJy-v#jp(2^#sMru$r1R{p<hYNUpNmRR>b4+Dv=VeQB}~Itnm{%#b9HA`tljFV<G_u
zYh{HDTw|{{z#E->`@3LG6EmxSAPhlvAC8*n$`SBfoWMZ=3Wbau^*+?BstA4pkhcH+
z*9I<vEV*2s8l%qvK)BrOa$gk2|DbGC%=$znf<!pt<K5C<ddr+f=Z;9ss6ZJ^<P8%^
znlvi=0*C@?z}HZj3y_NE1{E%_O_l&x8niGT)37+%yBvsJ&U2?PD+BXt6ds|ct|JNA
z^|Y_AxZ}VLJNgQVQI?%kx6jSp7%Fu^!{Rr<x+Lc1<*l_Z1D&WcQiy;N++L)lBOjCn
z+^%?KWo7vOSBSgT$>as%rAnkhCh*C2J7*vLCw_h{z*J9m=a~U?0rJkB!bdE7pPqgW
zQE=#eCzV2sfr$igQEj9v4A9R@;-Tx7o|p<x%DUPfS_#ea8<fxPW61$nWB0a{9QXOU
zy3XR|B=jFjUO+GS%!`*Tk1&xrHhTe{X6*2Jg<o%D90!<M4WNLnxp&>Pu!!Q-cR$Ot
zuAZ{$S+;n?Fi<EFh_Ii0z~ZDpnK!*TL5X{L7C{|JM^nVKdaqFDBpwe94b7#s0J+rK
z))rhT+{2w?Pn+rm4$jorm^aUkQs3A2`*s!<N1zhGBF`~1GZ%O*J3#6KDxA?M@Yv7K
z4;9x(<h&#Z%b|oYD2$_!X+R-~7%L#A0k5}wk&!t82cZyfv<{WkK#(9q0S9m&nFUAE
z`FQU-l%2sqho8ll$^h*UaQ`!G*UaqB;(dLF*5Sds+3SIWo%a?_g1QcFf`fwtr`?gF
zN<A!Mnmy#|Gfb%Cv|F=Li5#hc^H4uiQY0`Hr}jF>DyHVyfmDaSOR658Y(uTd`w9SN
zLnP42|G<-~m0tB|+wMR&s|(z{0PzrM1!RLMaNgXbiZxJ2)<#K-L;X|@@45h}K07ZD
zK$awc(=H?bM8p%@f4|~>?HV`@%4blDq2fzA7u5LaGH&nj@o_}Vrx}(=AvxmN0fC7<
zYr^mrAY6*pX8p^CrrRKNAX;;>E607Y)2yYl)8aYm5qZorqw<Qf-piO|o=X2(w#3fp
zQ1FPh^B(7xqn3iN4T}@z^NxBjuqz6TRTeo~>~+;)HhLf4XXLP@)p(}ILLppr>T1Aa
z7F4rCorr37*mV;?*HE<hjLf1k*v!lf)D|tDaWC*t_eJWpl#Gme$dV#(Fu*h3@9~D?
zmUI}n0m_EsNaH)GOt7#49Ik*!Zb_=}(d4C3d5Z``sES(J+D<N&Rze+s0)cgJp_pCC
zf);#~+r!%1+q*r(AOZlQ8L(#=A#jg<j+gtk@%;@iwZPiMDPAHFZ`aEO&P9ZW!wrlf
z6-PijqZaHI#HrOo+N<VJJc~8R()6P}rwZn5P}vG)K}|nDx*jxC6nQb=6oHKih64Fy
z0lODsAvOuSTx%Vu%|M&F83aWRL?ts=_UP19JplWog~GZ}>;sa1Jsd-jLzjR$DH<yM
zN#OZsmVXcm%Z=_6j(mdMLIfgGK44!x`#wT(q?2cB&`!d~<XCl>0@E7PEpSdxSNA7X
z+m(G9cR@-j9JmQHDD4nS;4)2)f}$ogfcxlC#Aww-<jXF!Y7F)jLHq}rVZxDJm;}V6
z+%TM$cb};Tk~x4LnO~c4Fk4?)2$ON2Gtxf|CL-e#L)wLrQJW%{L?D7_p#$FlwI-5W
zSdMB#ssd+{VXVzVI2_Ib!d7VFY%63hL}efe4x&mOqFzRt;x;L7*fqq&4EEc>+1bWl
zH8AMS=O{(7OhZ*xOf6~tDcCN$_n-VNgyA@a06FS6$N*^uMHdm3j%fAe`QfD4ziRDZ
ziKP;VG%mzmU9@8?+RekxJuh0JmO!0FRF1efh%_T$#+zGPsc*D(_W-4$Q|xvh^s7ia
zuYOM0U?mXM5aElQ5`t2{{C+rP=}lSSegM9}!2}(YNli#XNy+-tf)gvil}fry+yROR
ziHK-l{(icTd+80RW}#&qGaz1s<51xVsCT!Q<@)-1XQqjgAz^(1n2-K<obk>sE<*{S
zhSf;R0+s+P{QBb!B(Nb8$e>WPY(cLq7!|<4Jt3kL(Iy~y-2dbXw4C2wu4+I~)p>28
z4`gMD09!ITGV(R{x_>5clwiCIB1J1Xz-0p$1W2F;C_G1q87Or6v2oMOSr>@L5MF+&
zhG0<0S>P~Zst$t@zWo!n(x`}=H4h{WpxmGP`%Q}hd-ikZ%hntKc_WH^SeQheCqeBF
zh05sISan|MLV8b$r#O6nJ!}nHE6{dAP<$Pxy#WOYlmKy%_PAU<!9x*q2)P@T>msJL
zUxA-%hVTR&7F4eY@gRsUfj<+BXk&zpHb^rt{GZ^d(CIWPgCIJg6zXUDWYK@9Att6R
zRokpCTnwo;rXUWP`+Nk;=ySXQvJDZ!qA5sr?&Ilsh-wA%l9|Z|++Yo!1NmJMG{|ic
z`m&}Ftj^%SufDMJ;(;uJt^$q^(1nij8`}21zP{C?(xRe42)cfK%@bf#0+}0WT*yAa
z6HuJeNWAH8O=5$~U+16M!X^rDN7@He@}0RhdJw2I&IBKIn}Fm$*Q_Ko7_4nrHP)7@
zVPLTSM!Tj0$#+mqg~@ul_>6}7oUSSn_|O2@2wm~pM<aBgPxtYo+SsrDCOsP%0F*Ln
zW3bo_1`Zi^d%IRdEF5Qf(y^Jp_VxK;*X|;C0qUUgjfpeCibxeh^d>+?z&QM%YS#jn
z%<bld!c&HR7ywscp%SY4!E?I2%?ED4tFYC90MiCwwH&S#NY?eRCZM_oAZ`{o3@QSZ
z9|BUm>4p>x0TOU~bpzB>!H}O}4~(G<1C|)*DUAz8U!Ei07#Lchd?NqU0C@JGLG~d#
z-&arZL#v~zgzy+TH3_{NsX)e6M3mc+w$#<#jaZehhOS@~Acx(Aln8+}1hKN9TeEQ$
z*i*nf=RpJ;e9Uh6)v-a=jQZE`ZQxE3Y2VZ&+@ypgRk<zbVyUFk@UkWUvO{^ZtsD6P
z_~!eSz*oy6Ug<f9$OS+zP*R<QQwI2yqz*+2>^Pj!0GK(z^*TTSNIZa8I}Bpz24-=D
z`iG6RfCka$9MGdxkhQYzRy^5RRCow3{=jda+ucV_&d!7AT#!Ho@dN3yl1?MH0PIm*
zCs3gPv`~ZfIoLp?BEUX018~wz)iQzaek2HAQ@&3H`WZlU4ifSnRa$NiEe+yDD*}89
zhF%AR-xVm727Y{2k_2yKu9&VwB^D5kh{Q*dDFlq_xMS8ofxs=9EkZJv5qI<ZWzVS|
zonYS4kvrk;^9M~J;6O2tBM=DCBLFZXxTn}%1PQDEuzAGdqVJkEHa4J?LVyhl2?;^j
zGd(>Wcx4yOpX(g{vwZiqTaYP0Ry2X)WnamDSUx8DSnxh1qQQ?G$NyX#=l^*l=zqTf
zoyq@1OYMJm^1nOzKij1J|F=ca35!!QQe$%yj3u)tyq8=uh5k5~_UPD-M~{?px?<Z>
z(%))kM_c-1x4OPR*lMZQrI^m9XqE6x{c@UOTar3=j=9<|uTT8WyHn}Gr7JF(uVllf
zrvoJLX2)u7n(~Cv#*Z`<h(Pa<Jpb=<j)R4enLJKy`Pu0Xb~EyB$3M;P;ER4I`jD0f
zKM#&Rd+@X69sd3Q%9}?LLqZP9E)N9iW}Cl;a(IiuUd*u_kGB^7ykm?<3Sc`cC|!y1
zD=Qx7pl~lgz=V0k%Jom@q|;R^t5o2vug{MIU_B)*Ep5LI{`?!$&qj1iul-2laj4`a
zdvk&6#%MBQc4%$g{G_ez)ag7xH{0I+hPiZ}=bue}%ZWBViFxBB3crH_m(-(m>(*O<
zG??4}`2`gpjVfyw0MgVtU%q_#STz%7>*<Y;^%x&|0(uQE?*9Gz)wQ*v)6!7jzOCMZ
z*?N5Q9k+FKEMDHbfy1fE%Rh1Fh6R_$Y@E40)D+r8P;a8Pkk{7ME(LAH>`>oWc7D};
zx~-z(v+aA2e4FydV3rPV+^X%F)2D?yPM<!lsHk|MOi)m8?=>9e!LjTAbeunksIIB0
zbN2EohEBqyH*XH%(DyMmEUR3ZpQV@>6d8FImy({2GWw-4SS$Cx)~cp<H>3Z@4_@zm
zGm1{LTQF;W8+IrzxxJX*bw*KLP3<+s3O>WX`59yGre6*pPI>?S6t13$)PLL>*1CBt
zw&(X$O!RS`|NX()gZ{j*<|?n7K-XQeWd4tf+W+`_r<l07(u%2!FYie{zHlq}>IY@N
z{L?z!_XJq}h;b(7?Qh2a_+zW1ftJ=#M>q!a$$xWYATWlHgY@Az`R$Vu?7|HD=F%F~
z$~ieWl1sf;74KbR#*`B__x&mW)&iG3Zs!jAcYfA0G!%@e8=`f=m;UQ^omxK$c#9-{
zGR;aZF76>YgRSk`Jb;n2eS>esV2QtOEOC3~udKqt!YVZM!rF_?h2>v}y^J<~iRE9P
z9C;;HRp6VKo!x3D24Bg(`3L(wZgFvOQBy`}nUu|C`pWjPvaxBKo4;@IM#o6lu-)?O
zPwu`iEYxjI?kz%FUC!A2gc?k?fL?tvhkHi;`t{Ri&))n^1k9X%dBZyVLF1}*_GBV`
zzNdEYUMw%%oc_g#I=h_zxIe+Mi3egu@$tMkVx6gmyL$m1*@)1_rs^BVlnLPAZYSlw
zkUt3sCo*CBJU4Cjt8m}sd*4vsb_|AoZS#`)JKeaX-INp+y$ttZj^%G|+6h<F8AbF#
z%D?UkYQmv#*gboxWB)j5`SlE^89sjgi18mk?!|O~d)WT#QXhTA&zwo=t>oq7tG@T~
z#?OsDy0M>v4kp*G!nXYYQ_k5|{N(OZ!bQv_`%N#n4v}K0FmA8mXoN5;ur4YMOqlZS
z%{_cw;@ffW<EaxTVk;^o>%X5nb7s!n0P~=4)5H8Hdh^xP)G~FY>93{4#M<=P4<1}8
z7Q|ed+5G-(Pwu|qDvo+#_9~uAb#hcjMtO2`|88k(^GDS2CHFJ8Pc%0-qnPqgbsTnB
zeDfW@Pm_a3U;`rMHkb6!o1W3?Twb&1KOJLB<z8}zyeM^=f`{lsj0XZGJPvwfW-e}i
zQJ)G6iI<h*!Zjik5b?3GC-A}BFgqMJx3D}4?Z5+VB2fMVzR!RY-q`h%Q%5rSa6o5Y
z>_G3uAK!R=TQ67BF?4=jn|{KcFCBYaM5IN3x;w9dA58kv#;Nb_;c9vn8EGwN3ra<1
zr?p%vE9RHtf89?J&tZrN@-L5^#-)%*+G=V~8&FL6myz{N{ka`?Sj;l0fLz9S+W`UR
z!G>3{OKdNOb}T_zGay-pq@+Db|Jz@`DTw~z=O@6+`**_db8akIe|W&Y@i3-*^<S5U
z(+yrOazO#~=FK49!9gx9Z*O1K|Ga!*<F{|`L3hMuG*E)O_UcAR+>hn0GR?ARh}Ms3
zWx?z)+}J{=Pjb<h9Y`>G1vJ#X4A;r(IbCH8=Emlqt7adFscs=hNtgY9UjOLZ5%nu8
zF(^7ZqqX%muKC^7WtF?Y=|^vv(A(Y7&&@yFIcv&Jx_hWA$6pW*p0*JY8@E+v=($R}
zowR^rdq_$^pj!Yo{^z+sDs4QeDldAY8HlPnF%%jiHg0_65x-v=P8~7|kx1ml0U4lW
zW0T&2ZhO=43i~WrATTuPm_?n$!I@Q8Rn6+!faS~obEuctWh9O3Y-`^L4|xD_cNF~b
z!G@P#3Y7uVkYp+D?5zF14db>M@eJOlq+IBLXwh5v1<1iVF`mO14Ef6D_Z^XTD-wa-
zJHf*f91_BZlk%XxhswKHo%7#;qH&uii=CTW{*^0N3caJCz{Xihn@={ryGlL<CN;2m
zdw<+LG<4jmcaofwg;HE-=x|dG2J>q3&$qq3OF&EGbw4^|YHIpViV@%3(4CdbMZL1W
z*V%r3b$mz)i>1U_{l|u$UR^V&n%_%){(R3ruDpH2zohO$D0C-(;^gA8dGa^N5y5v4
z9W?+-X!Gyi56yUVK^|%OtOqjRvuC>kNMJmhA<XHo2%f{d7L9K&4}}Y}Vlb*3k+E@S
z<%5ovf2_kcBY;bKXm$Zo(?(i(pyDRor>vwT(((B5<D+;s%r8Bg_xML1Ga0PupGEH9
z2!&BICM-f|*Z*Z&`^bCU0*OT9oZNK@6&5fNjZJg?OStdF%a@Bw9zf>(bPkXN!R+0=
zl>Z&v-tJp`ATA@L4{tFnFm-Uq<W0`W(HtBcG%Vo59Fy3*2d0YiwGW{6wcV;i8HW?@
z`26{^VUZx_*wW^nKkVw#&B@JGoU30mkoIIDAw!zkv$1F{`^YjSSX!UcrhTzME39os
z3CE0-%gn(82izwM!7i_E*bxhms&zS}ErwE%>!LTJ*jwh8Ow$jTn2TKuDj%SXYt^ys
z->5^XFZEYe5aoC6-1+(%+60l!O(@sqIh>}I`rE#JCvgB8!4C|JE`m93E}RL!4eTo@
zh!<yRW21Tf`eSVpD`v;h|NJ=buw?PclP3*J;B#*O{+!E9mx3sY4G*R^K`8U$UPJ%S
zSJ16y|LayKvhG8pp-GMhO~D-95qTr1UXn(vl%Add?l2&#4w4<#SQ*6&Cj=_d&!oM~
z@8<*s>&H9z;bk}G&6hs}n7#7aH^axv`vzgFORk%zw6}7vA^y4f4pvq%-~Q%ib&e<B
z7i2M*l7D~UMBj^m09w`2Mc;OqVsQqj7Bj(w8R^}yIdeawa%edi50oVa2G%)SSR|`1
zVldQA6VHxIORLqFuCH0(tMqMu=I`&F$%1+7x_S3pU^YWtu-`|R4fpCYW1Kd_#^j6w
z3#tFl?da}%H?Yek=_AbK{CtU+iOc_KwAmGMvB&xN<iQ^0Ix9Z<hC-fb<k~om$oP*R
z)nLrz)j-j55z&#w4$y1EA`^zX8Neg|h6s(XqXftMtgoDgulWg=dN&NK90$F8bvZ)J
zhB!qW?lo^)dwaM+1qPG9QJp;qlyGU0s;^qu=2`v^5|Z0+eU%^IK1IBFGtyFmHfqx(
z`*@uod3x;He3QDm`n?QrMX$n8qjJ4Fci`=?`D`1<#(#g0n|JT0i0EkBjGgc;QyaOV
zj~6_Um-iZ2uUxOdxpQ^;``}x9HsOgc(83~p%8H7_wy|#RD5HD#_8&beh&yxY)XXSk
z5%Om0W&xK3<0eKHS$TiH_wkL<VbqSGr#C(?%Y-S}Jc+$f-PAf;TBhI^o%Rq%e{93#
z`)*`n7SO3hI_SIHL@r#Yb@uTo#V>BfQ1@<Zy@b8YNG+ae`t4=k+MkJN^GqamTw|Gj
zb^L1W0>~R{nfzc6o4Df>Xp6Nu{UclPvA%GM2Af~?fu;EF-MbKxs5Q~Ev$MO{*g|=^
zWZZ56t~0@7E;T6p69>vWz(ippg%5XF!v1ZV7Gv$3^~XD+UQ6~kb;xyKe(RY+n76|l
z=GSNSvY!LWtC5b9NhT66aJ&s9PiBmV;KuQk9{dw<0t69|2i6P4;e(q8VCC-aKI8!!
zrcbZu-+%r(Bqc81@xOo6Wv1`4qpbC}^JP=AmaV58hwQYpLh!Wh7^?opO?lOQ^mT=N
z<I^iGBh#}d*dID24GpI{)|oJ}o0SF;GM3ZQ%w3%Nl`D^8<Kos0EAsF6*De4ny<Uoi
z1#JMcOE&!+2=z;HyA7$DMLCBQkW0NiFIj>op53_I*X7@{)Z*l~#Xo`Ok~iZGScc3o
zF$~61=^;E<;GgFLcpTig@qr^VFOSMf9TVf_=eL6NMC=-3Vn1+TTl`+w>&;3c9(w)X
zj9+q|QEUbVy>>x?+k5{9DGWw-BkCL*0W^PFT3U~|pNlK~`Sa%w+8mm$vxkk-C<N^9
z&cJJIq|r-_HC0t=8XAGT7JvP*GxN?_rqsc-ta^1+o>+j;S>4dUcIO=CLCZ!^IyN%d
zThyT+<31F|ly)ssH@)!3Gl5zX#jXDZMvK|4;UP0iw#S0yPo|;7BWEwS+>sID1H<FI
zjsK7<znDXx{*cs}GiM$IXBP~GTTwxusXD)>WO*o}a5A@B#lP?ueQ58__Fi@2c=9f*
zK~4|wrY*kW{2m)hsR8U<G02EhP81N{7S^W;cv`P(%wH81C7Kt1aoPxr5_R!ON#lBy
zl@Jgy-$SNL*U6FxVuIC4XU7Q;P}+F{WMZqU6>;ju#_=G(#KTyJ%W_Y6_+b+>Gb>Oh
zyYlbHY51*Gkv+!G4Gs*L!*Is47cUlc*DsylIP*SHr*4}(3+SXE#)sm4bUO4MUqB3I
z)<5v*6AZm1u1*y-BxkJy4%E#wehRMb1j9fvb4M!9Q>O_GJ<zLLK+Q2e^yPHG(|~{u
zn0HEraVm3YSI{kTNe9CViik)B`E|TtI9dZRy!T4i1M;67P!7OP<P(7I$<scxB4YfL
zC);5brOZlPbUp_{4}^3J2n^MAb(%&-aTKcu43G=P-uWH9P|*uK${2{I7pL*L)?s9?
zdF7<x{V+fAQhKVffk1bh>?xQJqFDe<0u!F^VK7GzyV}{{^l2Nr`KL$QJf;eUYd^Ae
zc6NFUyjIvUca+6LO3KY(X0AqPr0<hVuVcfB>+(_!G9KR_n`z#%zfFVDf$P_@vHqr?
zpYdy|sl9`l!Bjhn5B*dpPew7H0*F1;Rm+M61)fd=H%cLLfSsn^t+?Pk6lMmX>=SW|
z>y~rqW~sM9M`?tPZq%y<V$Ue-RSWbEt!27G_J>ZB1Xr@V0xBsdj~_oSH5}<EYSn&L
zIY!Rl;BM^n!oeqZp*=eRPJqb!f*WhtAh((ZSeDvNp=%-Wo_8IlinM@`06{YyoC8l-
z>*tOl^X%NYGyH<dGZ?#w!v*qLw7yrHELHsyz)b=r*$U(#sN|bLV9znFJKXEkqD*)H
zbV=xP`S>F^gHG%8NwQg1r&Mu6RAr<}PqHg*N{BiXL8>`#t&*pGJ199e_GWh;A(2Ym
zQy1+W2rabUYDmSwP<3%QT;X)-Ds`zpK*@4E8dUu@lTd@97H{m@%vD+XQoqW(C<7%W
zf0%}g1t!iMn)i``LBF?`F58+B*dvHPfxM>yB0^Ju-Ts3I`EV=C4TKD^6Ps@zE<d?D
zGMW_A4#N+V)HUei?Dd>3372$CT%nHkH%9~4)8We|_a0B53nj)U#rWv%<5y_~MYLq8
z`T-draPQ(L7Dn?>oeg0IX1cBqM@wI2-u`zo3^YM5gU;e6a!Kb(pD&|D&#vTMgl(Z%
z`2jQ_cd5a#O;dZ+U#PSZ%UD+Jw7w`yZM=$BDp)Y^TS8KjfTW}jlqrUtH4q6PB6xrm
z)OG_7g!aGq>}*j;E0*9JrZYuGm3w<3)_72x6r5*2vwJ3N-zu-H{36=z$IkijqNze9
zs4Ry%v{E$#<W}AT31{XqCxwiHIuaKE)LqqN2@<2d;iDd_lX-1jT`4TwUbjF%R0F}W
zLv{P5^6_VSwgs|{CnH_v=Gnr&Pt}Q265m}FiJHpOT|LFer=^h?nVMA}o(5v>*4Ufq
z*hXfdn)aFX#{(<WHId;3A!%W%XgWLqr>NKr)2bwxn6&!#;cjnNrD@hEboVxcSXtK%
z1F>oNwIA#%T!r8bT_5iHLf6VWID=H^%>&uD6<U>`6a0w$Md~eNl_<9+r>8E4TV+TN
z3d1e3$x>g{V;Np0FwcW(H^`>%Q!@$`L-ie=DljIw-TIgSLZ22AQiWm#xox&}=LMZ~
zGFI*k7K;b5UIRP`qLwrGwu``<#m5i<z1L<drM#9tKuCq21YT&L6ntMsvc4Dry*LSJ
zX=kr1D2&6zE46AhmVZ2$iR3*r8GBu6|H1Z-(bA;ol_>BFQJ*!P6*`SD;sngdYiRJ{
zi1J?CZ{V{`tAmErPkMLUZbKMPldSGzBitFI*14=Nh+F*qXH!zEmWD>Us@!sbgCp&I
zMuyhtC`Y*5`W04gxy(RqWAkaUz72&Q_VS42*@_PqedU+KMMsak;1F31vU7mAnu<U7
zn)jT}T&fEx-+R>z+?uYPWx9SNPCn}Ukky53pY>E?yuaYe;_2^GRadTFbzbb$koJZK
z5)9_85#U#kA5ZzI(3&$2y)3sOA-1eT5#iDQLL5HnVc@;c7LFy9#>N?N2X-_!$3i2%
z_6l@ATaXvyxYuvEQoL!WFDu@;^CqKo#nJ%gt&v+oyW<(0ZmCxRB=i&rfM<B)wZE{I
zw1d+np#ZDi?P&u&g;Cqhp;xIC9VMx~v5Y`ihf(Y{u)8$mwO!dbRoR6gIHRMa#}R;i
z&=<-r59~|oPS;8`t&dEGmV<3)N77SMTVU3cq>cC1gt>4P+KZ#Ix~{#>p+JuO_Nr^k
z=kt5Km%jVye!TY>va&Uq{O#LFp3Z>mb<*>q%I6&X5$T$hFAqrd8shOuDJhB&#a6~r
zdx?>%rnNQQLh#oLzW|5Ak~ogymw6oW-95Ui*h#LHd^6P$`cfsuqC58-La>gGS<u7L
z5-yfXt>Wj-S=EL1#D!)hLK`%B#(S?iv^L%%efc8KCGkO(RFW$X)0{D_Jv}|PHc@p=
zIUSx3Q5nUF{gnbR>v|d&F6Hdu=ZOvoQw&qTax2iakOK3kBj5eF+o7joA>Gbf33@g}
zD<t6j0|;wZI}BW-$ez=(9Fyzz9z-aqU<$0^bWfDHT{P@pw93^|V6x4jJy_*(4Fm@<
z*C~B2X_s1G?OlW}XgWLUO;wvmL!vH4yL&WX>c6Fe4QuUF;xefV!S*H~m!k%G;GED$
z)xj#>W3QMmkmLMk>3qHP%6B%xd+46=D2|cdQ_;9OT}DE4bRMG#p(;CFPoC!IS8(mQ
z?+1M&X|SrYwu3j8mzN92Gm5gI?J&t+o49NDZfQ5InXzE*WUzO2Zo=9-kRFmg`TB+D
zu^&7d%=@{LKc%?%HYB=dFX^T#YI-9A2}x`d4guNO55cz+bPF6F@7f(OR=RrhLXoHR
z++w$#`P_g-;Yh5~{$GE67ZQ^DcDYU*-%`d{Cxpwmhr{g;QzKb9$-sWOKYRYXd7-00
z&C;^)?7`&t_&ek6>&lNGD=d8B90%m$s86T$k6N8Md-fIYNnYM)^`8ZAdEPxEBWDQg
zlD%4nh8fWNI0Bt3YG45%f-E#%WwQW1DO}bwH&Hcas<p}L(34%ec4gnK*pA>B3@8%2
zq&V0CeW8sKY`d4EyHL)|c1l*(0GT^%A~a+<_d0i2Knu$XnO2$v1<UpYQfN#D$X8To
zKr?Sk(KLN4>Am7q7uN4y+V`|Zu;}KtZQGEA^NzM@<J+Lc=3Td4nE?#IN3TGK-sb~d
zhn~Cxiy2v`Q95AW(`jw#{(LFqLFC_{Ac2?>f<PaDXY`N?h(+;qI<0JcWt`LjVPPWc
zC3$Sj(uQ2DgAg2enhDwV;~ghBWlKN`z=O1mlgezjheHa_kXsqyF@R~yLYM%JlIs~|
z1yn~J48@NT7&O8H%;Ue$$?5-cT>_%0ouXn2?`VyX>(~$fGD#gLh;+DSTB&P}p@&5R
zWFd>5{GG)3_V$AZ_8(DW(7!+X^G_#tp;f$k#+r$kZO>MN`=7<*xB=FP-u;Ugxfb51
zKm(=<9S)D*((}9mquGME!u3UWmAg(34k<n>Ez%2r?<>=?ojR&{J4irKkUUr@>#ltA
z<Y!{nmb-Q!5(eD9WuM)<3b`Tp)i1+2B`U70tP4Yh??-@8eCax8j*-2ttQ=e*XMarL
z`0;m4Y1{1ql*H-kcKPVN&(2oVO10v@$0O0&7InU(v$M#fjV-Li0CYA;BcOS#U1JZ}
zyBAREPN$tm5&DHZ=Y;UVY$Z2T97Vi17#fB%b{IVj(0`?k<dXoyeEB(fK#sz=_j|Ad
zONg4#;Hm|B3yhhkgLYZ8uB%V`la0jXg+Lng_Cy~fgit)j)1gH1SQ$+^iG$v_HfA=N
zzFB9`oid81q<ZgHat$%rYU|&B>;vg_oQ<&L3kAcepFAkEhbBTVqI+?YG{C$h1QO!p
zz2_qa!R>v)BI986I|b+sh_nc?x1>#w&q_(@0v1jB?YG~Yrb}pw@My7T!NCg@jP?*b
z4#HRL_e)puqPqo{$sazHKJ!<)b^Ur6JP@Ka&rUDYLaSXcA<4{qrtf&>oisr27nuz`
z%_VJn++y0C>B|&L+M-(Aml3d#l-Ssw(Z%B8b$o~Dr_((-;}opEMpATAx*oQv)Vm;D
zYC&e2z4v8eqC8F=XxokqcdB`*gJV{A3CYI#nnn?A!3E}dKz`im#DOVJ%gihuXJJY@
zdF7Z>T5K$D<H@%XDv^DjdPGaBRAz%m+Iv90fX>}kEPa;koIzph;4mab3vfI|6_u$)
zGIA{BAXZQzbipKL9*LR3iS;Q)KnCh|mcsol=kj`DYin=zdeeFh3k*ROLXf>FPE}Uu
zG=M&cM$<VaH3!e#`A{yfS}+ckN;7z_w_^EGP&Ltr_6j`gAr9gxzM!ChDOu<jlqF^7
zzVbjp)>g?)uqH#kO3loS#b^z2&d&BuA7opryddW=6lU`&-FvT{fk8TWteDG$o6cFS
z7e^!&B_;7;^M6j<9sBw^$tTX!f4${hU?&43>mFiz|9*EwP8RbwXk-<{Nx7tzbDk{x
z+sEDj`YG81wxTq&)D1J1kj{h(79xqn-%7x>la7#wUJebJmcT<bx*&Jwksg+Rt*W}e
zt5Jm|At3?kA@*XI5_p3wt;dn8-mW*xECAqjnVSvPwPfjs|71N&NY@uh)FM41Cjrk2
zNCoR^nbN-pfq@IC&M?U<jOB*196Nl*lRX@Xj~?aX^EfV^)ll!gY+X>$iE^&>1jfW9
zr;f5P^zC7w^P1J($$%^GNu)sD$9@V<Zpxk#`r6v#YBp0hZd$Ki`y|)$jT7dM042^v
zxlHfgxl{Abotd>WjusZL_i|aTUX|uCdZVuHMT_3rhjn+CJR@W`ZY$mY{OPklN4#vM
z?F&JTxw5ceV_+jJT(xVrw6~P!FK(Tk(#}7wb*~!z_}h^K+v0l!Yxw(1JnqJ!TCdx<
z<z<tHf}e~4io$ez5*3iKRIi+Q<1r}aG_O9{qvqyDfJz@4T?dDz(2?GAxnBmG3Fw1F
z9SBrC12lGeiNKjNPK<RLs?LXOpn1B@eOS3;Y1ZrvAD`|)|AbyzxB$zi&I9`gQs2Fs
z9ERt2D4gPr@Az`qHmh?_ltj~_T(0ib{Rf=K-e0Xa4_yQB*ukKPB5iK3nTLQ_kv<1q
z#R4GMvKu(PjuF6W-o5({;$%$9`;iF2x?g$Z&A&V!MZDGu^<(a;p%etZCwBwe4#zSv
z{}vn-6;~IL_hxz)?#KcXL*ZbEo>K*@_$deo2tz{Gb`D_3B-%Y_3`kVPs`AJ`zh!KE
z&9DEgv<Jh53g`t^kOqpMWle}`a7>I{=@39fZFYV8htLL|1nvU~X*7nhmPCj7aOt{V
z_U_)@0?!8$gNfbGm6gs?mMu}!Liq9ba&U8596+u#L-JgpFs5FCh71Y$XPHd#*51c=
z^L$Jo@xm>QME$;o=Jmhi9=?_Kn2h<Os>;UEEi2Pyv{}i_a+-Vuab9{7lID%S;M9HS
z$FIxFLyN(!Uw?fa^(2{Y%4Rot+0Jt!<{ou=nw-BiB7K$8mz+FGjBjrKP3+>uLYcl|
z5n<t}^@gi<S_TFd85Mu6{I-Ru<PW2fri;^2(>ByS>FEoZon3SSkBrO}@6~8I2I-IT
z@s{C;u-fn6U+-dN-Tk|my@IzlR!lr^{Vmt8zg{m-we!*Vm~?*6RhVhl#^Ig<mY{Z;
z#aIAJQ;D*LoZf2S+MD25FPT)q_yWUr$ng*p1;yqhxkN9o*Chg?()YRiC-49!3y4>y
zX7P|vpw*rP5TSSUcX7LK1&*nbQ)4}jLbIoS8MT3I^~|4!BVM_>diwZX`_Kj$0WU=m
z5kc~c5y~(@)C5h%VooECf-&O)(DNP+IFn-vI#?{Lh$jpC9_=n%%NpwssA%o#iiMyr
zHC=3{WLcvOJ?hTu3nW4tG#Mk)gH$pGowT=rW`wSi$SaKl1IFuR3>t$9&l>3%otDJI
zZO3>ohCu_gt;?`2TpkAg)qxg9-mgtl=G*_<Z?S+p4D)&%{I~Bp)YsCr1!j(JpvVn3
zcdp{R)5JGngmjCsqN2}Zl2n?lK&gHF__VMvuB}ZIrYSVF+HF~n%CxQxJ%7RVR?d6n
z6#uY*<NilN!_zJb>gvzp?Y<~vL;o)`F}}4`bpB!`c6x$^KQb*oel<(+M}L1^Sa}SA
zZUG4Z#YuZrs4Szs=5Te9k{Y1~E@8HXqkQLdvu^<9B3{u8Eg_b00xLkmBqnPlxg355
zvq$3qiCSUOD(>soT;@tPE~)sKwE=#pMeRcuth;lS;b}~%98>V1hw(78=q5-f!wdH8
zWOkr8DGWMI3Zm=DoEP`>q3i;nwMAVOI+l)l&0OY^Y+ls+w%FOoZs>M55Buj>2Pp=v
zCLLd92+UnUiy<)arup4Ts@%O^^K~A;(>Z@WpuHs16Wk@Y2j+vqp&#EwcBY*9*`G^8
zzEq>imy(Z*p#QuHDw^1$>Cz%?zanA~kr)dTVonkwlA=MBLsB{<R=1O3Ebnc43v~cG
zN%cP$sar*w`wOY7F!cnnP3R<!M?N1*$itWAs_vd-32+beE}eqMuie)39Q(xb;stA0
zp>tMzpjF-};g6Gn?aj>^Y(?hQ)(93$_2i9r7&nP^)Dn5u;VCjnJPJ|_rLu&C>qRck
z#oXuw-daIsB`j2M({I7Rucl6CGS5liQ&73sVd$w0Gdfv%Ol^|GkqF3~k&`o2*JJ60
z2SlVGf~>bv9O4alS6h@TIXOH$*Y_Pv$1VKu7qWz&_+B^X9YL>N{qmDJQ&;b{lNlZu
z7y)EEMU`B%orVI<9AF&R#n1x8UbaAKD~JQc+KA2;R9uq2w4k6$fb8NM1f!<Cr#%&K
z+_)7Z=M79{B8(P{0>*0uNVvtJF|y6F4b^$@@QTsgZrhf=Pd0Imh?y|@vIlz9^a&85
zw{+=n?Ev1Z1za%)p7jMWSozmqe+B8N#kI7B{%6DQ?O^LSY?wy<xJyX^So$rZsFQhO
z%Q`g3w?)h1$68dow?9ovNGK{%02q#nE|^gn`D?dONZVM6s@n;kI3P5xO_J9iL)ZR0
zPTTzJI9yfco%p%A3+Z|^aj284YmX!^Ep1I2hH*<BKz=XjvjN)96|8xPE`nYmsHaF}
zrCJyDQWZ!RUtiq?VliBf;gt#_5jwD31E8le?%lg*lC${w?xB=)*>_MUFOWSsp+Di>
z&V79DlN8XWF_^7^;KAnfwdJAX=D_|y5^034x!}miwC@#uR>0tyUh3vTui_ai+EDl+
z)(1dxJItp#!4qf@ZFAehqY$c;9GmWoCG?eXm3cLYrX;|*rv;~43rG^{h~7cfHJtnD
z<a6eTh)93PtSY;Y4_G7!cxtw`8SdR6|L2%DoluUJDd-jh->e(Gl~Z}t^CraN@x=wH
z3!u5ku8hPPmSu?l4Z#rT(r{_l4C{+}evYZWPvKBuEGGrVazq;wBHKNj%dKK6tXv&-
zgukAxlL`O&^;F*QH^t^ZcA8niECsNHaWJHJ%K!7qLh-wZ!nR+%GQsnJ={7lYN+Q-Q
zW<5q@O7-SVFedS6hgIkk1U$)ifSnzPHlJZ9Je0IR9^1MwTxepzK_TGUBt2c6kgg~0
z>Q)jIblbpvP;9Cg`sSeBzXMdN9H_`edSu}#Rtw*Z0^HNHno#Q?YJN;&Zn<dv?JXF%
z(^4d|RDd#CcMdt`d!iAsQJ}&*06L<GPFWigM`fr+dA=vjwk`&X&&jE0Dd7gy40MP#
z7)x=g<Z{hlH#{XI)B#UWth>Jx9*z<x2y<5r_}grzIg4ZXLch?3F;^u(%*4R%p>L+9
zFC_*ec=-6fM0VdF3frgRUMle0zCdbVOwT*04c_C+=oj`xSm$cFkH7j0dFKGXtNG$Y
z1Okxn-rWVIz3J8Q3a81OEaytD-CgUxBD}n}=Z7hz`36P^vc8XZf{DaLP!$w=UAnoR
z`HwGY7<AwRubailYQjiwT5871LlnNgf4AXR`ttYRV;~c?1G>+MS+|z+dyhswE{Q{V
zjmYDaG;8VS?$9Cx4oV*gwC{jMNfp9!4c-YzSJ%U>Kr5=iXkH{J4L9<;m0*-ac4%S%
z#^eyYFES01)cswq@a&C+es0D$3+r+UeNh}96sQM}HcZ)jRA!`k<Owtgp@afG62y3s
zdWnh&-tB#d{qMR7JlopG3dJQ@YB&F2SeQ}J>+kD$98(_wK&qy`{xO`?>cr@?XFq&}
zsC!_{5&R9plRxKsc#3nO+xACQ?09Q6O-;bED^#TehZW0PG=hS;)EynW!#Gc$=JBO#
zKR9UtgO9o(qVQV4lwjdh;rJGPwTKO3Dx`JMiV7f5#4KUuDr7-2f#Hn+2*3jr1mU?x
zD0%^H`DJ=K3LbQkl9{Pbbebv%eGlr0PcOsCG^_Krez7WX|KTpxojZMt+~ik}P#i(q
z&@FTV<scR@M(}_fB_h%$;7LC6K%86J+O~sKmI{UN_`sit{YS(-@-pB=CW6?u0G2`A
z6OHVt3Z&y21Zyo_-SEQx{s4w53_0)GJ)SyWzv_~g!9=}Q;sI@t=tmZ5NYOqk7W3m!
zfKf=_Gg{m0q&*j{;V8OEt0(MHHi3tA-DUvU+6ts7%Tc%TQ93jUg~3}FyIbJN1sxzy
zQvcqk3hZskkkL2@d+zh5h(JLc8uEk5oQ^XmPF#U=^Z@F4VYbp{_=HZn78IM=-QNX&
zsom7{SW&<_bM9PltIMCAFn`3X|LdcN0N}uTfEZ5%M!9FS6h`!|fZa3AA2)}m59(%_
zz9?P#vD<sDiXZp{br+Xh5SmWrIQudh7;6CtOIywKeL{4%q_!>*o(UHVXk=yaGgHZC
z|4vdEv`qtF^<`}Z;E8e2_O(Qq=R$<S0)*2M5KCU&Jp>$^KNYdS>nqMO2P^J<M6yh{
zs0EQXxu8HN#%CoN24+)1;a0MYC@8X{$jwVa1%tT7G&r8xJfU49pEqvP-d^g6W)?h~
zB{(SP29fXPo8PyaJ}CR+k3Y;|nlN0%OeK%>D-_1N1@kYV>Hv}Qi0=ha6ZMmwIB~)W
zRNhhG{yHEvl-!a0#cSh>FAE6^H|Y-#4sQEEya;2Lxdo-)pKxxuv$Ky2ih{;JcJ4z}
z(c#0bBs{jTnMr6v=iFKXSsVHbVZ!@C@C{HGS3p7JMGdHh@QhMsUA8{ugZ|goS7~fc
z9G6J<Se;N9fK;Wcu8umZPvYPScG}SBTy6;>6y50M(|k~O4#0Ij=wdKQ*bT{>s*D=#
zfc9X$!^Q1QeO&;j=FhAJ%_?Ra%?X#9->b#dI;*ItT(UF=>5afx#Up}4!Y`ExQ@GK)
z^B^A1KLV-ZrY0uIK+X{1iI2vQ|FC!3i+Ss~Da9A~!owU&n3C_pu;7pfH1RPdGhrID
zHX4PFoxXTc3%VSL@i6a}nv>HuEoN+D5;hJ{eZ`vqO4ep`3JUdn$u<XVp@9@nySS%f
zk6XoGe~I4vdq>$68)hVIW8as$AX(P^ulCL}s>!U|;{g<Aq|LnQh$t{>6qTXaL5h7e
zsNjHt(k;j!AiajBj;N!e6Tk{6C?L{C5HQk5Mx<MSASH=Xf&>T<s-fKfgTc}Fu65VC
zYu$C%{cv4=f{^Dtr|h%OKEM6no>o@T*28*kt>@utFU_nJJ3gUCDx5nFBdCJmd~T|W
zJHSsYfQ3Mt?HMVcKA&$l;^~Zs3h9zU20{QwQm-u3de6uk=CPbp5#;;8$OE)6%`#h;
zzj&d+KyVpw1k&Bw{gBWf*t~%B(p^X$^=5+OhG#(O50XgJWoI7zsr|4siR5wi6IT+P
zhQ%knP;(z`KQ3*bKHXT*)d(CDUN41)G~8NDkAf~qQpf1@gk@{S7g<~8q9JfzX!jwU
zDX=vpZRSm*sA8bY$aBHy+8UPl77%ajE$nLcj|9HkSh8%S0U^DLK~O1Q11VczqUb}W
zg<Q>9M*u<uzMV~asr~tn5w5flsIXO+FV|p%E?JCgxLHDUM@Dp2Y>{Gb$Zx+b$Bmyl
z70{1_aHcKX4laB3Y9m^ZC=)vfvC2+JUQaOz+%G|T5?TNz1OGHBQ(zqsZ$$U>nxT}r
zE9CGlx9CYpQm~5fy)`fBvLlXpF+S=eQb0E_G#vgC?+FeA7k$gYS|T^}=#L9IlcFT{
zXYA;GwG0z5@DwNLGCgd{xGG#tV0wDZ<*V3W_WHj5I$*aGUF@ac%AN?x`{1w<cX46>
z%~A{mUw%LB>Q$T6^78V4tSgnr*X-PkmORAeiC&{5dQUhBTxiHmSmQWoTS!F&9BOF0
z-1O?<WSUO@{$EM84@X9Prm84NJ?kyi3zc}XD(DNg3*c9L46GXh69B(l^V3?2PJV%A
zd1(UOR#D-XPLX+^*-9C~dGS(jcOQkFYc&YQXqrlTd`1RfoZ)BE4&->nAabe$($5{0
zP{**|)fV<UfxMaj^2@t0-7bY(lcZXEG6Hj{)7-9tn7EV%BmWy8>1P^KxZ2s@e)9>C
z#dH)xN}IL)bz@@?9mS0+O|BY~W||pmc*p%-I#ePVDG+|+AAvW1ZVd3)TAU6k`5)Mx
zQa)i}VR?Z8z}crkM+LMZ@)?oF@Ju7uFW@U=P@?OoJYu{DtgU&q4%*_m6&OH|=`mM0
zfBEt{2K)k$ig?As1;SVC2oBhOrnNEnI&hERY3%H-*hDAWRvbN%^Z8?&M48C@804{7
zNepq)3|ZkE-^kXCeFx1StDKXFj}r#xRo<6$XYT<Cb%|sVQttGUml;{f2DrGMvG7lE
zQtu4mDpKgJJx@SF!P%b~${>&4`1|*LzNSAxMpb$OpkSn$Bdt31176YRw9uHy{1y_j
z65a^STg>X`wYbKcexyNt$y(g;0sKj`2*cR_^N<Tf){GGj?|^`G@}m#S9WMmon}TRE
z_1{5Q)BdU(A-J<%vz%CeBYt17(#i{E&61Ln2S)s0>e`7sffg>3EEWjgE$Yu|c0}%*
zBCcF-d|aHbzb?je8jGCGRD9)vaLK=eyk3Ro`}WzS8XFnu<STvwNPp`rkqi2G9RoTh
z|3(C3s#;oh_YLC8dNBoTmHvNfCj7aI%mfi<ARSt<ix3b8Y4<h6Tq(nZ@dXD=p@~b&
zw|<RPdw4vi53OZaR#$Ieh)Ku{bYNHgjK7z;Za<T7E*YElO)d<qp9gpv^7I@bX7rBj
z+jsTno?)zFS5{Pf{kNrh=dxk{LJoTfHbNaxgMp(R{|Ppz>B)B*v2|jQxZE61>D!g@
z2wvV|Tm-CY7Jy+8?pK@VVHmhQ3z1&szh<GLp8pesKuo`e-sO6|yk(r>n$Z>49KTfZ
z6-$qvwTpBO`Ra#pnDXwpOnQ^zKy1toQ|s+B(yr6ZhN@}}#DDYiO~Qf2<eGd1vm<Gv
zXGX4dz+eP|L}Utj_FXV?i5SyUd?xXMOTPknIY(F2_Ihl3LX@<5+PdtH4*j)0%jq^<
zkDKgjl?WBx3&MG>=Ii#t42vy)tbaE_mHieXqtI2`@C7%n=CeBXFv246)UN?V3JOU!
z37jKA@{tWnn%Z+-&NP%cBmjO;IXKdHG|yx<i8P(YWr^F_@V}4z(Hba0$kkP@<9mZj
z{+tzns17XDQaYoE*r>a@94SVQj*oH6>~Ew509t6#*`k6QA-D+vWBl4ZJtG6;%Ss9a
zcf#M!l=^4&u|=tc4mf|nUcGjQzKK&RX7;<Vr~kYNY2NmoJ8wIfa(dfaTW@gxSaQxc
zZ`tQ*;c_bPjHz3I3>++K<>1ShGkHilA_?7`Cxxrno^fZg^Yg8sVR4O&+mRKE7dP1v
z*-z6pQO(7B`!+aBzH*I_eh4s`_s>FOi+b-|yKtk&3k$pB$3MoB|9R+H<g<WIUbHP_
z<SzUG;S_bbfFF19y_eee+Q=LfBaMg)7r$EXvF&0+V&gxVH5l?g-#QWu4SB_)MO&b@
zq?htl#*Fq7d7L0{v?o)DMQl%GBnd|#Sh8vUt9=I#{yK#T=TG~hl1)?F#hcf!{fWVx
z?t>qJ4#Qo(UgIg@49}6i4CeB4!k$|C6?+>TzSthqoaaSsAi&Nmw$(HkTUa=z8czmM
zhvhBoeCxj*3kv<IbYvNeH!Qhce72hFF3MI(Cw=9&d$RU?`*+;oQy!F!Voq;7`~+8T
zR`opA?p_<VC$%fYQT`U&Blzu7Y2&Pz&09s!7XRqzD0y!4C8yDoEjh=cGL#ld&M!?p
zmh3)YR=`RytrkSK6zHUT3=F$R+daRurFv#zZycIT3J&b#ZE=7)R5Jsy(D^lcNK($T
zh_u>{gy@;P4>D#+CP~dIKA^vEARd0}n3#=^5jbhm=1oyeFmwS$RCeJh2w46^A><!!
zwTjw|2!B8LlaFEcj&0T5^AWdM-*J;9$DBiOk1q|!SNKpB|6K3cS+k{jMqzKfupNd)
zIS8usIH>rD$T<qzaO#I#qFDo9Hw&AHAo3o2=G|+;Z`74!5@u7&4grH^#t?g^&7Uj4
zdXw(n`IvFq8_Tts0pVfAFTV!r3q9P)UuBu#d2N9Ot}FegOXRh^9EUxvXF5ZTiBO;_
z9+D!}gq$XFi9aS<sM#TRhw@R)2j`FG04J(A>BN>IqLhS~i<=S?lOn+^CE6{GGeJa`
z6YbSVA<8oe&_b%PG{8qCLLYfz9%Xr90^Tx{UeSFErruO87+#XMHIG#=8BWzT?Oo|3
z@?hrlU(L;>5b7hwJ#S-0g3XMH?k@xaT_mVG&G<o@RC8;pW#V@^v#iY5lJ<QXKe^yF
zDeMDyfP?S>6LWI!{d~C$=dTmDB!k~kgiU_BtqwZGgbEEAtLv`URF7(q63mNoaMRx7
zfJS7-PK}U=rloL%zWCV<jl)}qJ*yAo;3=KLZb07G4cC_EODV(AIi<JZij_i<&msDs
zhpvT64!IPFF;qtWAio7=Jsf~he=zsc8@~i@o0Enc!rQGir@aw4Tt`fk`=1)Ik$!rE
zue%ZxvlBwX4$<F!ew7;cnLKkzrW@T+){1oE=TGlflflMKts!(=yr7QMS}fwwmTsLb
z@<4i%D~7}-*iSJw3%ds)x`Pm^;SpjqcTfLmaWh0{AdXfy{z$Bj^kAv5c?g9(O>1my
zBq~VsT;V`^@5(Ncl<}v2zDrz$SQlinqo7KoyI4eH&(ut`3C8j#ik6XyLARdztfr)L
zwtWZWh$Pi>U+U^X;mdD_sEypz`YAlrK{xV7{E0#%CMcpHgTK!x)cDj187{aK8<IDN
zfEGma6kd}wET$g9{t1Fej~Vq#uUi<MBa@`;gpUcBn+t@-Re?s*h!lPha@?)y<!aoC
zC%DEiwe&ooFk3`pq41aEZ|{sB`uU1t6nXR_L|^c}@l_W{awmVT(MG3{=F}os!-v?d
zA(Sp%ASh(b(<Di4n|heff_Q@{Gu}5OsiRqhureWOFx71+b2jzuL_xZ#{zZASslPy_
zx~VrN>Re4tWXNeuO+pfB=9G#0|F0mr@qeN(*2OYKf{hO(MUdmj6i6I9%dsO9_dnDc
zb<QzMsoNkWwK;LW0lmqk|K8K%hdZ?Tx+{jW7O#wu_Zw@BH!+xq669>6>K5?G`MiPs
zD@WzE`!Zf|=Y}ZFRF_)S-L2hUJey@mTf6RZX|C@@Zsa$b^BNkg3>>(($fdXyz2(}n
ztVZkmf$I84VQaj)TK!6F&f`YC&XGB0Mr*a?eQZN{Jcicy<G@uT{dUS-i{Hy{nCL#?
z6tD`>Vtra`nF{N6d%I+3<gXpeBg&plICO!)HP~752-9=Dsr9(SV5r!-qZbLOzW>&g
zig;LiB9pl&cmvF`2237>22pXevZa%F1;|SgI`+yGOzUXqZf#T+^%207M2ChRrO=RY
z5NC1nf%{^%f3`bk2>alBK^5Bb=A}k#)de|y;N6iYGH<!jBk@UwwoJ{X=^5_$ifHG5
zwFVY$_;y=i%r*%Wrn#MNZ)}djMy4p~bn23#?lL|{y&}RiUj@obN#5ISf>%bQq96H=
zO(RBuLm77nWi~<x6)nj)^0)1wNADHpuZn%dTqPooYAtm-Bkb&IO-@$hgac@K^LyJ1
zUG}Az@cYXbpn8MObyf5AoQ_vZ=5Gqzidgfl2o+WoqGq+)i_bi~h2z2T@@tGYby<42
z0xm}?Li(?-Xvt=t+@O&TT*$X$Odue~bwfk3ORAV!JVLb<i)(IgpYE)ZjXQxThxa#f
zRRCAR3&tBKy`F2~I`<dc<NmP8|L$f*zm7(GCrkGR0XTpP++9WWwJ6tU%q?up@amTI
zWuFqD*`eBcG3=RD(9&z64>Qw_aqAd{+8i;Z5f_tGe{X5yUv@h=E$9bQn#UN4-7&BI
znmMqSqt-#=9OC2m%~A%QZaDfpxyh?B%HHvD$LaSOr{{Gh?_X&<+P~ZFL!9Zc_KHLE
zb^Scb+T1p3U*WV}-_UwW!XzQZDFARfl|Ixk%+OYQX{f5&!apVJuL;20V0Zb*CAn8u
zQ&aU3$h~<^DNqBZxr&iL6fyQrX~G^hI<G}k>IKRu+~HSwTOHuZ6>9X3w6MzSQql<V
z%ni*0IXAVvtq@OjL!1wlS7iBs_u6t;<C_lUj{LRfhQ=eN>--t05(O>J#k;Rtg#&I)
z$O~{VNz3PL@^6Cs4psl=uoIayR97<X&mXoV1cHfOEy*-!2@vtpZL%p<JN4XlOY2F>
zSzHs_8fEEH^q|S7g?H@3>x-+sx}JoWOH=frF;>yvrsBr0EVxY`ofw`}jnL;!dd>)m
z=cWn<lPyMjP_ID?my^4U^dW2o&o0{Bc|0#%zb&`#`9v2pFNyh!h<HWN9+Axcu8d9k
zw6S()W;kkZ#l^fU=03z$#%XY1-GNvZS?f_*82=41LeCeNwt>Z}9_gD6U96_G65a6u
z<HY?>-Ab48UZhQQ4iXi(jLHz(c5ig@j<-1Z0?;eUP2?Ai&$uf-Cs(6b)#8a}>4ks<
z9MCN|YYXXw2T~fAK&AlLHsT;)vF*S*xQ4rF5YkJko=<cjIneaI>ci1jSMu+WeOZFk
zZqIDG@3CR+U+Pi=V&Y3}G?x1uKF_7^&B{B)9(!5qnM-Z!!KEZOQk-P-V5My-&6YwN
z8gdG#@)>QrOreeP=hgi*cN<Y~2#TRM$<2}~%o9+~rO+-fU@(4~C-noh_pk4=efia4
zks~oRtn66QHsegQO87qf7fWOlj6Rf6nf!A5lZuCU2T5W>xG<vvxxWcb)+LZ6fJv2f
z8*SrNjEzdt>hnc3y;BP&#86x8D6wV^Wp>u#%~aalwn5?T7AKmz3^_lBs1?@*0kZ<;
zFY~v`Ec97IS$)!?++J2?6Wxvv$L7Q}N%4!P=Y{;)Qlgz>_z=}Gf|9E2gEJllP$A~Y
z;U~iLPZ)+m_)_2~RODA#a6CwM2^{rjxD410yZ$&UwG{2)YZtgTst+kj_y~~ZK{E{O
zujY5$V-eN2;*5L1>a8y5TcA`{7GWB<zhB&z3C<>lTv~h~%s@MVK3XjL5UCOgMv0MG
z--ImgEt3p+fLV-tDs?2+zc^rO8TB@UQ>XnemuoO`#(K<bXXo|KfvWs=)qST)XJ0qz
ztXBH?QW}!W_P9q2gN`VAD-MfQF>s_P_hY9DVKv;}>@^@uR7F8eGaid3tq-3$4!ePD
z5^Y>^cWfjdebCvtF}k&)abdJ3d+fK_lB?$A&biK7dH9=$CQ0~;+R+b&jp_cQOX(9I
zC-?Jq6&|{^-pNmI{6m3(TWkkBOm6^1Ilk{N@Y--&S5dMlaCn8fnn*}GV3HyE1zK*$
z)ba-UF-F?fBY&~%o`rb6&G9xIHn$dGhJub{p)i#WxQRY`OQ_+o=a+60*8A1HlGhDS
zmtE2656TD}e>zsAG)93_X3Tpc^}V#TvX%9nrgPcqy7LYn6ft-;IxA#Ryje<n%(?t1
zPp%onF?Wonu%2qg(ek?3{`=H6KvOW0KH5%HbBjdTxFy`Wyn~o&oMIIBc`C`SyltK2
zF3<LV`S$G}Jbw-bNWt5GF^&a$w>>*gj=O|W<oPkw+s1j5n{i&~#{j1|9}Ec(vz<hY
z03(U|MWq*3*AGd}iAuAP+iGwv2PH53aTj;%iht`5lVRU*VlX&JYuQ+OTX~aMr0nsE
z8xdqJpWe+;VI<C(?#9Nj$og~MUG=U3RrHbfb15{qtcxg_Ca>h)8wBPN6*waF<Epik
z%UK`Z*wqm$s`7iO?{&+YcU-*Iio!(qCgy!#bu@t*^X<N4-QA6};>P=@++&lt#pWH=
z10m1DcwRcbH&n@`H&z{eTK%*hnn8Em*PpbwLq3${=v(PlWb;a~=sRK`WS>dmzrR>*
zS9E#@7M4kyT4V`a1l*q{l8I&587a<gs>kRxG#HGf>)3@pJA5je*J{Ot_-F<8Ye6Lc
z^={>?93t!CU*?UjMk`D$2cyZQvVpxmX*miCC7%qVg@0dMrj+=Zt!HL<dZjm6<_w*!
zxKp5krFL8F`dv0|$FBm#Sc`&HFZx=X6}w=9bQ-g>Ir)N~y;MORwBK$2{@uG~)r!rN
z3VQy_ktWsb$YKVT#jJ~pQh;8mB`Q}{FQxpTp09yB<W$o=Y#l8B-NTI>kS*m<*K`=p
zxl2fZAFp(WNc1R&T)K{7VV)f+wHMA*ez<-dvAky+`lkszM;ADERU;3lfY4RMjNUjW
zoWC{EET!STcqbU4sL{XXP3V<g$etHvU)_xSjULEAV&H-^fcFddH><}g=W~b|l+Bv>
zb~66dUz#2rHWMgTCYaAbeXk8l;~gas$Uk^*f8l5kf=$q8)T1^Kr(QTO^midwif1wT
z#&&i)C)%AS9tv7uAHoipRR8_UhNj*>tdd-bFrfc9cQT`zgHk?zJx^L@W5<yHt!AJ3
zl1>GZ5A5vT%m&Z+`RmF>BlBEa%-k53K}eafDtmsJi!y~-sK0fr>B%+JpHeQrZcVc-
zS2NglruO4zSHjowW{oE%TTdFTKk5Hr9;<QVn!|q;W2AmWtvX*&ok}w$UDS}Gvn*&}
zXV9%o2lF>rv*y|P(KuD2e1tzCg|3q3Zwf78W8Mv^bp1FUD~uXCoEhjkF?M#eRe4ee
zc@o<pS9~2I48r-KzyCdg-j%MBOTESqnM;IB;Qia;OpAa2%V(v_54`bE8yA2NYmnk*
zN4?DQWGLB<jeP65YX+&1P6P&m`4*YC_OG-IxMa$`DozS^pNcHZeB?J6VDHfAQdWO<
zo?7|d4b4Ug+exIF+k^u=E7mA+!0du~;{MKX{6)-W;y}ouW5;d<XTtR>!85-L=R$>t
zc;yV^hZ_$em{WD3Imf&GqV<?#T3=4j#})|FEUG=R6yDyHci0xOmr(yivF<55J=Pxq
zS<RlodSW+(4RxXA%j)l@xzS@7^>D=AS-AdP7TDHlFBlpEw!$&g&V~hZsMEO)v!jQI
zF@$lhS-6$P(Z~9|hx}5pW3(d$5Te>CZy(*$U9!{aD!iID97-bKiVcs_BZ;HP*g%51
z-JAF<$L5cIgse3A=30BB_G;@JZRMs=s->@p&=>Q=ElF&~4xky4OKMNBmO8=<fY;EU
znHI1V;-h+P%=XGOTYfvEAJ|fRR%1`arTH1m{(7dMC%^W?M8!m_{0qJliy563ICv3n
z+@;9Murmdu`ABE9_2}r!b0<60A~rZR+;9%$D;cHa(|5WcjAdrdxS#2{_nK;s<_6{y
z=pEl=?XQPPL28V%Z+~gY$6OXOUln$<f8dRwEfzzv*`W1mr-p$wddj52(PG4zNE$D!
zd~J>OhUo>}2YQBT;O3wSyNN=b(qV(F5NUN(1siQz(bi+GJJx#!<$g?G<D4)#a%EBw
z?v_2wPbn{$;4ZNnTZ4=5NMHm!Lvy?mVOy>jjJ-{^6l^VM7b8W=5~qKS7bp#RuIk{W
zXhH-ZrMpVy$?{KjDvfjAq>gmk4|Y1YP6!HKrE$hyT?xFpWT(uqZ<<+3E0bCf2?1f5
z<#sfk-okNvXV!Xli3n-VAFA?*28v_*47=?-@$2vpQch0X3{26yc3nZ&8+y;vVP&mH
z*SpFh3c@2)f>u!aE1d&%)eCapg^vQ4D}Q((Tju15^^J871}Ae}58%?l4VuQuRsq&1
z{9~6;p+y%(UV|3lC}E*T;Uu6KG&cFcW3US$V;jo$E&JON6a20-(sI*RV1n7F_1Q&_
z9>!_mt%HBV;}0-$tSWeyG%`e-OQ8@g3w>Mm3cg9&#LqqX0^=1C$@U_oDrc*1L|&zv
zd^wgVPho%)3jPyA3fhZGt<?4#*aeIJ(LXd_=DVcYm?IG7q7T-wcH?|Pj<bNjZxJ*u
zVW>R`#-(0eMpwlSvI}^1Yo0svn(A%pWkd#aCPoT|<dt;&_UH{gVpZ29zCp=Uh%ggw
zLz8QXvR`F0JTSa;ZD0^;`0gMgG{IZE{+_FXL;Vx07958<-6x<xA+G&8*wV9>3d8PG
zR(~NqI~#{ZNF%>ZxD-(?jiMaK)93-+yK@~FRX8>$Ge*+J8=Q|7pMP%E0+6NRud^4a
zmnU15tA}pNlGy%JhGt6^9I${;YrKeFiAuy9B?A+P^A4!pgU+xZ5@lU>1cXAOW<Xv$
zFgJ-&k4k<5|1o5>8equB&gMN|?{q@D+aCt&tS*AXHdo=959>3aZ7r{?MQwRLXm!d$
zZ8%St@Q4HpdCsnRP_x!6*~Gs&>Dr*WVr2w#M;{Ja(9}Qhpumn4e2NHL75=zjRV?p%
z^Nx+A<d~SW>KI;nCW_-!_vy!d%v!xaDPq6G2<iuwz>XLCrf8-uhtQjjagx~U-{aF#
z79z}ko3QZvY-&%b*{cRaRbo(UBGGW%mT>f1h=a?g-+ve2H0Z)ET;RED^%LY9)1E(P
zHtHbX&TxR+M^tbC-4g-p7{x`*ZCmKnSAK_X>{wUF%wI<w|6mrYZ>uU{zNFII8&HRL
z(yeUdjp5{vxWXx;vIHtOHC>Mk=NBRfS`8;A!;{N}O8uSt@T6(2!$Co=l^J5fNb{^C
zLtVZBp)6R<ut?3v-N#A=@E=T9T)!v?$%9VxGQx@?_<3hF6I)?QibH4kERs#fiXIg`
z5=rhjSiwj)`n$)&NwxZL>9KRH+p|s+4Leyd-ZEd?%OdX{V&(R;u-`2|kBAG-s1N;0
z9^8aG+}e{fNrps2bFgAIMw>o9bj7YLJn>Ap6u$pY!lce2YN7H=4JUl-l6}>%HKe<9
zLOnbdCV3}sS#zZLyalyez>glb2m@_J>>TL`?h@W`$rbP8tuG`AgNF)aYamsCaxjuO
zU8FM~iI7gj??;O6pv)8$Ral5HJh3tTUkOuAY&xg3$p3<mE2ayHMA>1mJ$0+i>Hh#;
CMU|ug

diff --git a/articles/Modifying-Models.html b/articles/Modifying-Models.html
index f4751fd15..fda354e94 100644
--- a/articles/Modifying-Models.html
+++ b/articles/Modifying-Models.html
@@ -1432,7 +1432,7 @@ <h3 id="creating-more-complex-model-modification-functions">Creating more comple
 <span><span class="co">#&gt;         cp &lt;- center/v</span></span>
 <span><span class="co">#&gt;     })</span></span>
 <span><span class="co">#&gt; }</span></span>
-<span><span class="co">#&gt; &lt;environment: 0x55b14af6e288&gt;</span></span></code></pre></div>
+<span><span class="co">#&gt; &lt;environment: 0x55ff06d78570&gt;</span></span></code></pre></div>
 <p>Note that these functions do not modify the <code>ini({})</code>
 block. You may have to modify the ini block first to make it a valid
 <code>rxode2</code>/<code>nlmixr2</code> model.</p>
diff --git a/articles/rxode2-cmt.html b/articles/rxode2-cmt.html
index 6fb5a9baf..5699092fe 100644
--- a/articles/rxode2-cmt.html
+++ b/articles/rxode2-cmt.html
@@ -381,8 +381,8 @@ <h2 id="how-rxode2-assigns-compartment-numbers">How rxode2 assigns compartment n
 <pre><code><span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span></code></pre>
 <div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/summary.html" class="external-link">summary</a></span><span class="op">(</span><span class="va">pbpk</span><span class="op">)</span></span></code></pre></div>
-<pre><code><span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_a688fd3d12f8698d15feb47597e1364f</span> model (<span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
-<span><span class="co">#&gt; DLL: /tmp/RtmpSPRppR/rxode2/rx_a688fd3d12f8698d15feb47597e1364f__.rxd/rx_a688fd3d12f8698d15feb47597e1364f_.so</span></span>
+<pre><code><span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_0a5b43d0b1e7bff7f47de827bf0daaa5</span> model (<span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; DLL: /tmp/RtmpPXkf2x/rxode2/rx_0a5b43d0b1e7bff7f47de827bf0daaa5__.rxd/rx_0a5b43d0b1e7bff7f47de827bf0daaa5_.so</span></span>
 <span><span class="co">#&gt; NULL</span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; Calculated Variables:</span></span>
@@ -732,7 +732,7 @@ <h2 id="appending-compartments-to-the-model">Appending compartments to the model
 <div class="sourceCode" id="cb15"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">ode.1c.ka</span><span class="op">$</span><span class="va">simulationModel</span></span></code></pre></div>
 <pre><code><span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span></code></pre>
-<pre><code><span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_31957e383dd61d7e82e22d46f409ad8a</span> model (<span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<pre><code><span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_1e03154980d52540647eaa8bacb10b0f</span> model (<span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$state</span>: depot, center</span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$stateExtra</span>: eff</span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$params</span>: V, KA, CL</span></span>
diff --git a/articles/rxode2-speed.html b/articles/rxode2-speed.html
index 0ae72650d..b60b6c5ae 100644
--- a/articles/rxode2-speed.html
+++ b/articles/rxode2-speed.html
@@ -264,8 +264,8 @@ <h2 id="a-note-about-the-speed-of-the-functional-form-for-rxode2">A note about t
 output. You can see the differences below:</p>
 <div class="sourceCode" id="cb5"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/summary.html" class="external-link">summary</a></span><span class="op">(</span><span class="va">mod1</span><span class="op">$</span><span class="va">simulationModel</span><span class="op">)</span></span>
-<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_4e26e9d8f25df5e07a0521b7b177da87</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
-<span><span class="co">#&gt; DLL: /tmp/RtmpSPRppR/rxode2/rx_4e26e9d8f25df5e07a0521b7b177da87__.rxd/rx_4e26e9d8f25df5e07a0521b7b177da87_.so</span></span>
+<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_cb1fa822f5309328f76b6ae6134387bc</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; DLL: /tmp/RtmpPXkf2x/rxode2/rx_cb1fa822f5309328f76b6ae6134387bc__.rxd/rx_cb1fa822f5309328f76b6ae6134387bc_.so</span></span>
 <span><span class="co">#&gt; NULL</span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; Calculated Variables:</span></span>
@@ -284,8 +284,8 @@ <h2 id="a-note-about-the-speed-of-the-functional-form-for-rxode2">A note about t
 <div class="sourceCode" id="cb6"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/summary.html" class="external-link">summary</a></span><span class="op">(</span><span class="va">mod1</span><span class="op">$</span><span class="va">simulationIniModel</span><span class="op">)</span></span>
 <span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span>
-<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_5c7a85ad8a3c6825cf7c3e1131132e2a</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
-<span><span class="co">#&gt; DLL: /tmp/RtmpSPRppR/rxode2/rx_5c7a85ad8a3c6825cf7c3e1131132e2a__.rxd/rx_5c7a85ad8a3c6825cf7c3e1131132e2a_.so</span></span>
+<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_f8478a8ad7138c076241f845fcf72d6c</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; DLL: /tmp/RtmpPXkf2x/rxode2/rx_f8478a8ad7138c076241f845fcf72d6c__.rxd/rx_f8478a8ad7138c076241f845fcf72d6c_.so</span></span>
 <span><span class="co">#&gt; NULL</span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; Calculated Variables:</span></span>
@@ -312,8 +312,8 @@ <h2 id="a-note-about-the-speed-of-the-functional-form-for-rxode2">A note about t
 <div class="sourceCode" id="cb7"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/summary.html" class="external-link">summary</a></span><span class="op">(</span><span class="va">mod2f</span><span class="op">$</span><span class="va">simulationModel</span><span class="op">)</span></span>
 <span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span>
-<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_ea7cfb51903394d4a9db941ee311eca2</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
-<span><span class="co">#&gt; DLL: /tmp/RtmpSPRppR/rxode2/rx_ea7cfb51903394d4a9db941ee311eca2__.rxd/rx_ea7cfb51903394d4a9db941ee311eca2_.so</span></span>
+<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_881bb97ef0dcf691ed7a5d08770c2985</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; DLL: /tmp/RtmpPXkf2x/rxode2/rx_881bb97ef0dcf691ed7a5d08770c2985__.rxd/rx_881bb97ef0dcf691ed7a5d08770c2985_.so</span></span>
 <span><span class="co">#&gt; NULL</span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; Calculated Variables:</span></span>
@@ -370,8 +370,8 @@ <h2 id="a-note-about-the-speed-of-the-functional-form-for-rxode2">A note about t
 <div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/summary.html" class="external-link">summary</a></span><span class="op">(</span><span class="va">mod2f</span><span class="op">$</span><span class="va">simulationIniModel</span><span class="op">)</span></span>
 <span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span>
-<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_d7291e04a9631f78bdda73dc6745fe46</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
-<span><span class="co">#&gt; DLL: /tmp/RtmpSPRppR/rxode2/rx_d7291e04a9631f78bdda73dc6745fe46__.rxd/rx_d7291e04a9631f78bdda73dc6745fe46_.so</span></span>
+<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_a19506ad281e7f985f0efe169504bf02</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; DLL: /tmp/RtmpPXkf2x/rxode2/rx_a19506ad281e7f985f0efe169504bf02__.rxd/rx_a19506ad281e7f985f0efe169504bf02_.so</span></span>
 <span><span class="co">#&gt; NULL</span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; Calculated Variables:</span></span>
@@ -558,10 +558,10 @@ <h3 id="compare-the-times-between-all-the-methods">Compare the times between all
 <span><span class="fu"><a href="https://rdrr.io/r/base/print.html" class="external-link">print</a></span><span class="op">(</span><span class="va">bench</span><span class="op">)</span></span>
 <span><span class="co">#&gt; Unit: milliseconds</span></span>
 <span><span class="co">#&gt;               expr       min        lq      mean    median        uq       max</span></span>
-<span><span class="co">#&gt;           runFor() 265.60642 270.91812 280.10508 273.27375 276.22522 390.10634</span></span>
-<span><span class="co">#&gt;        runSapply() 264.82305 271.37265 280.13567 273.25948 277.09826 440.15521</span></span>
-<span><span class="co">#&gt;  runSingleThread()  27.71541  27.86995  28.73338  27.96453  28.08604  48.11461</span></span>
-<span><span class="co">#&gt;       run2Thread()  16.31607  16.43347  17.25248  16.52140  16.67760  27.95446</span></span>
+<span><span class="co">#&gt;           runFor() 263.75899 267.88252 276.54611 269.91252 273.66833 391.77390</span></span>
+<span><span class="co">#&gt;        runSapply() 262.40141 268.24234 276.38984 270.40109 272.75447 434.24999</span></span>
+<span><span class="co">#&gt;  runSingleThread()  27.45217  27.61226  27.94130  27.69391  27.96168  37.54928</span></span>
+<span><span class="co">#&gt;       run2Thread()  16.17052  16.32538  17.99204  16.45466  16.85388  27.76178</span></span>
 <span><span class="co">#&gt;  neval</span></span>
 <span><span class="co">#&gt;    100</span></span>
 <span><span class="co">#&gt;    100</span></span>
@@ -588,10 +588,10 @@ <h3 id="compare-the-times-between-all-the-methods">Compare the times between all
 <span><span class="fu"><a href="https://rdrr.io/r/base/print.html" class="external-link">print</a></span><span class="op">(</span><span class="va">bench</span><span class="op">)</span></span>
 <span><span class="co">#&gt; Unit: milliseconds</span></span>
 <span><span class="co">#&gt;          expr      min       lq     mean   median       uq      max neval</span></span>
-<span><span class="co">#&gt;  runThread(1) 27.73454 27.90294 29.28469 28.52594 29.34674 48.13157   100</span></span>
-<span><span class="co">#&gt;  runThread(2) 16.29712 16.53740 18.44910 17.22503 17.44130 27.90322   100</span></span>
-<span><span class="co">#&gt;  runThread(3) 17.40113 19.47790 19.87893 19.73595 20.33017 25.56011   100</span></span>
-<span><span class="co">#&gt;  runThread(4) 15.93134 16.07433 18.29293 16.73419 16.86182 38.19777   100</span></span></code></pre></div>
+<span><span class="co">#&gt;  runThread(1) 27.45592 27.64329 29.31929 28.28103 29.36378 48.39516   100</span></span>
+<span><span class="co">#&gt;  runThread(2) 16.17788 16.39189 18.17140 17.07855 17.39019 28.93695   100</span></span>
+<span><span class="co">#&gt;  runThread(3) 17.42292 19.35593 20.04593 19.99469 20.21361 27.37969   100</span></span>
+<span><span class="co">#&gt;  runThread(4) 15.80004 16.03155 18.46875 16.62245 16.95347 40.09879   100</span></span></code></pre></div>
 <div class="sourceCode" id="cb19"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/autoplot.html" class="external-link">autoplot</a></span><span class="op">(</span><span class="va">bench</span><span class="op">)</span></span></code></pre></div>
 <p><img src="rxode2-speed_files/figure-html/unnamed-chunk-12-1.png" width="700"></p>
@@ -682,8 +682,8 @@ <h2 id="increasing-speed-with-compiler-options">Increasing speed with compiler o
 <span><span class="fu"><a href="https://rdrr.io/r/base/print.html" class="external-link">print</a></span><span class="op">(</span><span class="va">bench</span><span class="op">)</span></span>
 <span><span class="co">#&gt; Unit: milliseconds</span></span>
 <span><span class="co">#&gt;             expr      min       lq     mean   median       uq      max neval</span></span>
-<span><span class="co">#&gt;  standardCompile 176.6202 183.5591 219.5272 191.7218 238.6262 350.7461   100</span></span>
-<span><span class="co">#&gt;      fastCompile 176.9396 183.0521 219.0194 187.1348 235.6830 351.4314   100</span></span></code></pre></div>
+<span><span class="co">#&gt;  standardCompile 173.3969 181.5005 218.6953 188.4703 239.3224 375.4217   100</span></span>
+<span><span class="co">#&gt;      fastCompile 174.1782 179.6236 214.7402 182.9570 230.5892 352.4374   100</span></span></code></pre></div>
 <div class="sourceCode" id="cb23"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span></span>
 <span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/autoplot.html" class="external-link">autoplot</a></span><span class="op">(</span><span class="va">bench</span><span class="op">)</span></span></code></pre></div>
@@ -779,7 +779,7 @@ <h2 id="a-real-life-example">A real life example<a class="anchor" aria-label="an
 <span><span class="va">Stoplapply</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/Sys.time.html" class="external-link">Sys.time</a></span><span class="op">(</span><span class="op">)</span></span>
 <span></span>
 <span><span class="fu"><a href="https://rdrr.io/r/base/print.html" class="external-link">print</a></span><span class="op">(</span><span class="va">Stoplapply</span> <span class="op">-</span> <span class="va">Startlapply</span><span class="op">)</span></span>
-<span><span class="co">#&gt; Time difference of 11.31994 secs</span></span></code></pre></div>
+<span><span class="co">#&gt; Time difference of 12.08713 secs</span></span></code></pre></div>
 <p>By applying some of the new parallel solving concepts you can simply
 run the same simulation both with less code and faster:</p>
 <div class="sourceCode" id="cb26"><pre class="downlit sourceCode r">
@@ -821,7 +821,7 @@ <h2 id="a-real-life-example">A real life example<a class="anchor" aria-label="an
 <span><span class="va">res</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/rxSolve.html">rxSolve</a></span><span class="op">(</span><span class="va">rx</span>, <span class="va">ev</span>, omega<span class="op">=</span><span class="va">omega</span>, returnType<span class="op">=</span><span class="st">"data.table"</span><span class="op">)</span></span>
 <span><span class="va">endParallel</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/Sys.time.html" class="external-link">Sys.time</a></span><span class="op">(</span><span class="op">)</span></span>
 <span><span class="fu"><a href="https://rdrr.io/r/base/print.html" class="external-link">print</a></span><span class="op">(</span><span class="va">endParallel</span> <span class="op">-</span> <span class="va">startParallel</span><span class="op">)</span></span>
-<span><span class="co">#&gt; Time difference of 0.1070809 secs</span></span></code></pre></div>
+<span><span class="co">#&gt; Time difference of 0.1223662 secs</span></span></code></pre></div>
 <p>You can see a striking time difference between the two methods; A few
 things to keep in mind:</p>
 <ul>
diff --git a/articles/rxode2-speed_files/figure-html/unnamed-chunk-10-1.png b/articles/rxode2-speed_files/figure-html/unnamed-chunk-10-1.png
index 8ed1c9bf75dbe61dbfea1c2f142ff3c1105c4984..8d796fc192b42a737f97510929be7665cd4990cb 100644
GIT binary patch
literal 36489
zcmeFa2UJw|`Ywz;vBnr7DoTu_C{k1eL<BU(G9pD$5Jtd)AZ0*6dNoE(R73{_0TEO{
zr7FE+VnI5B)S)R&dI#zEc?Ufw_kLyFweC9iuJwI?*Evy{J$wK5ue{}Xp7-@nHI;n}
z<}RPh$H%vTx&PP0e0;O)`1od={AM<OqBFf%7r)IurL^x?{ElD4yHi~8+qdWTYgzH}
zt-M12pXSlnbB2%aCqCw{yN=l1Yp-@p_3!hS=q{*a+H?;6@%GYBYxmEr*Eqs5Xx&nh
z{Qlmm<`cSiAIf`riS!mpN~P&v7YY3)^ifdgBZDW&d))WVvCf#a`e)0<-+p`R8p9@f
zXuE5Wqkr0&F8$`@(gQVJ8TZGJXP877)+)3mq-Xr{{Fo?>l+P_xUA=QU{fY0G;&d7{
zpPR_qx>@vxCx`!|KPXABn#EUQJ~1}jDE{DQ>raKN<sAoef15S$u%4b?r)8ken^8%o
zmxSK^wMt&MlYTD$?9l(|)-MlkUYaN1mGVRRr(~NB-+X^LVJ5$r_U+{4<Sm*r^Ge>o
z4^j<K=u{OsoMO{alw#Y(VCv}T;EE$nSAS|Li>h@HGMOnsZ@2T#_uae~KeA!kfBoRU
z=Hq{HJM5gN5q|8JeP7d^l=;|QZc^A?PJhQ56<hXZ2{$)*s^&L(>)k)~{8%T4cAqw*
zZr11D!=lo+3%Ptg$!Yj&j!<sE<XAH{F+MHaGEHLVk{E-$Y+O-kTAemj!wNB`@0A{W
zF|Y7dv8rNPWpZ<XtAdbo|Dk1})9}eLN&1dNTK{1q`;?A&CC=8ZTN^G3=&sjf(ibBz
zo~=D2ZBfF^$^`R(VimcF*mR@rdH6_e!^~7Z*B%T(|F7R>F**LWHL`P7*_|-rZj4Gj
z92ret*M?2_d@Z5IAuKkvEi%R6h1;~2+6-5Yer2qYR=SgY(t#Pc>g+8yfjWs0)c{@X
zRnn%%=3e?`@AV&V=^quUk*TqJbnp(1GZt^qEg)%_FDlj*Tdcx7JlD?i8#{dBc6ALs
zBV=0F4Y|&{j0)$WYNfX2!YAm)PkxxL>y~j>-qD5=r2bf?!oZKsVZ6MuNP;c%n*e>j
z6`wP%2}tBW+;fe?;F@PR8yjnk+qIWQu;Oc7Mj2UhU8$ZP3L{TUrRdN*aohxcnsn_q
z;~Wm;o=vE>*(kQ+HO<A5?=g#*4x9&T%$WYs(Mx-GU!HFiAJOuw3C|n`axmKHrp2jv
zyy^qB{hZW^kv6|LaeT#1YB9c&x5h5zd<*Bn?knq<HFycp<B8|2yvW)XicFJuR<e=*
zTJYl3>SSxBu9`GWuHmac76&($={mAwO6a0z>!$0T2upAp?se*aw~xuWFk_Ba<>swY
zmJM%<dK<DlWms%n#b@$!-K=<r1q&7gH!fVcJvOo6zGJ7BwoBt3O=){&x`4g9Zr0h0
z{x+2xs#9!LqeJ(+yt;H__VLB|$jwb4P}$hnIC`kYQJ9H+`(P1SK()nqadvE1`ly#z
zfb(F*77d!JXIF8x`@$XFFN%643wEf#Ie;JXY2Zy`-_b&0eCwDYD*nE<way8-OKFK&
za+t&wOXps4>a+^j=qzvX@l|vQ-RmZ%t~Bf2J|P|L^vQ|Q;=)3;H?*zYBK+M1PONNC
zZi}p7CscD7R?YcRS7~9_hYdE0yNP995puaJpy#x?Kw;E$o$CF2v*z&!85Z1MCTppG
zW823+4zDkLa!j=8bG?UNlu5PAXh%!ogWa5lHoZxQrd%ISmU)_#PV)JRy39i@tDL(3
zy1qh2y!pPPnL?jWL2RPx^JBR(={@fK?G=)fy%LkrmGNCp&2m~vmS-XpKVJP|?xnm#
zgZz)w);HW%SIN4&`80<yIcPrVH9p9dz}!k$!He^~DBb@eXYcc~?;hT>%-#?d5g=o(
zyUOMBhaF?xFDqOp$0ZGN@5Y*xMaHJ3r@LH?($Uk)IFzq0cJ=kC*0wfvQGQl+v#RTe
zdJV?zG|H{jWwRcry<KK_?^a)dqN1w8us*GTvQ{U5OZUq~1%E7F)7zUPVF>5BOwPtM
zeR90ZyXO3foOgwVH~(BB+>|98kc;Ksa_ZF|f)=k9w^YWNBq`w_xiO~eN^{^uFD_Ep
zM)oGAFJJDlW|Z#Jr%%aI-nVb^Ic`&N#=iCG<6ghov2~MeuH#Wfo3XY-#?7H&VL6)V
zy>}Q{5eYTM`VHHS&z^njE>lB$z~HgJjj=_5j%~Hn_{cy`f$Kz{;JFvm_Vz?R^7Zv~
z>T;~j+3@%t=S+@Q<8X68f#P_vVq;3}cx0XVyyz2u{WjKpMKPDmjrI^0v3ZbdUSQws
z?~UV=aJ>Y3>V=3ycl4q2T{ZFo$ItB=taY92@8HUGRK_Xc2xi_`Y1RGLoMzn5jB_+%
zX{mH&yis}i`B({dU#?_<{)fLW$qqMovKE-lyY!2S!f->bpNs@7WsyPTxtwS9H{=Z3
z%by23_S3**o9}s6#29)fZIQl=x0Y$XXQdKzF5w+*z{ye9NpqL+Ha$Q6MVQNI$=?#P
zw|2;+=x1@NFj~x=;TG44vExoXua`*~ABm}nx2>@^yw;qAy>9h#flR~S^CWVo%@HhY
zm7=&tt}}5Btt`baIyMz+h3}pFWw639z&g}1DsAvUbZFX(kFT#4hiOI)*!7zX!9Itp
zr?Kr7R^;8?yk@A{Ru4x$N%YvWpJvRS{}!L#lW^&|CU#>>WK>i^5Yw}_{ItiJq9CUI
z<oHnWqkSINmrIFc3R)essj|q_5ZdSL?7YgZ?gDnz3?DgLv&-|BacI$Gd*7}p8xg!N
zq^Q6w-Jblc$683y#VIC}$#rg$49Hrld7QT+bG#)Aa(yKGTqg=9voL1=jA6fw4Yu~-
zk1=_Mxpz0)ccl$+81fGNZ{dUPvg0e(DUMbY^cE>e1mmE3vCOh=Z&0`FOll6<=dtQc
z<Z?0scib?agMR+oEKWtHbS4{ceQz*W-1QZgFJIOeO_bKH@5vM@DAClbZcp`Gr<{#<
z#8G>EyK;T(o?@%>Si`)`h-(8+6T}J(aio=Hs&C^iVHJZul7<E_1H&EWu6{}qX<J`j
zUSJYGE|Bqg51xPDpsgv-&-=<MhxLo(`_;mvoX2`HWxG=Qww!)*y(syq`eTFE_fLH`
zo&0<G%K?}j{WsT_!R$0>2Q-W8B<q(yJL$9K%mIE0-SDMv9JYT7yl7ePF4M%Vov`5<
ze^$Jxx1Ni=%j9Tn^Q$GQxrg?xjC&VZ^4Xc2?mQYDS8^=X&f>n^$3GAHZr#w=T%g*V
z?@wpa5Yw0c?BvS>&VOhkt{52^d0$!@5@Il}C>m{A5iP2dyp>s$J|3#4wNW$jV07!M
z&Z?xo<sN#lSuiPnqiYm<UM)!)>&=OQMPg)0b4@)h>ffCCRQQ0&Sz(sAT2sgRDD7nL
zK}8%9t2fIG8>|9Gt7f~l20nPO+B|JQL)X6X4pvxrU|?X`wsR+L?C5=azV*G)#ry?0
zjb6+btYBIrL6|t%#@p&e*rm$f4<5HJIV#)w<dJ0&Q#cq4tM&ZUo>k8M?~dbUNsV4C
z%lC)&#U<5#GKos_Vl3NX*R4In$MStjNOK?dau$rooz=T9>-T?pS7DL4vKTH@6!SMZ
zUOVY;5}C9P3zGG@r`~VYCazfuJ6~~f?D~(Je<t^)tgP%M#`0o%u4@!l9A=u~SxzIx
z0r`QuF8Y>O4wPsbV$~|lQaUW(?GZTBUH8`p{%iL5ao;1M@w0J(Nl&pxvdm~i3UMv6
zqy3K+{ju&%*lId#hOIhHM~z**JAZ<8N2L@FouOZ(=F{afUJgh(eeND>)y&9#EueQV
zbCshgv-iykbA4>Zr1t2X=p6BsAlQ7P+VpYlm-(lj--YGHf3cqtoaw*5y3|L?M5C!7
zAOJs!HLq2mgX}hKMiE{dUg!P$_on>;Gu-30mOg`tp*SNf?Lt3}HD=O`NeNDx;}Fe|
zkn5NQ=O|kv9`iE)$4x(Va1$ir%$d>JY3?3L>FK;E(IEex)?95?F}`A&F+SK*ljgXA
zk4(kim*z>Dl!lAyrfy>vC0RAkU$%vj<}e_Vw$Q!I+{vl0DK7{9^XT>=8!p?MHW8*T
z*cy*0z~S-x@bZcwUYBA(e$it$g(g2|Cbfk4MZa7q7j!!F2HQn`d96MR+p@K_b&GZD
zdwwA+6VG*Ke}tSoj~9@x@H1Caaj(7BX&(Qpvg-|*33<LdK|S}+yB_}2cS27?UA^9X
zR_J}VgahudUNB-cK^b&R3-2L5363%`F|jH?d4=au)_CmTWO)iF7;zpT86+ph<h*;d
zPg6fFLOy`Z6*;QT>XhcI8zc9#*=!@0x44ewtHt|*u;J_9xaZIf*4)_kFm^mgZ=%<@
ztfeSeY2mgyH=gf;*WYsT@89p)R2l_$%;Oh*OIO3O(R1?lmfR4O-Qs=RGHZFj%MI!w
z3?{{#8be0LS613g1&Ai|?S$+bZXI_XC=oMTvHkqn&cyl~(mGLx+1I-sVJVukH%8_D
z{_PUWj<~WW1mMi*)(P|ZH8NgzHflC?B-G|3q)j_ikn#Z0Lf?lgLXq)i>ZQqcJs<f;
z8(fN_&*U7a6=FX}Ow$AltDNC7u3hDV@X^pie(>#3#^iYF=a1KzoZ-G^3K`hxZ<6my
znUt<ud8QUyP1?iqvT=jDh-r&-3E$lEAFBe6BoB6QM_Cs_4LNZl);P=TUfW%&J9=L6
zm&we&VAYhX)h4@q`SL9`?d9Zu$g5mlB$R&0#MF3qE4yRZv^~!L?{9^-&a?F+i^yS+
zoAWcNKI5&!VHC&@RT-BlaUR{+9{O5u)9xz^bt0Djs_rSQs>12-s?AWgHbwrz)R>iO
znuzm416<L=v|heN|E>1q+qUAzs!y#cC%AoWaq3ayx)&Gx#+Eye^cz_?Czw@pg8CP?
zm;P|=H7|;RdAFp%;JRdf`6nsqxYYykgHF=%YZHf?{FFSsoe$UV3pdR3jjN2;iZN(h
z<obEHuHA=Ax&Cr88#Thu+T;rFlTry#kY|>b_nyF+$$9@&omEw8=ftXYHof<wF>)W1
zD)aYG4n?H9(1{t?AEA-sCC06@3P1MzXA0?E2Ez2>%DKL+mayjrrQuq`Evl}SABT#`
zcYU}d&{S<(TP-@!-&SV+$TdKUol{u;^Yr+wE9GnsBK+%Z%*oN2YkFQ;(YXh%FUor1
z!tyF4RoiVl+b4p4mZ}NJ3RPVhDOT&2_dv;8yv;cr2SsYox9YaGpXiea=d}_UW4${@
zK0TrUw9miG$<WhdTg^=emZQQ7MKfJde(qKa9d%?sELg^P>+J?voVCnS!NS4zikK3%
z`!r=%m&cp&QvP$DNI)wbyV6+I=}#5B(vwWjo|8tPgCMy!3vY6uji8y+H)Pl8jTCtk
z>QbWHW*rAVm5L1GX;>Hrr&E;SI+-_dYXtVMOwUyw8}pd1LrZXOfTE)RcI#~TI!Z?j
zF&d>P-Lz=rKcZUE4DxPekvcIQ@!UwOh}Wk!&l3>OP3cPa3s&+DIVKt@ou!SBi*axM
z(VGI-UORpMa7F9ciCu7l4ukDJ3eF?p&*$_a6t-;kw{2?jGgm>}L(e4xlGNmD;_F;n
zjo59t%yqo%@XD3j&%cG|@viq!RKSXAKh7Lx3ORM}qHS^T{tgNYRE#)rQCQa%@^u|Z
zGGg2kDr=uy@Y<?!pk=~9>3y6@nZdLunbs#;rKF^S-O9$9!AA4~Ze_J|k4&G|rEzxm
z%c2Df{7Qp1S2;J_+82#<3#q|o6aQnORoYk>;WH_9zs;j<n=tGYV_1-ngvhJ1vhp5c
zNlK22B3KDU`Vw%<valiAX}YHixg06w?MGQGKNvn~)#as$0qMuuy7FXda}_7*6`Qbl
zO3|&l+b2HP_hn2DO{#dTk}txYadb~yj4Fo7KU~mYx#5So%M_eGYqKJd2gSUfIK4Zv
zghJ1S2N5(o40Za$N+k^6MY7q@HPP8czO5l(ayTG{mU*z!M0UKt*eYwScaV8)I_+{_
z_{qoTb1_Af`nV5P_O}!Z!yX=6DA)eX-BJN3INs@wfAaIcv{zVfhig~1ObhwYS>xE{
ze@#63Ny)9alECM+9Rzi#WJ;-oT933mZhkRi(ZPcI^2L!lDTi;SlKJy6X>U#nVJ1=L
zqLCoAs<v`i@#$vK8l%Vlm9dI>%064xPIfPvRGVvyZF&Fs>C7gfi4Q`*Hyt%ajGbNn
z{IrsJ;z68oz42DfzDp8L53~{yr@BaM>e`>FxM)kGG@D){riSB)q$D7yG5r~Ka~^h!
zE4#mq|Jv)Z24Tf4@&<<<?QIO0XbA}2UZs54Sg9L%njxYxwnF6d#$6^6-*+Or<uCJq
zr^pJO2oQ7rRUh_%V#lyDwxCwUkJ}=Z>s{PE)8E<9j@#?8N@s4?oCV8oheakEyGMvV
zS%ovoav7@^lJR@??0JIa=wOExGL9xB6$c-y_-7ye8G~gk?lyZqF%mVoooPAT=xtb&
zYR}ZTv;;<z@~MP(epND;Q(O@mZ2_J@iONx-&UInh>+U(UKIQq#W@P8=(x~{zNO8iS
zR@zqGf*}J|w0Tn+{Q4GlW$fWiv6%TPr@p-5g2}N0{;0A9^IAcq)lU&kAcqRTP8G&c
z#sAbp4rU?#PkiIBB8pqOxYB*}Tf4Gqn)S3R3Gk7;$wOSB*%VrS7ryoJamS9hVFXmZ
zF#86DfxCwP7D&H|praA)lH1_?8mVBQSbtshFDnsMCLvIkgz3NMGSZS4zhH&5DA^xN
zq(<hkZF%?Q<>*B@;l>i(ccPqvOPr&UTSF1u=s(hNRF}hPh&lJ+?<(suouvLEC02FU
z9L3MOjO<~6a#lIE-@*o1TRT#kpursqkIB!*Wn~}j(}(lSn~?lwBQmxs)iRXgN}sEH
zK`=*7m+Ryxzen|H4|%w|9X-EZlXp>Ufz?*!w5N}DnpAsFb|Fv7d-dlM&+u0F(eW-<
z*Nehs54Tvim5RAe4!H(oOpazq*lxfg7XCb~i$63zUK)nd*#&5T!N*sBQW__D{!<|@
z;)YG{!&xwmHi6BPy78TvV4k)ZJcEi{N30;?411W5zV?^`ayc!S?%CK!v_AtcE)*GP
zWAA9XqBwSP-NCyzM>|s_u+9ep6r7{6&W#8z^KP!*t-Zib&bsw>z<9f15)4!?o~56E
zPZkmTxrK<n1RdL-57!G#mK#Snj}GS5{dG~&zPCYCFWt#qNJuaFyb<9+CtqF;21L^~
zIg@?KEV(752ta7Sy=gOM-+qp4kMD$u#(E7_<?QfxOH_B<qJufsS5UC}ieOQz%MxLg
zT!KrlNu+NZ#cRygiFtl>R~+8j!xC`?LrZGz2yD5TbuBiL9O7<M4i61NrfO_r!m3I@
zP|<=!6cOpOcie|BE*qyRdg>4NHflxdGkrGx<{pLI1O`tl@!ZKIr3>zn2-1U@Is*T%
zL_(Ohgmiwy^v3s5dKm)>8jEzx|8CEioM=Ii`Nl)2w)JcaTw70l=3^MZb+cIeC_^_I
zo~QTum#`0n4!&smsqT{4L8L1A+v=}VYE>MfY+?}rXmkw)*#LV^K73-`54+**wMcQS
zdkTM}FBLtwvdR}T*8u!hkIh1I8&{_5hg7M77Ie3Di3tpi&vt8Ljy}#z#mOrxr+~g^
zagtffO)miIvjqGQn-~EM3xQ@;?o#$LN;->rCq}!>`z0|tHJiPJ=&7s9AJwiY{Idl=
z;**|DI39wj9asdWhWpJ`n@|1fvyqUz?{IMz{6rsaAk4hHV1*Z`81E!9L45d<B!Ee=
zNY|EN?*lp5E169qbMV0l_nE0LG@@WXX-{(WF=(;Q;;nVnfO%w;bm+S#z<}c&F8#84
z#+-!*ITTz1)YWANAU{ir-GOW?Ph$lxd*KTtk#qQk9|N1$I6eZ_$~?7aFDJ;htA<z}
zg9v1)e1TGg3SXSFNL<WyOPw?Fa!^YFj0&7(_4PkUoA7|t2q2k3!{5{_lQh`M9c&GY
z^72v~c&f^d_oeZa?tL-urc}b&yF2VHjqJSv)0j#T#2M}m06CAfHTRxuOhtUiw5Y{A
z%y=zs#%O`xM2L1z|L6;rD8DCzva=dkO9m5hIVH-`CEwr{-Ao!CUKRK<lSAnj@{0YO
zgYa0GJ!Y&SJzvCw5cTCVY8Z%R95_u&!oTHV<p;iE<zN^%jHuM!14jN6Dfq(9U<_gG
zHJq%P_I=tB4LC^R5t9Fg$ur8~8xQ1=X`UnGbQ9pnIxVT>1jnJyMiJhu?IEoF6VUx+
z>ozaE#3(Sd+wmQ+VvQn%kp~{73)cTIH&`Ks600cJ@wTFPMD;Pp&(Pe&9w3h5?X5LR
zIm)A56D4}%M`g3=gJWy3HI|6YN+M6@FwlDR@;^8Y@sxJXUl-@duxyc&-F`|Y!cA%k
zNQ*=3f3O5B_84cH$yvN+Zvzg9gw;`e&L?J~2l7%RPj9?Lud!Ki(%zupzCR7*he6t;
zN0#Gx^JdTg@g`ZFFT70ALmCQS-~rx5^6`b<b|awqzkcxloB0sQB`}wE?J{|X^9VT8
zEWMFX?wp0)w)0XFvJgo`Idl#asw|a~a<E?lf5Yqq{=EY+Zg69!kZbDoJJ>R6Eg($P
z{xUWm3k7o%Xgy&v(`{}ae7?6AV&-KV>5E0YFA|j#+!%6_Xcpq0_(TpCyCuMCA-@ql
zEd^yj<_71lS^&Nyp#S)e5K@FJN18R16-9Rfx5g*(77xF7>Up^-2ooHTQkcQO-^lUC
zStIk@xO*)gaAXF3{W~73-S*3id3;NQg545=t?>~NFo6$aw`!MtLtMW$tccs(UW_=w
zXeIs*c7X56dTVRzun5cr0udY_+-VJ&oMQs?;ZN(dFjXgXc+bUsG;w>rohCF7yR=*X
z;udYkZT8b&58*$slk<3Y@C?Tto{3Lgwmc$i4ow7E-kq}m5xzEVyIu#$0oiLjLhMqK
z!@r+?(`#?{TpGYxm|Q-QKUV$Ywqj%@z6?xzEB^eiNBkAGhcJSvM;LAYEdfr4IrR{0
z-b0jU)U1+sU}!DvMhtIx2dwMV`p+^leW#|buD<*)F1Qq+FJ3XCxRD5136cHv-_+b#
zDHj3?GI*ep9#{yz5KHyur-b<Uc&!)Qq6QQE^e39uBJCGbQUfu1&r9q1uSLY4Jw{w<
zHl}uJNB#1?tn4xHr>xM$sYv=o*~Mg?+`MRdf7yDYM)_bgcGgbr)ZZ7_DOOz6yhbdc
zHt0XMx$n;3HUpFeXLEc+`=3AGI+dF)PNhw;j|ku++i0MFgiJF>t`K_`KRFRW%dy&b
zHa+<uGV}v_^LX?8C%x(`VaKUyKemGRMb#U1?Cn{;yxEWal{VEcya$%vp{swzZiao%
z+uIFHJXUZmUPA4|*V`k#KLBXvfSxN(LF4*w@q%-OY4|-^{0p!Sr0;mdz=%ZdeZ3cN
zGN+bsYW6;Z6;jHff#Q9iF6si!n|+iY-&qPenQ!SSV5z~4SekWA_?0_To;T+(8TLC$
zybBt$@9>yX26Lpp#gjp)3dogzUSi5+Cy0q<;P{^j$98Z#0E*siuC)H!n?=1xHFZkD
z-DyYhHa@_}(AXVkvTs*e7090n$AY+#Lw1{M^LF(WeFW+f&g1XZRqyZuT@l{CJ%+e1
zGJT{KFg|EuL<2mge<uYDJ1>PjeVT<t*^_aNUo4xmRw<kH9lhFajQr>cM0CN87**7i
z&0*5C&Vo}@)8^-$lKpg5pzGvBa3k<N@8{p+t0J_9q2PJHZl<}WF?rI#jf)gM2mbRm
zBLi)k|J`kjUW_85`{zw|5_SVS#6WzApk<Qm*@3NtA9N5qgE~IcL_Yk<OgdEGkB<&X
zOcCijX@7>k+xDBxyDKz^qcrNZ@zplmqTW>dzU*Vz1$=kNI2}C&Ex|uukkkmc6i{|i
zFOVJ<k8mad03Lv-P*kkIUoK(lC8jQ&y-<*Vbp|3C6`wF{y}<8dC2Xg$Ji(gm7<jMl
z13pO?c_kTP$_gl21^Ocl!*4q>)Rj?bQ9D{QMATGc?L<c{;a6LZq#Lz@BalUwSk_h!
z@Kqkr(0+~+D3gMFvR0OD5eXqMDQ!BcadybHB+aT4?*Z@_&NWmh{-}V=xAyZJ83I|x
zax%tpjJfH6-}JEulRjSEc;Mt>wPEBbk|$ocQ5pa|H46yU0mXS3d-?Wn?PNUnT;p#9
zg*0d^zoZ%D6l5ihAm^+&4V5NQly3QT#8@_F!+4m-YnK_EE{)louMXAJp@M=C4N-pl
zF#tZIRe|!~UceU3ynAtxAX4ga&vjzh9?6(h-Gw=3*_IdFO+gFM<PoInIM_ZPdV(t6
z5#Bq!CNB_l-NS3**#x?+v?*UrFBT2XGz9r6;?!Z}WLuDfAH28CB5b*`7J@Y-K2h=M
zaBh0T*_s>%L?b52wq5B-N@2$MLQ>D06_QAr+adlC>x_YzBt$R6Wq7#2_5O^7+fIZf
zcLi2<r1s?p%glPL?6;2mLU}gk<RYO8fRgqD_vl0JxeaDTgpI^ahrmY!eUe1XT+5vN
zv`?|}c(BIrAtz6RVaPtlv~~bY5{+E{o^`2K#Owrw+`<7)mc;F+BJ&4T``#e(`(t{f
zClOPGi?{E~lj-X(RyC}MJ~PbC80U_8O!g?I8sYpuTUzmOdQC{$F>g^V4m2fDMD!gb
z28UABM|<6kONR#X<+@CJb_~_Jx|$_^xO^PYp=_Po4DB70z;mrKSK2lZI}ZcKgV26o
zqViS&=~d-$4FDQO7NCcBqE+lLsex3TTHoan$CI|uohVl)IvTk`mLZr_LBJoBfhU&g
zMn(8+KE1aS>$791%8I~Gig=0O*BkpNAk*?k6;AyL|N9r?t0db-?}9*(G%5<(F<5@u
zXRFcS8bHI~+ty(0*_$&3JpuSUFUdTF9Rvx`Pm6YR?`+EV=cSr`0h3O|0>&xKsn7HP
ztbr$*R7LgFtWtDIbrT4Ei!*jqdxBsAQ?7*Ta87?~Ntm+YMjZ6=EvAR2nV*&@C2ay=
zFq8O*2M-=pxR?|b78aiJD1OtYh=Z0i(i-OHZw1Pl2Brd9@v!5Cz3vY>fgD01QD(1S
z;_FZoX|Bv9boicCzN`-fK}OX;<t4!^jeS_;%9~0|Nx%_Neyd_-nMQDw#_=v7e|==F
z3<ydgek2cIbEQ*$JSf~iv7y`%g%Fx4uvw125q_lGaQO6KM}Oh&h|+$9M$OoaIoFp-
z9%x&x=l$)Hb>W+z>0y!qYTmlFW?u+Mu82-=2IM7Kxa}4=1*CGvHqM8!0-N_(Vc=;h
zcD$Hb1~>;ItAcTEvO;CjmvcWj4d>{YSy<-6xJv-Ya3o<Cq3F;-in2$M=Jbu$PAU-6
z+jRGI5kNNO?k6HI;kMv)hJtibYzP59cmk>QkxODCb?o!i)@6<OS^+67bkiKz{T`vZ
z{}YUCFu@<=D@Fswk0)*_)3Mb_A1R57|Ga3hDSQLd1vbP0i~wOjAW=PcU7Sq}KG?Lk
zi(I-dmQ2r!PsGew;(KS_-ds%~H7N{$RINtvkwef{v#&`c>1MXLjC~$ho|n3=txs^P
zcCma+x`z>*ZXRKVNbG`;f<lLq1vYC7b|?=U6icY$Ko?WIZ0lLI7;SQ(z{=O)OFH7o
zpB~O%wPPD>ZB+V*u}l9Wp@<%9dw`yvjM0t+J-_=g^hH8NbiukAJSw@6Z9_cwZ}Y_W
za|nzsJ8q>+c&fL;$!<tT#D(wMoifPty)#cDbyHWyqzk73uuFEvL|caRXcFx(;tc3~
zaT<s!5{H0<m@2ZI;)!*?5!b=wV1qPpl*F?L<@OQk36TL$V&M6}rln{tZL*x>OMzZg
zvTsf6dUEK|7ML#O(u>`NyBFmF9Ml;}#^xZ5{60)Hj@lX@zz#7UnYN&v4e#3<L}pBk
zRN=sqprjc4fsWuZU@|~Z6F{B0#+3F$2h#97bCnWIpJl}vZ#Zu8U^XmUL9n-OWxEsm
zIgU;hxNPEo$pRLAdahrhn7vQX;#V^N@MH8`*^hs$k*-@DWu_BwF+@AXP`Y8WGd8Pk
z$9CH1kg(kBa_({jWO9%U6HiR&4DBG0$9;#@)jh!tgjYPN#PMjNx%l+%5zBdD$Uz?t
zqXPOSR`uhjZ*DdXl<Dc6e&L9<YWB_;J9CXo5)?qaaV-w4u#iWqPbycAyeR|Md7lIe
z2)Xl+;SW<BBij}p(;FA%EC%}PJ4|ZEv5rK9KFu)Mdi6<=DpeDlEoR2TTIKKP%atxU
z)<N<But7ipq}A%XtfSVl9)`Fk*mTu6`zu(_@l!{DnyoWE{Lqz!E5{;smT-TbQRAty
z-N;jAlhwyRh2s+^+-5A~k-oB4P2uq!=1_pe5CuX?7TEr~Fw%-k)*XBc4wn-?H!B|w
zsU3#HFF3?o*Wu|A&0=VXphnmN2#eSfcH{>v$N?aE-wNL=6#a(X3M4N((W1T=Qc<&h
zH(O_2@4XS#l76k5JK!gn92~Jo<QJcl1+Cv75{p0hBRFIVWS~=sx2pm1EzI<*-Bga9
z;2A#Gu1p{7b?{=RfcMw{&=JI9PB#`TV#kSRq>#kOK#W!TKK+ul=K3F;y3BSR-0{{c
zQf+-=!P8?d0f$@0=ATh|wr6%KqOGc+X7dhz|2MYl$U^fcxVEYXWYf7P1qwu*xhu?5
zZ#};6AR>|enRj?#m4y{Y-gywepscHY+O-TPJJul|N5{~Add(Y`n=-XdRz?q_RE-6-
zHbHkF9gq-FI%Lh%&M=hj-<;<wwbAwSOQBw{8iWs$fF$NLsp1SqA?%iCMq?u4wcGH6
zGUFFqP_d&egK%F=A||9d5e=khIF6K-mM-5MJ^RuxfAzgww;3imb)b|b3nrQb6REw~
z8ym%quGKpxj7UwyG{xyFz2AEMoQL|;F~Jbdu<U{0ikmXa<KMa^IFBliK7(|w?Bhs$
z`ggSRyFa{?#0m;p<!;u3aED9)G3!Y<^1Wet<Hr7a@2w+@cr@ymg^%_4v0;K5^h*}K
zi9e{H@SOd&t*BByX0=BP!rL5>dGRWlThql&=AJF<PkvugvesyHXlK0(>?W(;g<=#E
zt7#+@1Sn=GmvJk}rjkY;_e?Auox_Q9m5_FTTB$F>+<7BHE1tk*e4wluFtG%CGGMaL
zwFzlR9x1Iz_tNw6HB%$)sU&Fo`75Nu&%35$>}0SNfUGl^pj|CN;OSE|D9zR;V-KzU
zI;{b@dZG)6;lCwhKb!{LBUEx$2MUf-E@r#ABy0*z4|e=9o&m3<3=NZKxGy8Rr)%ev
zA3%e)fPR9=tzjbQBS|2)${hY5HTfynZ7a6Nx~Tu|VM1#}v9v5g_ZSdOAeQF%IeGQI
z$$GPV@_~6)@`Rhoj^<ec$0^3K8^5;kVri$(fb+rs2d)&mKS*tT5L7S-y^e>Knap=!
zMTeA;8wSupB2T|oQ-m=AGXl3Cx2?7gu02WtIcz9vBpDuQI3;5=g%MWy*BE3$d9Gub
zuDD$Tq+FtL21Z+r-#@83_i{lJ_E#|kT~3V~^)|%WR{&wPEIF##Nc<~d&ft?>;{IHn
z@B_f;e8A7%r2QiajX-+$B};Fln}zFG`Y*T|aQ0$1Modlhp2<qN98u7++IYd8r`pmO
zFlDnXBLsBqPWtD|nwCFfWPuvhv7TmbQaZFb4N5O22S1QU)K+KAU9s;Eh#T5SBbs1$
zoC6yu8KG-S3p=G%s|s0M?tZd|=%P~R3)!2@Yf@7YedU!(IiKzg@>DS%49R<OVMd?F
z<Yx~|Hk?^u>&zRZx=NoIh%(>#y5td94ptt+lEh%A!`f_g*Z&ZWjUjB5WY~&yu}^Pu
zK+i8F+(2GUCbQk$Jq-!KXeSS->VUXcr`hEOrC*!|*!TRZBrBQJw{xGRdpzwqT^bR0
zAQ4%C_VL<s*I}S@dAOY>R{^F1kX&#{o3x8*a@@D?6rNgSykXY5Wta8mB_MP7$+b_$
z)e})qFd{MvjCva-T$6a>(Pv3)%oAs9KyT8T(||cCKnltUZ+*U<$V226E7CNZNRmb#
znqp6W@#E#~V<aahjjYp9O06<S(DLo-miJGK_olfhA#L`iM*_XeC3zz#bchk~za}Ex
zu{EFNH@CxZsK^hzXOR2rLzX1nCak5Y=}k%#=%$Z7&f;bpISZ+uKH-2s^$fHm!IwMe
zk+{V@(5jQBJq|>!k+2hqEH7z;hT%X|4jpG`M<SosylIQK{=uQ%WxDQ@Rm!fVG{oV-
zgbe@Mu_l?}>Mj>=q$kA(OAvvxd?KRvyac!FPe}{W!7YyiYD~EZS$l&eoHvlbcW+7!
z($HQCg)&B~yGj#l>C-%O7pq4)wruifPJMp0xnLZ~`9@U4BNS|`ySRuxiBXYv_kfFl
z*nJZYktSJ2jiJ%M7P&mDu{Jkr>$^)a3Yef(Py%Z9!eE)iITvhi3#TI|cE1@~V}lxp
zwg~Vkl7{})8Zm4*pWJ4Jp`@{ip$zk$a{0mXHOyWdyV&yjA<tE=V?B&4%6WL(7qMAU
zFVe2#Q&_=F)|nS9$RXOIG88h#JIo~^Ku<-;v<yCTA&50fp6rqQO1BPBeFACPNGCfn
z+}sCNv1xp$7Q+6w7_c|d4Uz%F2azm!EV%!Zna=snDp`sfNfC4K?xxjbD!`ZaLNXAq
z^|m@zvAYt=J(7u$khTxA(?qOqxK>OND9-uIx2~r|YuB8WM_Spe1*OfKp_1ompGo_L
zH8?|rEtUr%LKLyAV5|3s@i4pIn1FFn;uJ}>ORxc%65<8HI7_=CjAcc{;Y?DPP6GMH
z9PcQLt)FprC~_rpuRQK>A-lVjWxiHRN^$^$&OT0bPou_qX%`zog%6bEL(D@VpqNc*
zzu5m2FKY!k*^2XHUZMYFTv-{<&Wzyn`OrROa!7JZL2XP246ba$&E3fQh>>YP*w^Tw
z7uGXzkb~hG!fm!Wi$6I6(eQrEKXi#a^%Fdbm~v<mbgb}@canZS8yOO^4hPr5_3$m6
zF>eI2!de3|HuW<b)Ks87XLX&HYJIeaq&n-AeQ!g_W{4<#qpfWCMXanMNs=KVEv_7u
zLFz!6Y;C%;(vqxccgzuQ7@a-4KRWabKy4s>38zSH1iIfB2i$9~Kxs)yBeKLCB<kJ_
zQoSB5S3g7KD$)v~$H{A68>Mv52U#RREqRsvGs`!@Gl`2W`f<}2=xsxM_NvtI!=n6s
zU6T`C!$U}s_!GLYmxqhI(_?+YPMzL{<g*XO3)>+87&l(>aw+s|0I+B?EQFwL%h7Mb
zJrIZ9_JUY}%cF^R27lLW_L{Iy$VEgb;U{r8Z4(rl<e*%Dl#cokp-Csl96*&^!{rrb
zXDz74;Etwl5;$Gy%k++3qeESKUjet8h$?`ReRS5V&hQn^#<z8i>~}Bkp20V!IM}Ih
zu(dZ`VA9d1V%Dyk!n+oKxA=#duYX))?9SYj_ov6VC5+M~>_<(}5wg<4f4(VISN4s~
zDq*y~d-X&r{??D|tpiWx-@N*DtH{~IfByXN$v3@>hrfwNCO%8aQ2g9+Ueo4e$w~tQ
z`?ScMw&do~O??#=7FX;G#>cDp`d9geP2YLvtWb9+JjxfS7PT&JBF~Y$eYArh`=392
z$KSmWrw%*sOe7DdC2U@#8|T~rGH@o9>3Ag(jUaM~Ak!CAgCJ;9fFJDS0pAw}r$sFz
z$|07J+NIB4(6#=&KVt~Ke_^M5xe%<cKqg^Gr+Fj{j~KW>!T&a*&8OHPsvQcP3w7Yb
zZByVa=w-g8k-@=hDSqbYL(NE;E{|gmJj}zI<Kp9+DBB>@w2n!^`#+H5S$>f2lP&t#
zK`PIGLebLj=^yO??F_#XRyf2fv!N;}z9%P;5{liBK*kE$Vbpc<4r)e(x&EaQrK=<8
z@#E%)Q|RiS^uyHvDzDY&`N;?~2}5Ntfn!5Wa>Oec-+OYLZulL7me)OR+*f!q+Dapo
zIArCdq-x@UW`24ryayL_%%}+o2{EmV-Qe!-&O9|patMEvSM9v?iJM?<F$SY_2CMO!
z3q`q!>;a-j{wAFH7o7N|ndXb^Xj|T(!nr5scu(wLZB3Q0Te<1P&VODVYyPhA{rlTU
zFk<Yu^NHZv&O_yU2Kg7T=Y_8&{yqB;;p0nHXx<+>U%QNPEu7W`^7-)pcz@^FQ9S%S
zpGUr=I^T*Ll(*wv$-vhOdCT^(1Am{BYQRI`m+@GVK}kCqcI9pN6*BzhEPa?)w=dsL
z+p~MrZHl^?8q+(4I<~R#mWu?Ackrgd&T)~hBoQ#>t9bW<<_`jJdlQ$@`MUVQmtA)W
zy!&gdRLlY&UosH{OM@a7b#Z6Z4366q^vxq1_G^CGwC8yP4T{+F5)$q^RKAy}gYWR|
zG^DlNRcX*IEhE#6h~prK;$&o4jiUJ2Z7$0Fz+g#Vlwp=e=>_O3ePXTuA^RwA>Ue}p
zkn=`lpVExyyz`(#%;_v6#fu5<JQ2eAJ-mU%R3uML&Qf}zoB0N9Q|CpK_i!EwRp$>S
z$<A3gPv90*cpLsae^V>(SWR8FeV$bwfxv+qY2|<7(R^R7B-v~jdul3pd{~jzwb}AC
zg`ZHg_LVD4UBdgK|D`1VI8uDL@z|?iC^C6lB$@Y3{D+V4%V6H|I5=A#bNY&RP0hFe
zcQ1gE(Z0M)1#wSi^B$4{)xQRp_hw(lOOL&jjw`09SE^dN^vi>n@uW!Li5x_T$rahJ
ztemGIY4eXa-PB>&4u=yQ85ucsMzd(vPFBuIg^$}fWn;F}uzq|%$EljPV#P&8hXrM1
zWWEqMyj$l&Q*~G{1Axsxzjc=+uJAdId)=4cL6{g*)rh;O{RM;sbNvXqWO<pYbGx-!
zENJjkj^VRpVDWBqkF5XUFt&mYk&sLCV}HVL|Jy3Y@rv}g|8QOU|M90pXC{{O@lEvw
zpkw}@n%|PgD8R?BXMJZ;$5-Go7(=xq1UsNsBU|VDBT_uG!zcHvl9Cd^t)kko>zTa(
z7m)quZ?HP5R?XY#d?&6zNxW-wl`@C&mRxv(oMVfRD1Q0&EAVC8tIUjzvmx|56k;4`
ziz82u-Mxc~^4NsZEfcv=L&gl2DGNZ(h62>~XmEK&s7Gd`T_HQn=f+{hn`-&``FSz`
zfY^b@0ADH~A#SLpHh`mnU$?6e?p~{>C=Uf4N}O<3f;e29g!)mn@vm0^T)tY5h<_<0
zRaCUHOyswkU0<FIvL$*O=|SXXrlJ5~{q8-5axh7_^ZKMIM-Hq6u!wr9SluOC)vw>u
zjuT-PK@>=ZGgP8tEYzJ3??qpG3im>FJ4{js2dk~uhv-<O(_Aq{z>X|1jro0!M8<K!
zE@Vbb4y0f~Aj6m(#1h`Y;-40Ly~bC{<Z3<O)OAd%sm6(AXeF5GNSjq1^QsK2{Q59R
zR)e>h&9n2Iyr1WH{$As}^8sIG4nw*DzrL*o%)zzO(<D;mcvh3|0Hj*M@R1luU=YFN
z9}}HlfOk)|{qlLJ^9wluULO<3=hqC*dp@li#69ogO6{#=%ZQLQles8RgOXo*owdbP
zRreriUCXpVHiJ~?V>&8?9)}nx?O2OB+xKP8!gW%lBO)T0@iS99N=i%l9edM(${86M
zReTuO701(_@TqNKcP!)Fm=_*rWKvvMRaNysQ=(A*%eV|2c>wy=)9mGS{$H=}Fl<bR
zqb;&{5Qhouy6GDz!*eXS70`nXH@H`l*pyg(JoaK<7ikS>QyF1u1Z6G5#t~D5YsVR1
zc|S9a*mK@zwVDYrZng-bUO#sesFD5pj{KFSWo3qlG+7-&s6fVX;V__tt9=l+%N+3~
zz8lEE;@qW)s3^MnKyAILH3>fnQe$YP*s!UZY^C;^BVT9jIoRBfcBrG$StWz~>*HtC
zlZl>T7Pfuaet!bS<<IQcU6tUs>EtgKV?al5M>9TIf7u_eP$aA;n6SvA)(lEv|NApc
zzb?ET3VE;A1FHzksRsamSdjS5wGrPsP;iT1@@A-^ef`V^&=3`Pr0lRkib>_g98;rr
zh2`amyN!RZ**Udp9>#2T-U6Abn_gaCjI4bMy|=Kw;vNz|%%!A`!w?e_vj-Y!yWhWX
zMgIodX@x#bO-)pOLNwidXtXIaq7nwGq#~@s9=yBem_d}U4sb_%!0eU)FAYH=$=hUL
zcgTXyi~;7<$bt@AW0dOsOlwkQNrWj~djjXHuKIJ2hwD9fF*zpD4iK+er!-T+^YCj4
zms>nR09A)`O20|F<mP=f#$|kjIJbNdlX9kdF$T)Wwmp$ZF_@8%hF@PLpDeg`@GgSg
z$4N;^nO#E{=B%P3oId2T_F>N8jST~wZHVjPaU`hhb{y-$ILhtwkZ2LE-7`pl9u&(8
zT=_1KiBBFhEKC<AVt&%>1He5uC|%Os^+IWn5hinwfX;2~R9>+s71NO_0l_rRK^Eca
z5XR_(c{8|g^8gtF5p#6(Au+af$c0KH04J-NN-&^-KVPtxR`X#}k}TC7^$hwL-uIEr
z-}Z>rybvTA8BD51r=Ac1lyYhD4&SL<vwX!062armx3AfEL)2EbJ*dAf_+dC$XJiS=
zP=)uRY)BZ@wg-?ULo|CBunNIG#8yxmoHke)R@JBxR698_WRKOeLY0Q?vMPRHAXso<
zUCN|0UU_^qMweDXq_=TND1)jv`vhEj1iC(rsk-!fzU~3fcewon>rbkMMl>QRpmY8Z
zQ8!~FWf|hO%=Vz%x@sm=p5sR~BDC*dayGA!HhoK_F}x=erbL9$oR#O*CK^;H7)(pR
zP1X)Vz_n~6F-T(JNkQv9lmX*k1*D0%s%UJ8b}+A09eypZm`inv9R2Lu8%RSeH<1Kq
zBMJVFxL_YeR~Isnl-xkKo%5&g{qn}Ot2^3sGF+U^(gsS$I&Cv@;QB~aflBG&gX>i8
zmDe}hggkw^0R)LLj6`p7z{J+LxH$h+LsZ1MqW1G%g~8_@jS}cOaE*&Xv^xx0JSt&$
z8#^Ngb4Q?Hb>t#UU2-%={qZy4Bo*a6IX~9t2aBB7!`I|vcM?WE)^4bnV~BPJSCe~V
zu?Gm2Lb06}wcSYUAjFe6JHqL5mqy1d?(Pi2tC6XO9HtqX=gwzQ9o**~1DDxGMocR!
zD};S1sgu|>+h}^_oa(9Ei%PW&?^*06W`1FF+@)kCFHxq6Nt@inj0x1yuAl4?ntXSX
z3tvUmZcT8c9sQfRWsYzJjbMOVo^=6#%zNaa;7jr5OPN3AEzAtz#zCfJHc<U9Led(e
z@9vupuZzJukgVT=3xp@;d`poaK(7{VLYK7AV<??ajr|Ml6qL=*7u@!YHGo67(>AsE
zoD37`suh$JjXaijtSw{0#0-^tpzqf~WT8s>g@b<55uz8ma7O(I3k4djynFXcjZsS6
zfIEx34%xaUyfiu8ysoY~O9x9%>^l$QLE-Cb#hd<@;=A5Z9XtVV=z5?{#TCka(v5Vm
zsA_@eN*X+Aw05LN7`6k4yicuOXd8g!Eb7Qy0k1)J<e>j{DY#Jb4WtR9873_fsr4KF
zSaai3blOX)vg4j8rWFP6Wj1@!s(bV#9R32}Ha-M!(w6sjX4mE<ncKpIHqk`lG!+@W
z00mr03VMq22eBF-2NBzCVfJ2F<P^yAumT{|Kxh`MW?YG=DR$ZVD`0dRuu0D&77Rcs
zU<At>F6<02J}R~0?QHs#Z26AKu|Dd)p|b76@0i^X#sG0_Nhoq6B>+3Fi!20HMZ*ZK
zLDp~o>A{s=&`7JHIX}lV-td^}YX~XUMG+G9;~@6^^1_wcUfMt|?onL|0FAB~&<0It
zyJ4@Xx4xf$dX0Y-wba1^-)Z-Y!d6N~S`BoZL;+MofzU4ioR!VlwVNQWB8dpq_>m&1
z8DKFL)LUX4yn=o(>i~b>HP$<SToO=j%$l4Sr&=Nu@vcQHgJD0C_Q`@bk&}0&a_&zL
zuU##mrU?_DJtnv1<>e){hKBX>wmr(r3hSHGtD$g+#=27z9g5DdSXCz{K}LHsSUMrm
zz%NgURlnn3NCo%Y`E~)IsCR%<_3vB(#76<@72CER8)<fj)CeHU(|W!;RgG#Vh^9hP
zskAKYLeW`Ae@Exy6$34OwL*<_(HWM+dhfx2YpDvh;@lM9*c5V+)JOn4c>xvv195qp
z7}b$oH%R5ikc*LXt)4ql>FQNe+8js0DOjJC@d&}1s31vUqQ|2dN=sD`dw|{x8mT0T
z!)sds=2Kl$vn(PYt_X;f0di0yQ&UkhKZqZwxQAM|$c<gCOPEyY%$S@Iv?)Ic?LzOu
zh)00DY%S@SFeOYrzkjSy(lQ8uyBS+pLR#0n3r{Nmnd?o51_Xi~*;tx-QsZz;zg5iW
z50*Fs3*x*ljx04ZzCW@<q)xv6y5x3eg&R1U)N+7~Gm5dnwADUpPS*jdNriY#AkK4W
z{-My|A0iwLI1`8*5a=uOS&%PsOeJTUN3kyzkha1vQy!tpFxZCLmULL#MvSbhw7w%8
z5<$Jr?@#!_fIH+VDm;+vh~odP<_WNNdHyFmRUKta!qloAT5Z9AQQLz?&}RpbFlj^{
z_aOs{a7F9NvxGEEcDQa@^1f9W8|iThs4}Dix0Ug(IBR*-tOAZ(EX%7RQtl#<6@B1g
zjpA!y7~}>rhG(YAB_V^j(?5~o8=3(twXVX_XSX3-%w2P*`W0A;ToNH5Jds9|qq9Vh
z@H!`;VM*86OU1F>)IBC5Uk3>kl_XD<I-?5qqtHaWc>|h})SOAviBo%m1~1ScM5EgP
zBJ}tluMfUY@+U0_NlB>bKo+_yWSz1bqtY@AaJX)cV&c(riMXE8GMHe<Hs6noth2NS
z#^l+fxY$^I9LkDRd#fa+mgt3=?87QJ!N*6HQVu)_Q?f{wGf5<44#JIqdSr*2Nl^IM
zb>uM}3?7pe02R>6&bAOgvusRJY~HA8Mje;SZu>r;3IxFztkjs<@zhGL82OhIBtUt<
zLj9ngJOdG~DyIrNi&@n2kd0|%CQmxD<vM|NRW(^qJhb_?YYH0IXuL|WzVoCbdg#rY
zH|+GzK*7p{Gv%>H>fU;ZyQGjk>dn3Mw@qzx;PR5PGJ&Z1+(e7hdwyS4sEbX8Rt-3M
zzsL`ybBl^NBoZa9BLVMb&LPWjs>z<^ej9&BJ~qzcESp^|a*yrmi0EPEx~tV-jfb;T
zT{uR?A-t+Prm+8Q2vn#j^mRoFlcYf;QUX{cI`8j+-{YGf#KlQrepqp#`#FGT(zaQ}
zzgINjb{XS_B$K$}jL`9Dfb#h}Dd(@a*$NNoPGoM}cS4E^6kC+^vIE8|0)nA6M+&M<
z$`J;AiOw$ULi0ww;R8e)QU4zNLT#LpnOSFHYoMfg`mlkw`bZgkK~S5^FI|mQkxsCf
zZ)u|Gm8o#HxV-!>?D{6E>wzHXSxFB?vBt)&gNlKbH@Ldyf5+cU+bEY4EZlVjAuxI-
zc$2rlBgAcG3wLqbB$<duYTpT{sfi&VK%AMH-g2i>HR!Cdv53pNMRy*8@`0SLd!*{y
zfmdBAm<fnu6L|cIf2S*4X4%PGaYPXk$Da^&I5HwaO#D*~z~BL3gnggMX#%$>KqrO!
z5QoHKXQDEeSQ?5Bs19}*nZ)rO6LPh~Dvvb<HB;y?Qb2Qyx=w*5Du5C&0BlQyXxvsn
zrsRRC2Kdb1zkf@KQ`khU({$Yfp~xGGeG&;_-!0Wr*c>u`J%;CWEm15=(!AQ%R0J&@
zHA7H&lNKl<9x#1U`J2EF<v`(r4k}2NMAIgUIO$M>8TVuaO!PmNThN^8zQP+j8%y|r
z<*LuEd6jG|fw+fHt`bYZU0t<O!O4yiOymd6P%2TQp_`CJ52bZE%LUy>X*jvV;qpuc
zFQDujQDAt^;^#6r|2q@_*tH`ztKRssT0f^Gi-NqQoc+;jg94x|h0yU(=V6h*2fdf-
zKL-3qT^|@Mhst!F)2B~U51Fn{(VX+`0^IaWbMBd+bmD~1ny{aV!BD7f@9?8CRR$AV
ztEy47YRJ+|697h!`$x*Gi0Ebvqw!N!OAkB7RL^C&(X85)Yr&pZNb#Jiy^Q7r;Az`q
z4Wm?n3l5cd<BdegOG48V9k-(#29JV)#>a}|Mm%Lg+=McGbUujX==X2(m0u`%D?p62
zj!AqpZm=%PD~I6)>24I18f%jp#=wH-dq6u#kylj4gu|6g!oq-9^-^ae=9IJ0Xf~a<
zVX{$gVlClKmm|nZVY^f{2CwW8L)~8Oc&BY2wwG|lmeE0aGl~KkvIsbpEIZ@L-!M58
z|50;3DxVb>IAr3uFj+1oxr3@QkheW>p7#PG@3i`xwn+G1PlO`=F(>iQ)wYuWU_HrW
zLc?wIoUKne87UjM^TKwk7CpQm4#loT68a;M)s9HN;EUK;WqV5yb)=FXxr?lmSHum$
z$!Wh#Wl+|%ZL3dm(MpbHkpe`LJtS;zmYDb`LA`R!`&^)E&OPorK0qy#QcBVx@DoE#
zu4q;gmw+ycwlYu&&0!E&=$)y8QYq#Z+s>+v!4)bTa(Yb57t`jXhE?L-0@EP+HK1a9
z5*GltXn?5IWuk0yM6Y_G5m(Q76wLyX6MxiRTZv@NiX^PaT}d<{$L4+`_nVuyI4mzu
zLP}y#N?6dBCl7AQlJW+`H96GxYL#m)wd^6G63MYiF%Ma30jAv-hX9$6Bq<0H&-|q#
zkp&KAxJf3%2bNlRvTJfYgPbMRuaMY+5>mv$xi&LuJe=E46SiTQBbGp_${auQbf1e0
zx33R?(0>a`o;m(0^3=A=1;U6R_%d#46w02@05GD`Vn9IUColJLrKF{sDAESUpI!O`
z3a#8~D6XT{V&afsF|r6$A<K)vpvk5Hgg1K?(o+VDUrP}x0wrw|)pBHW)DyxxWxkE&
zrP6%g%YUmh;Xr4F9u2dET%EEax%R5g5$bV8-Roo#AyIcM)4p;LyG=r{WmbR968M2D
zcb?R|RPs*DuGkd-F^-{K$R_Mnn};;OlM+G`jnsh?yY<4*!;S$T#tIp;TzCxB6!)i+
z{8w%L4_x70=35YAe>_tNAK7*5Xxjr}(8{D8(+!Y&P6;}dl}9%YoWtnMZ7+0@(b1b6
zDMLLhdIZU(ZBGzJEeyZFE_2v;@tq4B`*jW6G&^_L{*>jFo{7ZuH0QjD^;=4PF3w&+
zr+>Rne?*4@rOBA%j%P>r!Ag|d^RyNpeqX{d)iRA>zw2iEfD*N!J|N!V(x2EfQ`P>0
zZz0`1)nCZAKYO_aI=YVgsPXbd4k#C<5s?i^wVXUps6ojqI9AQ$F%H<fp?dK0B)31c
zs2Us3R`%Ns6<Zd!U?(U=H(GTd6|o_(>l?cV)1`o{AoAli1^^BsToJwhxGgm)s6d)%
znI~0d8W#n(MSZsjL0&q39ru5tLLv_Mk^qK18voD<7WZ(5gc)<^%R@=7vCq17#JpX8
z<QaA`ZE)(+CD;0za3v^~;&Vn*A^ma;{S|lnR0JhF!8wd9XeaP)qvsu}&`<YjJ^EsG
z+^h^t_^$hxqLzbYD=o}4sb~i2Z%k$vL7g^1h?5hxx!~1Pc2HBajW>i3-c>8|4NTy%
zVZcd$M4v&6RHWkK9wvZHSI9)(Ub`#m3wF%l!K#E>w)ZB2iR51S`^DfFSo%-s;tY^?
zFgd8E3keQB6Lwl#TRR$ULl{)_M7>-735CWW@L$J7G~Wwc!Ys~pptYm|;TY$Jq|1fQ
zuW&A<v=*t5J46g7<kcuo$jzUf{1uN!Pq2sEIz{Vr0OnZifIx5M;U)cOs)^U+J7C}_
zNt|uR%M*c)U%=B`WQAAjsYzSdI5vaG4FYHWyYGDN{_whIa5~u85C6cLyCVV7gpISO
znIl_5q?i1%^|u-#4T;1cE<j3(_4AX2IjKk%Uy}q|V%41U0h1;)j`coZcS*gT<;A}M
zQ0HrB0Mtxm4c~VxtE&bbjBC^mv)Rir4^YdfRZXC5VQJ=cT|3(^qdubPOfQ0o7+nZB
zm&z>hYi-HV7u@(4+NWkqq_Vz3_~Y$WR6f7WYzpbVQaTf8=oG+h49XUO{~U>b0sh9k
zmjyt2?X=kPy0kCQ`r(e3+<3PHY+!20!=aHwn<l2jR_pjWp3zmQ>=+veaQzSjZ)8#?
zeET01#u2z>f-m%yUT`ZbeWCXe7@BAIq58`MEMbZbIr^Y&0FOtw2{`=mWrQX|AH`vu
zA&K2sQ-d83{=R?imp3y*>!sH{;Cu)n`q+U+4Bn3zpO$_ZQ5Ev>tM$N8d0lX8_`Cm3
z`9fYa@X(OY@crWXbt@s5!eKZT86DU<0MweMvEK#QnhN#AfFB_Q`*jC7<Yi&MwHnTz
znrKZ(`4DND$2kLWKl5a^!=kAbkb8vyAh9GKZNP3n`2Lgp)L`1bBZ2)DE(!Wb{Pj;3
zj0%kbu!__upk*Ja`nPDzEGhVzw@=!?1L#aDDk~J689*Jv6LWNO0$ZmJSNTaBILxu(
zfJytL95>zFypqiS6kqs%oEZI2lA-_ga<uXPKTfSXXY-nj@X1}Fk_Bo=N*!`}$A^dV
z{Oeq;PXg}QNPVuTpCf~b(wN?oD3`e0qq_!pCE9R^qu`hh&A0Vzpj;4^N#o7cuhh1h
zB^AzH;SwVTnj!nhg8o%sY?d+fcv|RO1$G<6$QH2M*S=UfZSde0WA8j%w&Kh)sa_Ff
z=?8bTO?=e?q+R}UCDk)ttw+zPEj*f=iWs%PLimLKi^^uo1D<`|1EnxUsVYO${zZ2h
z&$DPlCq|HrsMvg-*ai9WdQGjujbDgHALPr17$TEHToky93iFKdtE010yKBB)!J{I{
z9CGxL{Jv&`1*mw3P=F_);S1D9@q&h6f>@+9!mKG5$Cnr!94r{TE9wh3s!w&pg5U^h
znCL~Ub}x^MMEz)(#+_j4e^K5CN8N0JyS}y<m0mFNZ_}+`S&+XHEWAbdB8mR_U!2}M
zI26Pd>ne(D6A%Z#cBWkAM&7?UdM5GB#J_Rc(Ku)2U<ceLAP!X31h~3fDVzN-e`tq}
zHCO8g!8_{%x1*Z1tC@)2I}3Ti6VEGu-!a;CW4V-Z*!XByhO31s|G&xfQ;1bw_i$NT
z8PpZ;YCS>rL?@Fbi|FLDwGSV}EK5bFL-8BvH59ws__rDoXdz%=B8Ke$;<$cgUAqjw
zf1SrllMEW5fF}cjg#F+t7Dg;`p#Bc{C2t0FjVLybKDy~qPs_)d=yw*pP5RouxN3Pb
zRMdb^WO9hz1P83*uPsOj^xCL-|K@-Z7{Bg;ysxCu0#ZHt2Nq)2eR7`~PJiOyyl}S3
z>@GrS0ktz&2Hhyn_ez!ir2MZdB<&KGc0;X1I%+0|_E1Ua(X~+v9F}|?Dhl*p_h3Hb
z?M<<w=ybDQOSy2xlo$JC0rFhpMuI>$HRpc)Tp_iaWOI1MPlr&CDITx=BNe#;HDa$)
zggoTqXH?LN8nFwhF9Po>njFwp9R0D+JzRMgF!Re0u5ua>kmiCxQqmUs8UZ@I2l8^Y
zzH??Olc)gxOlm}^>OC_}BfwICucRH_LElm=oQIBd%RDzmCZ3veG<FXfjq)nGP_{E%
znvfn>l1^>8F}8eS*?xKOY_|!Dg9C>gp&^F=Wsn=in&P2J+{?%Uj6)>X?I0v1xpkL@
z0s#}TCOgiOm7q@0h}9kc#V@0k@V$Mj?DerPu`#+5PQAYR(GDlAdb=T77oc&0Du?K2
zd$cn;aG0fPjK#(9l5$Q6BR1)Gqhy#0f(gKB%xeTLpMy%n<ka~FfJ6%b<M&b1HgKWI
zS|mW}e=-;#$^C`fLGApBh2RzAVfzt*+6c|L>C*#FsyMK>NLWl93#!<96(=$$c^Es@
zZa0d%``=uaS~x>w!%yR&5dx^c<KqAq2XAYr+yAdT@cB-{R<r~ssi20`)gZ{XXvD@*
z(fc*Mu~R6xoY)A{NhAvD@h8OeKL0I1`2s-=`$3(LhGmR|QKbO!O30S1kX|)Vc`|z3
z`k+C|jx4YH(NvrU+C3#T0Eb-;e`}9jyK6Ml4tW-@$tpogRGoYL{JTB!HRWlDZK>*^
zmsU)+BevKQ$_ZZe;_gNAv3WV$llzV{vXF&x&YP;#`T&rz-L?wzm(2!UJnX4B?8%sP
zzSK(9AV{TybMLT+cPJX>-y6%T9aDz;;4nzrYGGMIdID-TNolHVj=pfvo*7FcBCd3x
znU)p67u)i(i500G14&<)*vHe1xu>a#Jdc_9;L;3mgK8bo()2CLe0x{9`Vz=a6=>9Z
zd?EKszW@RzX{+8r_GP2h9Dn~Bb*4mBpe!1G%cXe?!r=v=SXKYs!#xnazeTlAZM`aW
z>FUzq{#}Qn7(-42Nba{<hW;cuc#QcWgap0Cynv3JPZ@vd_#An76%NZVT;m<!q|Xy6
z7#>E^$4$tkHc+h+u^<4}ojOePa)id;2|+ZkLdH>~HB`?*c$&n~56JTJsqmb5LTc5O
zC6tL%_cIZwRGvh9J7HL;^|}kC)D99Q-T=}Isal=9{2M?){+JHpwTPmlZki#GS>Zj0
zGf}HW74gF$=Kau8@~q#`yArD&)K{)!Ds*31bUy4p+Wr2FQtIRP!Nq;NPT}SH$J9I+
zcrbObrrH-tz#9c5o`qH9_pk@7K+Te=$3O8RTTqBavSy?wIXL8!z`#)BwBu_-cC@$^
zp*~w^uSzXEPnlf8Yl&u{x|&CY(yeheEpgLkqTi@92~IwPXi+FL<la+jiDt*zQC%vY
zL0g8ciXiCvC=jt*+KqjAo4S@$u`{u-g0P*-@UL!w0k{2RRZ&by_7?`2R|JFI0d=al
zEFuFdpmVoVQvJQ$yhAs^pi&Lg+VRphR4L`_O*HE9o0owrgaIUkocKJH^*qaopF><0
z6%DHlRZ|HuECx};=$7ow<nRiD0YUfNrN{y6Z=?dP5m_p&IOC&(;Ph0w%vlAmdH+BM
zO0+sk3#ejRm`R;G*|Sq=VNq8qbUxX<D$!y%!F4pjOui=H7%ST-J}v$!3=RLqKuwd$
z`}5MMk`~pv=d;?ZnoE{QnxNtQEdocVi8A$dLhn`U@=5G0kiU^xV9<*jok<KxC4x9-
zDFb(b&{JJm%Gf<Q6s0>+YkQ)$7M|>&YMz%u6PpOXCVuzHkKaf@_ed|4uqs6x-B-mi
zZ1WqanY8qdE-t$09eNIpEANbPE~`4Xs{5On&r{Hh(2mRN<7%%@9nX+vJ6i&ETx2t{
z%L)mcERsA=Q?2z88@UeRmFancn@_z$Cpt-#%Q9HgYuH967SGcH=wIn@D58Jm!QI?x
z4biz)ua{~z5V{HFj>qj_FRJw;kc>cL6~m7pHJS=0+6&0Ec|Az(tprV#^xxXM(x9l$
zHQjD8#)uOWw~J_QZfg`G#3X>C;=VLSMFOHCU<5THi-3X%h%s|xq5+e}7L~9>)3OQ3
z*6fQKA%K7zAT)~x0Tmh$v=OoCd7g$$=Eu}jP1RIQ)znnwM=AmO^f}-8zW05$cisJ%
zb63|{O~EXJ;u#Kr=E7VXVLl8NGSEDv6$gd=O>p-zA2pR6;vDepo!w_iNWq}?J5IR9
z=6Srz=@+wA2jqF{?NPTHozg}cJmH=Vy@AsfSo|&rL-ixh3vepVp2HT;J}#@?le{bG
zJ2RVeQw}ee@2hJIZHNZ>`BIu6u+RMa*EVOvU*#?VHr`k6op-lsY-GMR&@K%|9lX(W
z0pP7xMg?p2YinzJLf6y6T(@(&A91vVl~LF9+z_SCaJ$&$5vJ0B7_S#gEWCGXdAFXG
zO*BeB_`QYA$Jx;I=%1>o^1SVpY33b~C~Km!!ezSm&>Wr)kARt%cW?V1bZ#=mWw4?5
zu3mFvudn|so4kVBD_^caKxDY_SGOXeZliXyS%1{wI#7L#>k!r?zpRiO*nI0T6xow&
z>b(=&)8UP+>8YuYV9_WCl3LU3ABg!0B)u^Uhtvz8+!+y|%9*7xQE&5s$5$6Z14qd{
zaA4;D0s3KlIss|gpXrbyW=*5sXlzzw0wGi>BLWV=*oQUUJ}SPbni(B9Z+5v)sN%r_
zD__j@k+FCjOaWzj4c%Wm11l3CsA0Gew^1l9K0s2BhS9>Ia3(u7R%uszV$7kKw>r}n
zub6qWZg1$Ax=GKy81k?-Sur&OivYCTT~z7N#(BqGk~P(uPQwyinn`H;HF7)Aj9ApM
z;-^rD`Zc^pjD?}h&&GfCbU2r=rZT31tC0KujO$}`rftI4!9v348r=cI77P#@!ekPP
z<M4X+AQi@xkSX(*Ut<h_zkBj#)VUiFMUL}7<^JE2jAP>C%gAM*&+!PkrrubvF7Xb;
zqE9}rL!)x$m-9i;WQ91cGiXS7HEJ<lK9S6CY@h24hrx4icKJ>D8JTf}cR$9r32p^L
z_{0$718;htg0_G*WSg{=_cw889`FBnMc=LCrl-lW6UpwucH@YWpH487)QeNF`4P^i
zOkr1yBrJGURWy?<vDd8^gJ5uuQ=wkzdyyEBd$F{(1mLk)NERj8wLU9eX^b0*JG53j
z4dEl5Pv?;l7qYm?KIkWRZ{XEMk7R*Dm*C7A56M$z74hC1*kts3F)`H}|ND=db;^ew
zz=8Sdy!xR`>%#|9>z9bcIWHvSufdgV27cE^v^LXubT6o%aKl4%%6t)Nr3HhUu6?)<
zDJVR5hJhbg8#4C1%SL(VTJJp{rR%hop-4DtMp%Dx5v`is2|w9<H<R?Mx>r5WsPQdA
z8bo!1f#1Wi8b%+0=&yCnN$-NOvU^C<hT1)XbEw0fUU5UUvf#<D!A$8s6O*M@B8mC$
zbl2jvAi?L7rV-(ET^rl@LbdTvx(Q3bGrI_e&g&vEB(4u^m(0R#dL5vs_x*A63;!iu
zy(wd}w~t(<e_{Fi-x(KLx<CEr?3EWPoqqdjg6O-Tq)L%?)VY=NSe<u^jGkHOyu0f&
ztEkE;>p%N)r)8>x{N7oo3x9-JZkqRC)H|7P>gtE=>g`iqJr(tScIhsI@vRSEoUb|8
znbPW8-s4m5-RE)0Q|W8!+LY#={(ML0I8#XVC<t7aJn*HM%BJ*tRTFuL`?c9L#(GWt
zab>}}Kr)TwKq<QVY0Y#^?g?0b-ag)oZlr=(W&b8@Ys<kgM(e94j|AWO4*Fe++)?WK
z@g`?wDe{w!K%K@c5Y9Wo77+z6k3tEHQ3;@lGw`0eRvym9-MNysI%wL~z0A<^SYfe-
zN|6$`2OGc(g4L|wJma3&!l&uR@}t*$smJA^1Iq3u$zj)g)+Rj4G-?oBo#LJ^<KDjL
zua1QYa}RW1W!QwgzuUEZhlBPJ(_p?FG}ue@jwKjFhH%<UN*iva{-~_&kXsS!6t;C9
zvH{<PA6Z|C{t-&8-k!{NnHk%Sl4yyDGb1S8w5e;QoM*)NML!BP#V$O(t}*d>d3njF
zuc5`$8~^@sC(a1)>fATZE_xmn9vY5=MthZdedhLZW?3U)bPsyO#Tr3sT1RliYe(r~
zQ1J1!t$>3!Bp>lPzCvrkuQpGQVM5zK{|3oZ2*gOnG?Qire2<J<W4s{6AV{){v(|FQ
zn8?cJc}M+15zFtnO8eBWHc^`B=s$lE_%YY`@lH^aJKdtQ7JNTbvqN3<LQ4Ap^qgiu
z3I%l&W7$2Tf3DmGqSQ>YTbx{Xj3ACg3B34TeA=@3XFnQ7F9ScvIV!bvn{_1~fPP=>
zUxlk^cg90K#~@)*H=T_rpcKk5sIH~ChT|Pa44hgx*e82vcRW}x#t@qW_r05r_JW#k
zmz|HCG}MKA*l2ep*yWlqPDkuW9XRNj>|CShnS<C>Rvo%WJw?#)2IxzFML!?{j5L4z
zh~q9u%EMJemMU;#BwlNV(buZpTuXo?y=r6{6Y6kSgmCT!&Jolf`6{n$6_tNxHU(ql
zG~@CA$3cDsGDb{W#|E_Xp(>-1IZ}%`1e;D(j!qjQSJLla(Q?D{#NJ>e&qJOKa#j|K
z#2X<?fkf2pRMqITA;o(7{g=!3!NXhU+>h5rSHZ@7jngs4Uh;Zv=tq}$*|c@#VQ-Yk
zny2pQA3_u+k&lqpSLx(1TUWXmnuqBzW1NPN9M`00IypL{a>$nAG&p=&|KIP?jY97(
zeb`}wU9f#!{?|OF9H5r>O*M+1ICL9tpx4r@{m{g#{V1fYYJoL8xt$fs*hgH~<4IDB
z7>~)FxP1|jX<SKmfiYR4WPLh4ah%f-cjKD!?2~aU<fOUo1q-ongnu-~iMBWF&aNpa
zQLSsWN9WKOuIJ+2bxk=)wA^xm=yB8Dd?@inVvo%2V-6usEfjs6mDDJD)X)W%A-h*k
zfwl^>H33EtTHBhxXhn==teK1X*g??eKxp8IhH>7CZ2^<Bg8uweR7i+R{JW^i;Pw@?
z9_xqOwRJ9U9r{+A@hwH-A?Xc=?-<M9aGo)U57j#X>ys#{VZ++8v>SWq7tf*3iKBVp
z&H#CBEL@?8j(wDPR&gIij~Tk&N{9h0tkX@ueT6L}zqt47@<E>PRy;8yRvMxHYG|hu
zmK&Pe4E@<`aH1~V$favofb%2Ha@;+wp-bstBpWjJ^b$Go1bt@YiHXZx9$P60{|Tb~
zA$=@m+lf_az>YX3Dtwzl964*~YrexLjx2&4r;uTWHzv?xFA$6xq8Db*9{iA{%Ld;$
z!CNrk@CVGz2fxeFV}s<M;JrU%i6aLu`2YU|UfBP;_afS^QeQATFUv%2eeaA#%xRCz
zpV2m)E^T%S84<b#Kp69@990#uON#DY80VQ7;pVgn3I-I?x{0cvQ!du*AJ=a3jm{<@
zfz8|I&pU46Cr!RLOyFh0-b#~Kz#ugJoUUuS)*}mK8F>FaQI5<MF;dMB3e<RyRe~Ey
zn8Y_Y34+K4rw`y<E`MhK9+&8UX9em?N}D~>PNIpeDv{{XLm@22E0qFOofSgjWkdm}
zb`Tab1r4PEuF8BGvVqFrzZVOq!r09L81=^@ga2IqyS}I8RPbNCAeC{o*w$??!Yf1n
z1F$-#Su&9Xqg+Z5vKY+yV6<ick26_R2tASq&>s!Tg}Q~YNUv)mR9iU}5_v&+$4$p9
zRp^G4QYXpAjhqoR>Fk3jvI8Yw|HnYsxu~*dOK)5CAFf_h5)^BxA5vPWKV0xQXU0I+
zag_H*?sg~#RQ=3ks=(n^TQ>kZM(ZCZ4K~0!v0Gjkq!Wto{0UkE+sy_A<&UwQl=SBH
z=9c9J0H3CYxov-A`OUj6hax9g_GmtY4s63eu*ckls68Al-F)mr>oH9f_Ftav40rgQ
z$|v{<AF)la(3Kd~^j;so<R;NPz^xiVD7pj+H=(*Rw%+c($_{!jMw)2GrARl9HrEaZ
zlJgo^r8P}dZ0raJB_Ka*?i)FSUpqd!>P3RjX`L`bxb;{63?zq@D+l>7)fh)N?>wYj
zo@ivEkO)fHg_Epm@)&9fgz6KNag60;=82iq0vS``#F-%<rIn1<7G1wjj@G*KRxqH1
zNFf^yY{K$bmC2_YLz}+dwkf{NwrBQ2x}n=g;;^NH5(c#y{;9<>yhzxPjWu;-nNk5g
zFPleUF}wq{Ws>1t?wT)BP<|mh!Mp6@dtc)y^BNAthGgMX)Kr<pGPLWC;7~*snX?Bn
z2pEbO(dOkK5u{2#?MYoQ{r=Vi)w~HGXpV@Z9bf2&^x(O0#v$AKyq7l&T$ODFfCv$d
zfG(&$N==20NDOVkLAR>{=S!MD6MeAWKX>1B>6=*a<_>%qjQp7sQYwjh4e#a$NMcf-
zMQ!SsGQjyTVs-Y%F6&^*3KrtES&t9Oo}4Z?kUO+eUOduV(v_7X1otBGz)NnpYDNrT
zB9sT@P{wW94}G!Z6w}?BPTmVVHFjxl+s#ETL8G0lHrjTcUo&U8H*EQV%myR;Cu}qD
zo2<Y6Ebx-K1Sb(vnuMV_{=p}2+2vN*YdLSRxeUO5uEfhWOg-a`$D+a2G52XPAIz!;
zQCb40Ef9dPO6@$-n@v)&8)E!s)L-&nD-hCoUbq#7b9NE2Lh*~q8AqDmj@)x_ceqi}
zh_r<4#J?U5))YmD;zuag8yw}>RDy+MARl#n6PSx5StOq4P<Y@!c{(s17i_qY`=Wqt
z7<_O-5mTOaM}R<^DWWf^V~*BQufdO87~GN3V0g-csmAxmU>7bi*#)1?tyyxcQB95^
z7?#4_{Ry!CGHXTr2*_zii`lT38JhdQW6y2d>g<+HCs>CX?!e#?`e%9=a{;@hq!SMP
zR^U#|WIT^mAVseR75tsu{WOKlk(iM*rj5obC<)3IT2eGz$n%Zr&!8jfOsW`Jpnfv=
zyp|dA=CnUHYeqGv_yB<uTQPrSfMaa0n-b+=se(R3mZ{Q&R-$(@O6M>=rWgU%xcg+=
zFt6PhN>L6#LN^hyNhuE_9JQC6ZH0sCcTB4>u1VLd&)Hc{_E;G;KcRAoT+LmTQ*MZy
zQ#(r)9l8aLT_^ftPzn0kbI9U4c<2XCd>t}(g=1eZ031?AEj46uRquX1EUN%oBZ5?O
zmHkXwGwNJAxDrn!)bbX3S^ilVB@ZfZ{>MwFf(=g?fU#o53LFbOF`Po~HXVUYbhD|t
z&>j&*2649A#B)zoRQ+IYJ?)#PCJ-AEpCeXfd!!xTDY0zH2*Ow)`I7|)qIi;HYSe1H
zk~HZ&Y5Li(n;oOCg?<PucXgvA6TiSQ9U74jdi||RXV-O2DTYne6N@6V@MaGArw_`X
zrMLJBuKa^#yZIE`cXAF`%Mbb>@qApLXyr0Xb412fl1|?d#SA)4p^&s<3+_1fzc|~%
zS%-J}y?(pBjxOx=g=*|r<v1s{+&CfF(Z2<uBS)7<gGkKW?%KN#i$IdceIZ-LcU||$
z37s!c>U)hbM|(5i^Y+qtecEpej-J2bKK~!l!vs6_JS@TQhuxtT7Fu>rvV@>C)FlAB
znywr}ofN`iVh+en&c~!$2jZrV;Ds~sb{hMBT*J18XCtMUM0z<USOXDJvuR^%kFtiD
zvn`|oHO)0_=A)<WU&@-Aao%D5UGiU|DvHh&M(H`1`BIH<o&U<}()!RdcqxBPN}5Ne
z4I_CXNChzWm+^*LjZtxh9$=-(jq+!q7K_5qBGbm2CjwLaB<_1FUibC%AUVyLdM#39
zG`q4Pph%qHg-(aO>Lok_a)J;X7%GJkXaV3J9&vlXa@)M{DQWf?Xbl+Pe4kRj6!7cN
z!&Y2@`MQM3E9YhH^b<sOm7wk)!Hj4Rv4^-D2X@wt0gBM2YDCapsZl}>d?{H4hE%r^
zC7h&shMH@y1R*;B5SqL3L*^YgH$znr)@`@T*#B_SzEFJKOi(3?JbQ6B<R2w)zQ~VU
z7|5r_TjCIp-R`8y_m!XXUO!oCu3k_QWG3Bc>laQtE^OvR#`>NrD+eE}8kcADMJ4EA
z>zo^Lesnit3afX9J&KG{XrsUCRV^_cAbp)(^b}`mA9k{*T|FqdF8P5BN%Q>y1yTs+
zO529RXSYDaF-X^T7d<}ltmZziN>l<7L__QVTtfRf3&Qy-k!{*S>lGT;ML=IjfM#Yp
zDN4XdE=q<!VC%i}(6=7$UFH~~_%&%aMyW@{EVO>E{N2OdewOQi3~+=W*kqU9d09>$
zVK7jk-#|}9;~(y|a6q&^(s)rf0$(6O1ueV7$)apLP)`s9>Djrk>vEnYh^=bZ2(m1s
z>uT!<0P~Q%+}pb+O0RIVV6Y9BJs>;B;990-Qya+wyt>)=a<4BEH~9-9tY<oqIn5A9
zBs?$WK$Jp=R*1xS!n645g?9EZrX7NAk+>ZqgG^ixmAqKo)(t(>!vo?D7BXRP^C%D7
z_^9$Vh7gd<rlUao=!=L~mh@!u1p;y&@7F*E9#H!ldZ8UtRM}1+kvHay=9nB)klUJg
zRlE$u*)ipZ<hq7qF0sx_t{LFQ&R}F~JIpj|<eUUOhi7Ed$g%J0$K_<+kgn83+EbFq
z=yp^|M1)W{F-DKaz$^`R02nT#Of>^K=#uMN5FU9j>k#K5KNzT85Af?&CRPXFg=xXD
z;MOxGOlO85dQN5DKc~R=yxd#=_AOWry1@p>7nOsM$S6+y2AR@o?@*N;93y5C(3&Te
znF5CKS|IX)qplg+(sUbNY`=Z5@ELqK2U*SoHm@(cJh?ex++ugr*xZ%Z!PV-?2VNE}
zk@>t1I#9V~LzqdLMG<ptT@oCwPZ%vKVSO6v4;Z{uqvB~wdiK)zv(HXLwsWr~iIbhp
z@^PdmmYD*1zx<qYd_(=QW~{?uzmPT-g?#n6{befK|Mw&W-*${;!T{t@8D%cojjEUM
zJeY90nw)Gw|1R7{kr)iRdx`?LP~!^|(z!~8D$QmiJX18V2#e7v+-VyB)6?&JHo3sC
z|2b+ZF&$&_R5$CuW{B6jv1@)H0_k%65&R-bqsrT8Rt97EV8f7QPg@;h$ys*Rnf;ug
z*3p$gdWG?ZM(G@&hX~`fVtRXBs-B~!nI+1D|FZeYb*+6p^HG#*#$<Z!dwlbyU@2#-
z2nr_g<x{xI=Rbbk9g<(?nvdp{4vFGd_V(%w@XAriZ?qZbJqtuW^%BTm<NZJxUBbo-
zxS6^GY61FzA|D7{^2MI#zx74SAQwP>xQ<BUjKe3FqEd>Wi3$}oTA_r$m4b@jTubSe
zne>j4|F?m?YA8n9wEC5p9y>vR`H<Py)57?!Xq_?yyDMy~Sh2=Xb;JJks^dkJKVZ*{
z09&{ldoTy+;NifkeXBz@&U^7J5?3aRHo)2>PMsR=GVNPKW8*0738;_~{Nf=aW3`RF
zW0y|#xGJO#S4K$naM}0YfvtQlU&nGZnw*!`#~@B>^cdNXKgw~8tsM92oR6;%c*-9f
zytAhYwQEeO@#R4IPvfpSixEH)TYp5mBgeJkTXUIID!qOleQ^ClchsH7+ge>a|GMwa
zim+h$;p5GPwQpVbvc$UfdJlWh4o-5V+6}YBD@gXa!J4uaWr5~xg|?eJi*XIk!sPAx
z*0qb>^^LV!7v4Fy&1m3r+h=z2PfZ2poaQ6AGGYp}1X2}Nkx{dUPh7sNrQ>sxD8%S8
zc7SF@dXn%7ztf<uF*~u%H{<gOtj_cY{Upn53G9pvH+E~7)bEzF+ZD}9dY?*Gh$B%?
z!DNE#dCeXRNO#JfX2?JJ51$MgL+a1^H)(=wH5Wgu@+EEP<I?YR#3t3@U-T64z8?wg
zJW_v19W;H*86hiJyp<usK6s4ND!qGL<Cslxr=jcpEF~xU@*sR+suMU~aF0`s1nmNn
zdQc6_643-nmA6&GSFK7|TU$`QKKa<Pe!ky!G$qn{D~B5N$vg{0-N+bk#f#H$N!5?p
zDLzN#EHbEaP$+XDK+#>djx$0U6Aq`ZrlTAzudo<TJq%))U&@X^>d#aY`Qi<Gc!p@;
zWUqg*w)a0MU5v<kc6`i7O@{rWq}K)l4sdX+dHPUJv@!y+a&}saKU5(MhG)b)Y`r5`
z@X!2wJA{B7e}>T<33rHgHaM#BEU;rF-x(-0&ijc18yuF5cM!iQox)gUO-3xv#1FYa
z&+4~#zcEpwB$*24mLY}7oh<VBi+E)=ib->T@vD-BU3NKu<!oKcSWnocB517<RYfpM
z<<VJ(peG+s>1@YR<?J3of!0pw^fmrUAAe95%}7#Ti{=ra1peVXJZgN%*N{C$V{zs1
zsLqj8g-{t!WCJSz>z|K{SO;fGiybjEh5^yWHHj$e7!)OZsX_?r8929Taz&d+a2`9I
zC#Jr%qVuYERuj}hO%|6kKw%UrJ1eCQv=IzDu#z>RFPu7m{{ydx#CUf)%P5lz7IHhT
zt<F1xyf(}7VgSwE*cNxAoTdE7WKoqe2n&i?kAn84dCIiJ=IZ=`gA7PiSa>;+m8G*A
zc=f^iZcruqphm83uSSIhFV-XUy=Ew=()X!BjsJtY^ul(|ZlU3m{w-PCV%aL}*R<MJ
z_;+|Jlv9mew{tp$S4DvJ5BT7>G#|W!f+&V_k#-4BDL2t0=Zv#~ImI0N3DeMa&FloE
zt8>3kD~%U%H>Zw{);Ff4{3*7=Z37|sIY&UWAYii;2BKtGiBf%k#%iK%aw`fk2A)s@
z#Y9ptI-dgdC$*mmT?lspj`1$5Apg%dkLMc7ar>AryB`&2LQ#<d<j)T_bcNV6-H><<
zmZhfr;ILO%vLTu<rFMa)hZeZ_n9wS(&yV9Pvn(bj{?tz3d~dlypClRmHA>o9I7_S&
z-i&QykqylswE+BkmHMLvvh$kb_rvP@_qS}F?1kUKe?9ygwh-&SOaqA-`-8M~!;dNk
u<;#M<5c;jmW+dob$27wKb%_3{dYH`zH}lgc9aHmaM5{KeOjy3-*uMc>l^weP

literal 35849
zcmeFacTkk+_BDz+>Wqrcm{3$0M~ozuB-t>dHVCK)NH!2el4OuL!%=6%*a%2AAtFg~
z1|3BuBS=n)<eYQ3Yd3n%obT4XRrilu^;LbpsX9mLrr&<w=h@E=YpuQg>$u_(_9dK4
z*x1<E8ApFT$;P(Oo{eq6+3y$O747+@I{0JJIr$^M;!pfHekjcqfBbOisJay!+oo&u
zzj?Qt`}Nt_er99*df=3OKv$i^y?evHf_>&9#rqeq|9L|1#3|vPXD3crtluvFZtI72
z1p$nJaNl)GjrYYR&TB;r{ng5;5asysJ0BjMt#5fQzukKHx0~1Z^l+|G`&IwMfdlFq
ziBJ7gUC*eccAH)J@!761r^3$xNiMx3r&R20`XhCv%^#lTTf2bnZhu+WPP=*ZH?|0_
zd4lxkey^2%^XV_AcK-kRLWfSze76oU$B`~epW}O$TsgRio%7yp*_nMF9D)LN;>Cwe
zO-(yGJ3V<k*7E26v2baw)=v=*?u*!S|Gu&`?!t#hmOk^1Pvtq@;4!**F}F6wDo^XD
zc(F^LUJ8qfUcP-GyS+S8bJ6mR_j%M)t&Lg={MA~nKghT#R&Sp$D&o#T&-Xy?2Yc3|
zKfmYpKQH{Bv>p$<Lgal~MrzE9oJZ??&c&SWu_S{~edd_5l<gT$g|?k`Px@dX52Lp}
zBW`wP(kJB?yzN%#`)X<5U-<PUcAiZ|t}|Ba!^wV5i_a5$xm{-Lle=hAgZ-DAdsXQ8
z6pPH)aZ^ekt#>pva~=D9&FI2~9m*njgN=9lr-fs0CVR6Q%)09A8u$-wrB}7L(5qcl
zaWywtX6Yukja~>4x2nEb5Fj46%!_`@`lcJlr&XI4rVAP5@A$_De?2m<H&ig4T+ngG
zcj;`w_r=V0j8Ys&d*~+|YW-Dx^PL_&e7Ko$Qdc*)K&(NEvDdDrmQO4BMSwm%g|*zv
zrFWjBq@=9wTfT1h&sz@P*b(V3<9yloIX%tkJ@W+j&&W86sdM$2r8Iko)F?kaw)4<$
zc#HKsM{C@YTxVy-r&?uoQ>7gSnk?^qhqqpw#3IDA+fNR6`fReRjj^6=_R$qM!eD)D
z>!*cB2m0S}`*=@|_6946)d~BXCaN==x|13Tr^W~KR5<Z#`$KX6Mw0GHnVB+-*w!G&
zZH(*L=Ym{k#$!J&q;L3zz9H~c&C0%>jM4fK+$UZ={uh0!!Cj0oQhEM%Nu|7CMtq?$
z>p|AzK_r4+rF}TOLBFl|A%odo8fy6A(Qo=;?D*gVAzaRSW8bzM+Zt0JJqPakqy`tq
zz>5ET6}!ErhA>Z#XpM=oM$lgU+*H0F@VPS~_`-InPjwf=Bjxf+pDJjEiqNV?+;HRg
zIWXx$<v9<2xxC)Y$sUPF*6-`%_s1PyYHDiwS65e;jJ-Jf`s9w)%<st~x#hwZj+PV`
z-+9vARmam_6=x_~rRT|$1DgqMX=$<1*?>#$H~cRB<s$R;k`VP+JqB}Ro_Va(T63nb
zn3&;6S5;@cD1OIn%};&vH>AIP`?ka~S|_7A()<T|&nB#s=62TYSzmGEc;Myjk`~Fn
zNg_U8%Bm&bMtwOx_;Lx1Q1#L9vikKx2KQoC(zOzC%e<RE(NgU4a^pC&b?6^|ho@GC
zVd*)%)Z|+WS4`Ki#99ihlX0?V>S2L>CGE_7=MIeG%caE&HBDlEGjf<1YS)h-Q~A`3
z%WLmG_vX)%fq0|RT($4<;(lJbMCIkqNclYFP(}XpTk#^BD-KZGAEt>_PvfUr{LQ`R
zZZ8WKZvE=hEBDeQ^9G*01@%~|y=rvnJJy1B^CDzh3IqK)bnG`5$xNE)I1iWQ`wDHJ
zo*XSIl<W;~Um>%v#yrc}zuP1G+%1WYXJrx<+%tXGhT1-CVOW+hw(GyWu|B2!ssBke
zwNc;RBOz(n%e}T;l>wJK!m3<nrzLBjp5|+rY?yVKfH4P8HEr)M4|JJmr{JXTqwRh0
z;R}J>sLLWEB2{OvZ5sHftfZgA-0CY~b1|{{Vz{Lbt*XhV@6scF8f+~pvUs<t!px~W
zLfxV7)#AZV?kfzUGe&Z&41)vB>QcSRqyJLBQ&3Q_;iF0JxtRNS&ea~-ib)PlcaIF<
zVVV}pjvYv|xpHvx_JB(V3-^7wAvNCQS>@Q9QDKnM>7{L3wN*)a_!P4##7Ac?)a!5m
zq<?eA>Lu9gW6LGl9++edm#^P_X8*;D7fUKD?;8d=Y~#>%vIxSnPW9Mj<z-FfW!)P|
z87!2n*%TfVvqyJ!qQa9$tj=0_Vz@I&u~$ApGtordWjs&U;QgJ=fs5`0&rY?*^}pTF
zn%Wta*Y@FYL1XtLxW0_|samU|=0CUG+Hs!c8ZY<7ULESJ5X3hcj8q#(FXI&6QI~2X
z|M2+Hq9AGby!%rXuG4DF=EyawqBV&&m6ewh1+@qBg^Kj*(;dWd*d^J`mR>n{OM0X-
z-%o^BFmtR?BI?EEZhUWhSvWp8ulLb!9!MpOOPX>#tqMdcnv`58N*4`Bl3_jfU!Sh>
z)F!Kn(vYM&Tg@Pc7T^3JX-W7Uv%@zz%bq56n<mwjf41mJX${hH>`oBxjELKNd8kdX
zD8|0Ou@;8vFp#rd$aQAQB&|D9e4_1fO}<{&b#1#kk(_~%`RwfMSl5>qYbU)X+aBLm
z4Xclm3gUmN80ZIkk}Q%Q)oRK2^?c>N*f9V8p7C6@%#zQ1U$DQ&0`~23o$k~XvS@tc
zgkTaD9bNS7#A8eR)i94gscXmu$G)}Rv9|zc+uv8n&|s|Zvv}JBcgy#iuSE>)Stn+u
z#nihkJ#zZ9Eq~Euvc;F`S|tfJIG^c}8o?D(efv2#Y_R(98;8CJf8zV7w4T%h(pMD}
z6cDtI&dy9LE62GwDURjxMmJ%p^GibHJ$Yb|8rS$0v1$)WzPArQCZAPxjoW3Pny#~r
zzfGPuzj0jMbw16+h*Fs-{#N_0n0J<CDryPhJ#YS6MNhHOuGS(%bcyzdiFMLQ0S8>d
z&)IZz=YCP=TWd8vT+xc)Y#!%Q9dG0-YN|mIN6Y^6mG$D5dj7*%6J^nsh|vk_{Y8vb
zx$HmvF4gyPL11~jUJf(Bw%Ujem66FnQ;x5cgXP=X+*yBLaxM%Ir<ZN(9a_C)3T<p`
z=u+Dh{O(=7x+rpYYTL8Nw{jP(kh-HcUQ~PjwzRjud8WATWOc<rU9XE>LX9?`dMuAC
z?0GNl^#-TN^Sa-DknN64s?SwclJ?4Uo@{x!V1?H9BHi=n&$~?4G_>53?0FAE+k57v
z+ti@!tW_rN*#J-D&=D327vi&E`M#frhlh<bf<*ND8?#f|9(l-pHe=SplT_jRWAg;k
zaGJT`BIIM?X$&e~oHZRWg)jMXTQ<n9-eG%<S(>SBgY&3J$$(x}tiHok|2rY)iJ^_w
z#RsaaOBG}X_aKm*mi@s#{9~D#ZlIt`zlV5F@)x?HgpxzU*LkDfKfK1}cm9_6Hrxm0
z$Xw1tAGBP?n^SO3t=b-5d#V&Hqb*Tx9G&6KvtHEX^ySf>jJL+q5yMe!{pp=iX_k}S
zNrq%*a4vDp-kML3+~!h`zi?t^z$dGzRd&{)P-e30#s;ydq2XyBx+kl*>kain18wE9
zT&7PdD|`Otu~xtF&0i@^OnJ*MH^mHsokwztA0DdtNS31C@iDB*B(A`+P`p(SPrq}m
zheNZMn&r%RYtm9bO~>{SA104uPl}-rSDf=m)ppPAN`<`9=_O{{?2OH`rfutyh%K)e
zZZF-Q))g!7Q>Yf^=jWH2-W}xF<Lka$V#f;Uq2vDjSVjKZG7}dP>uhU2mX}8j9kl01
zda+hO>%Nb!i@3Zu-_DyWxTN-JnkJ~EiEf2i=MA-&3Yk{N!!O31kNa{<@@ZzKOXRS!
z4r1^&#emAE{Mo2QSdJyJ#w>3!w^Q-cpiQ=wn-OWE$7Tw%&+RFaopq9#8T(?}DS&6^
z$#d`K86S>z4ZOSMh(W1>7>=F7T0xxvoTP+;%nhdTC+O}Fh|VV%Yv5aS;CAwZWn~!y
zZ&oSkbydcMJExbY+RaV|Pa`-)CakD^P@8Jk+tcbltk{_q#vfVypk(!ug}r;89gWO#
zXcM_NeAg{W?k1<muI<520~+i~%lCb9o2tp0)@-*$$u-l@J-f#hHoW)j-~URV=m_UE
zU2unQ#SsCxQl6Zg?ZH}ihiymf1$5H)Wlc2;mPYGlY1vlkheby6M}~)o=e=6Y9iX4*
zt&bb?-F^1pa^%G&6VWci;oXJGF#HdXkF*|<ojAcV4sKbQcigT{M$^c0>FYjz{Mgj!
zA!sJVYt<TPS4f_Djm%ZgZ3=Qslg#m!K(o^7wVEb(c^u#&)YkPW-rv1hP1{E0TIb;#
z>+WirXC%nChMx)_a6GkRwdJdY98I@fr_LJ`2Ka83_ilRXZzdP(baf;`J+Azyf{<Zh
zKH`5`ys4GdphRy-dVy%oy@RvmpUvc@S8Z3+3LTEvDOP*wP*iHibE)Z}(vUMr3Eovh
zf#dMnwC)if{3zj1*n21~F)^{gm@`7@@<>;^tXV`ws!eB0joK+bEVy-Dn%yX8?q$R2
z3qi^ABwFsDFQ3RycAW@y?YNY__4|N5=ikcRjxni<)w=X*A-uNYa@naq69gXP__yUU
z@)?2BL5N`IV3D_ECfXDmo^&BhFqqi4*t`TY1Q>?5?z7I)R(a;371HsqW%nU#zfiH{
zQ37m7gb|yw&EGogmUpys*33kwc6CrQ%pxc<vHJOt`JwpC`rh9cEH?P`^7m!|f&x<W
zm(}kQ(R+U<!R6LX(duWf`{CBMG5n)F?VcdB!L}4y6-oQxmYe>yO7Zj;k;D=8j_$60
z6W@pSui3@!3FwR%@$0@0M#+$7QyQvNG(Fl7+#2jEgQeyh9y6)Jj@BE+P2bz8%+$NW
zzhVTDU(xW##IA-ir7fBhPcnzgHF;~zK0Z4UDC4?d(X#g_^V6&|<;qgK_p51HWUmRF
zkuuo<cdpyBvi$m%Tl*f>ZjDFEnf2D?orbG^Y*<*>YxUgY-5nhrvDWP+qRt6toln5v
z^u~FvXM1q?q;Ot{w(G4oGBX?Da=)&;W3^oOFaG}iYJP_M#p9;}bvMng-&K}+yRmwp
zy*`*fb+1)2p1MR{u;B4+8NZB?YTnRbmkEtviz(Jq2OybcHjjq#azq(-BL^)+MU^f$
z89y|jbz*-_mg}ryMMC=Bj*g-udA7Vu_u18$CUwCd1V#l*JKB;rQr#-$2rHIcW;cq3
zm>PARwSxiq9&lgO3>U8xGHaiK!ZQVamO@S!0>XHSPv%D>?g|{PW<l3+`5>o(oT2U-
zv1^-bPs9E{G+RYRt&pACt)}CkAGYuSya=7WjjmJu4se=-Fxn^Z=;BRx4n4nkwHGID
zAfwkg`r}|?HfpfoK-=n_U-N*t2A4^F_#0ClWlEWm4Dix^QbBZRxwziVq4nk$F1$xd
z<3)i;7yiEDG!-(MOYD6Fv^9pg3-=#Eun0uep)%9Ae3$ma;}=%-U1kqmIowrM>bPvz
z$t1Hnx!%%@t0%+0OW*LIh|}9e;<T2h#@}^6tlCCd%6nwGbx$8UZ0&%3rR=EjH4O5F
z;aQK{BNm#O!4<$+x5Sq!<{?c8hMhir`g9xeUF5{$)fJiL?c)-upOih6t{smoD=&W<
z6BCoi!ERh-W~x1A=4_3G>7I8IvhHQ;_wrD5$b=uWo7j9v%_Oz`>3E@CW@(72@x_al
zpMGD=itF8jk+*1Luu%nY8basa-Y|*DUK`>&h=R;@q``I8Mx)4YFR-To%iNvfz3GE;
z=ULaC8IBIcok@1RI21ZbRrPGA>vXh;jxr+PK!Lf7)lF^yPj17v-Rz_*YHMrHwh7yG
zd`z&+EdSht<LzyQ;3srNr@6eOcceWuSSQ476gj&b)2UUmAqUxBotg8YLx-+SCglI^
zgn%efE6AXXRk27CWz;c0;ho4gc>~&O9&P1|dRpb({IRw+;2PKE9XSJchPTU(GkmyO
zhYL_K3EB63vh;~Ld~?0fw%JM7+1?xQ%!jTmFF+ll*3W$msl@)qUyM6L4nDlYf)3s}
zpXES|l${+fGM-9DO!LvQWNbTj|IUfWhnug-PMuY`md21X{ejD=?^Qm)i@eDMC;s?2
z)gP#uzV=13H{FU!iMR4K-Sxik7|6svwmWDFoEt^>unsHq?$!{PczZo#AeYw{KnAn9
zsO;*xo~R-URRO5|#oM2F%4MYCz8;kPU|*rEJ5hpkz(?n@kwJl<m!1Qgq<-!_$_075
zx2HXaVIARe;Gu6Vlpd{pzQWG4nSWL0aHRQGhN8@PlN8c))5{YgqM`;UM^oU422p!>
zz15sw+8LcCW1car+WzPw)-vVu)eQr(vxBmmg3OB_9H4T^&)=W&8mg(~V@@B7Ii3H6
z;b~c}ZW#HZ-h0s2=BjO;?fm*Ehmk5n`8?gRzd3SGQ~m$aQbRZ*bU7@e7U#-$ufAMb
zX3&-+xAK2Fvhgn6=*_@h6i5YIZc8!s92|d-M)n_Q=9s-kO%ExTaaXNz&50wMyc`Bw
znhD0b!l6^yZW}piXO=ZRQiWX`9LN)Fm7Vz_TW#vt_}BK9j_~N7zI~wqmU8tjlikPs
zMftY}*`M>bIdpYdj_lNHSyrgLHi-=J7UD~g{iD}hS;2&zY|UBXcw@j$Fum`fkpq&q
zU2`RP9<6#pLPAQ5pRpwlRMY2c+=a;h@Bg&raFd7b)H&9pq2x5~uFXDYaYg7~ng(&I
z9<U#HH<@c2Esypflu1o?#nJYY=X-eQngN`Dm0g3g{Pbnbk@5c9t^xgD-nHTv3z2li
z5nzL?zW`~m@8{Ro>^S>II?8y-+w}lesl|7}M;yM&zQLQ4Yizh)*kY3-1S+3>eYF`$
za6x0XM`^fPv|^HW@};xel!5|~aa|NCUI_aY%$l-9RVkKOW#Io6G+2GT%sX)&?+_ek
z(;rI(-mf<dxQl8jy0n|iW$d%o<@OM9xbC>go;25|R)s;*lS19PW=q%dz5j58J5X<&
zuWf@_^20jOID>**%CvcKWm8L}u(tLQScb2Tz7Y3Rq4VJd=S5@VIs7ELZ{gCl_fmd2
zJxI5HWhqCZQP_8#?%t#87$qM+-b=A+^-lTaf;r;0k%`F=LRo2~X40~a`*t%Z?3E;_
zJYbrn_rDF#h~`Rv5mbfrod7?JFK=xmG*eIW8Pq0QtQWp;y#1hMMCAta^b;EO*uT6v
zqip!W{;aezQu9@c`!=K6evh20?6atOdjDZMSM=(YgJ(CXoetEH^`1{e$AgpK<0IBD
z=PlOj`}B&ep*oUxt;cw_pvzrsk|8P-{zEop?ZET|_|4(CzkxkkdhF$gfQF=?%$L+s
zL&7%R>=XUv`tyd5VaoD(udlN6eg%6B7YXj?Ad6*`JU(*UsmCgq*NO!ezy{~3p8Ndw
z6-vsOJTPU3!c8AmMY{LsPBzs>Bpf^WSFGRr<OT~njR)X(`mgeNz1RrVw{p09wrb!X
z5pQ6Lni^bE_LtW4!lSTVkbk-K#mdM9I{XAcy2<_WX`!e+Y(1nId1Sn1L+Wqt>sgrU
zRu;I%C4Kn`N*@Mue6Tf1vtec5+9g*gko*II<7ux+GG{({`0!zYJ(d@qJA)^uRc18h
z<d4hn6~})7^u+*l;=@A)6bco&FTEL~I^O)6D7_@CFQ{Ln46Ti@L<SNpxxC04;bVak
z3GrR?jq%3*5Zp*puaLcGqA4ex?~(ze$uWxm0X%2w-NlSH^eF~OrNm1)jrG}RtitDB
z<^pak5&dLZ_vmv%g$^%2JY9Kb{5HJh=7mc>`LaG8UN2-VyT4hg5fHw-liz?xHlrH%
zB>hFbgGI8zfS`ED=j5Fv7Uku2x@iBhddgCT6Oq6~4f^4AHr2xbm=lHUTMPVso|wK&
zL+us$k9S7+Ie~pBcVtOgw-vK?S<Q$(7Kkl+V(M`ftJ~_Orn9rI+?BrO*(t((%|G-_
zbVmD_-QFl3yKyURUIe(DpSu^-i#0gKSv!VDChqzN+DryC)Ke$Pg1(RW?hT1I8!QkL
zR92xc;w{gpUC4Io`ozQppGN$S!IpwN98OQZ;3`dBUEP*k`W*>_MS%JV;p0&U!nKxP
z;~o!Hsk!Qekg`BinLWq@Y|qb~4kk=gW)3oa{tdZ@o$mYxa{uPR0r%?vg8UJXYi_b{
z5RLdJ<R8481)|SPPx^$h%+iYh{l-#`QyYL}fIahlPLDswMW6&BC-}hHk<Q;7w8mS(
zdOW75qQp)B1q^L%ZSv8UK4e-i4q`P`7cE<NM_Exkq*fm<v04AMmJo3^HWe@f^yhKH
zxaq(DdEtM*^*DYOnaJR4?wQ>UGb2fZ<(lSRX&bRuyY7SA+?7w*R%p=WUF6hp)HPXz
z*RwOQ(r3E~4>uYgZw(e?dE8SHpioZT_v3U8f<?7)&boi+wxqPO@^lZ`__LL)&mhiS
zuu414=N(2b?h0gmN$7%lc*uMI%S(WHy#DzT>bjcRwL~!&gNwm8DJALs_@5u|z@4CS
z5Ir9p^nsR{A4Yfl0DNLB>SWRL0<qw@i2A^bGSZG>a8tMt7&x|5KeDKw@Pn|{HBTP9
zB|YAQj~}JqJOMnV*6bLW@z;+^EU}VY;*96}85h|9+lNQ^tD6c=s@oDIy7*kg61wCa
z@PvsQaMHfr?2oLNaC!|b!p|tNy5F+Sgf)B{+nQfjP)Jk-r^2@D=n-lA0^&!AQNr(n
zvfFeMU$P(3#O?aIMZ{9g-D>vL<pbZe0Jv>6)<q=10J$wG`Cy_gM2rx*eq0l=BW;f0
z3Z&a!*uEpbjCJRQ@VO8EuoiYtdYnmT^bY;}$a;0l#Si%DKhKQ%%n$XW*F$n4bGvrb
z%F4<pMKj4PMBkKNZTpk;Rh{e8ohL`SFOdytu&y2-6jmegNgD{1crI?~!Qb$0_|CO`
zbKh~~+dj$VO-~H-@*A_hdG1w-&R5~Tk)PSbx@{wA_ln=j@#ATh0o?KaISEw~K-mxr
zY1Y%u-9p~7dWx;USA?Vf`Dx!U)S5NhKCx(gw_9^J7k>e?!Vx5rRc5SP#-8n8Xk_%E
zIL@$0RLFiIt%~)pv`dd*wHkYzE-#8^`M0kx>W`R}WG;^Oh`^&2|3rKLXSwM*WGS$V
zD46H|D;G93mZO+di`EILxy-VNxnEOSoUS2^tH0Pw8~tPC0=nz?AHMGX7S@ZwJzp1q
zm$k1kkJhr|b;J|0h9w_9emtcS6&a}jjzd0am}NC{pNl+H`+vU}Uew9B-2MvI{E{RL
z4-UrnysC*aRAMdOX;N*R`ja3J@3~9y(#~v_k#<i49UK;>J`<|JIti@3*{qRtwhaU$
zB8bu2|5}k@zB=zSNzDh14gT@4<#E=U0x!zb?0R`}fN!4!*T;(bUdWl))-;w%+j~BV
ziV|SZVHFk;@d^>ZQ||I1llt`tKc~k0dSu4*<QEiJdjY;Vz7`2AkE{6yGFrT$L~lfZ
zd89{RQG2`Y(&taZBwGWl6oE--zP*S`JQJq1J;Sm#teoWecemzZ-6>95f)@~bBo}l8
zQUHGC%?g^k5QEuvRZ`|<dmfpM{xi^nZgM=d38F|lY|m5LwSE_w_ez#|6HGcgRWw^`
zU;;Ky|2dqGn;dHoRgor_D#=O%7P@R?0=v_X0AtC@7TtEe08{AgEenKjuPS`Ue##X0
zuUUr_(@34Q_8vXW0%s1DpdKf$4A|xp!Mc2>U;Ipr<oml@sID*s5}3Em-=xGWi+j55
zh(mui$6%-Kti*!F%lQs98;1tLRX)BYaWP?J8jk-CEQN=mDAE1|+|d>y;{b)JrUhNJ
zKoUT3toZOpWl}TPB`csz9nG4Z9_nq7l^xGfs*aQd##xH;0=xqsQhe=u>`sS?jW<$f
zxg&Uhio2A75w%zkl0x(g)iuCU2gbhK-p&BKpi3ydX(ISeIf7KfZS>3aq>UT*0h3uX
zJQiV!A`5>7j2~z=_<qanCM4%MD_r{AZ98Fliu*qQE5PFIdLqx_Aeo3Ru>4)*2`K*z
z$I(LKQ9w+yGL`*Te(Z}{fNPN@(2WofGd?q#70jPhyBA41D=p(YeP^fgQ>sbp9~dYj
zAFp+24eWlqVZ(+&0PWHYOTg?Qx>+ursqxPfVX0G~Eb{McJ``J~s%ev40P<p-YN(W+
z)J`Jz*gIO6KUaBp{5?qefwvpX^=78V)nA-B0I)#0Cl;2tLug*FL?6<6VlCS1GaLm{
z+a3*dl!eDf!LbrqCy>^~1o0a3dK&2el#R;8-!5ztwe2jAyy0~!l=xLjH~-omRB3cc
z6SWZNvKByoc^-n9_W`Lye%bYV*&gKWxTy^cCim2*#l&A_ju)64K*OLPg(||$QayTp
zeR;-dSx^fEDgX#FB<wH%EXjJ2i>LfEqlwbI_72pgqh)tOMN33nVJ^I;q|^)&U8~N>
zq!z&Y`H)`3btl%epr|TD?OoqvPltl3_imG&$*LaPvg?=Yqs$ce5s@uHSn5rU^*hW?
zj|$=LHiC*1?|Siip!m>=$oL9l!~YjzGgj5G$z|xnk?^5)*3E_nHoeM-Lm*J@X8|<x
z-Dh!5W~%S?$BGI=U;^U!D=|#Amt_fV!R@pV)k#nseJdz0LN!$<)`4w;>v_lJ7;!1h
zJGdJ~?+^A5<xCRFREbCT<bh@-2l(^AP^nTYF;t>pi9~FJNDLJ7jw=r$vslzC#9146
zRd#yfb?=E+(YG};=GhFLJsWd+sCxz6&X1icFKbFXh6(ZjtWalA7ptFFZIbj0=%LOR
z?~L;LxgEP=v~0>$OmB&U>Ll_Oi>DQqvERe{xLxkOo#DgWkJb#QYuEWP^}zm(7s17(
z<G_c+<alm31ap~!V@e(%I=-Sabd5*0mh1Fz3c!b&8%ijIAuj>sQp;Q_n7$jBQP;qp
zw2L-e?om_u5k1Zz8eP1A*@hQX((wc@+4TEjuD1B6M1l}xO7-=5{-V6G{x53lDvKb3
zfU=^1L=Zv1rc<?<AIr-0o8I5`6*fA_$;lc2H{2P%I%$JHkT^3~AW0sKn>OM!i+{Nx
zGyrI68w2o#72J~{aBl4_?{qbTz{wGr2yu;)%g|%)gp?h1xkRFZ1<=vK8&2KEx_KxX
zD?+-1VW3XK;kpHOS(A0c0$o!;Mm-y5XGTdWVF{-FDR6s+_pEg4SfhtB_+d~xdVt^b
zqgEq@fYGld(m8W1Td)Q2Ri4|t1?6L|sZq&|*Ee2ndngXpg%A+hWFRvI5FiBs_t*m*
z^n!oppZw0psUQuH&Xi~vdn;C5@J72rWNS{5%SasU2oaw|RV~{jAq=lWNC;Lc-S&3_
zwOX4BZQ2vE!2n?KRlre7QD9Q>NL2B_VDPMzVX(99+nXDbG~#^K^sq;3c_)0;>n7t7
zA6jk*6~0%L?7rtRJrRQZhh&Z*$E|X<`ZU4G<l|dQd&0|z;G%N`f{Krv9sykN`@p@n
zf?b(e66^!mb0<q0W!t`|24A=r@~d<pr=|ws3%rZ~<dXx!sjj&pJ5+pVd?YC=<M2!(
zQOzNj#MJ9Krm=&CTYa=n#%ylEud$Z*)gKFeL9PpA43!Y^t{pW>B!^Iulqb)4x4El7
z7)gGDDMM=-pm_-CHU}96r^Pk?H%De)RqxR&_i=<>2K^g&7&`d;6g+fM)=X#C)3C6e
zvopQ3I@LX39g5U42fQr30EjxB%UEWA<R+(=r;?0A<n)mtw*|}JXHC~+d6`uJ>gS&A
z=N?STn&w}%?bv(lRdq{eKZ@9#(ag~dEs(7&mqy+QX2IgmOBE&c!A}<<dd4am1K+(p
zP?0qw0xd~iHivFNYHDgpDx91E^gw6*zu08M0S@Nr&g?}pkw<bJ(q!S6Qa1v?rjYXX
z)+T>s5>1fe%q$yC&QNh)rqS7D*b-M8`u~;N47dHpZIaal$}IYS<TqzNuW)@&czpD0
zo7dM?6oGLDBgRe;Q1z6~0kw!OcZj+!Hyg@Msv}fI(FQmG?X`rr(St5*yhSW4rM%-Z
z@#Y}(qY}PE8-d)MIyM{xWm0(jw0LVkTy^X7ML2i818EJ@QFZpWxuj!(w0b>!w1#)}
zlCbb_-U5lM!zV_2>Ual6SCAkG2&$Kh|Btyj#ydJ65AXdXO|)hQ-{U!2(+T!Qyf36W
zAI$1}KoOG=hBO6(e=ikmdc8t6AC5=JWonEk?@%n%6C%JEhb6+=u`PUMWhcHDK^{gF
zr~@KYs@7|YfRyIkmPu9kvAP;auYb(>cND0!zRc_GkP_sVnDkZG91Uv@FC_hpzgtqX
zgE;J5(!`g<WH&a7?}av?r!Sd5t|xWk_gCP7{h(4y`(-JIKt9p)pJ|`LJX^OrZD1aJ
zcswR<Q<#?axNjA52KhMo#n2E-_9UN=EQ>?jY#|!P_+4n~nk84RJ5P^t=3jkq(slNw
z@ouleqPg|ZuHkTR@A3>H`c%E%KYB}zNt^;X4QMTQlk0}}xl36eVd@ZB1W|sOkMcX~
z@8~O0VSXADX+7}Mfm~Ljb#}TTc%IxIqN3Xm>IRpW6xku`(3UY%i|iR>TF(d#s`r)x
z9k)UVZpAf<5ESEx{+xt<Rnjyu7|a4w&qv#~J-!^&M_j2}lKyPeNNHqrv=TF&MY&Rn
z66mr-S5q@h&e>1wJAU)EQUVd|B&Z;DELaHX+foDXHKctm`K&?ywq{(SRpkD#dguPX
zi}rcF$CKw9@9^r0bLxAcvnahTExInP3>p3McWD#9QYgXp=Mi^^t(PJWrSI>hEK>-T
zoi-v55i#JI`dq{r0v5&$my7Ka$VkIay@rUwlh5^GSMHUr^7x{Gbn6j3YELQHrvPYk
z9K1%19CTDy7e}=@00uu+GTiXHcQ@d;_`kbJ`?V1P_#4-nw3(D`fV@d_!Kag+T-$tH
zUv(s{63RB$*-_UbXl^O3P8HTcN)N&(Q<yhPi16MOwgm72j*qV)LC=K>TZ)S*aIL(y
zPhTJ5cD`p0p+v}v0TO6Xw9^i%6NM&FROq4W)E92M>JnteP;IrYa2-3;K7#*$O$2_p
zx^punX9K#k!@40g#8fCs^}pfMP89}&VsGRG)z5g}>lJw?8kev`00%?WYVTcj${!p+
z&Va}6q;#b#a>O=5>25p5wIDZ)s`pM6x&YEhUaQ}V>k&ZGt0$v;=Jgs*uV}0Oo?Gn+
z@?EL`1|!X#d9e=!nwL+gglP(hCO(o)56=8kN$ob>7EbP-a9yqSSn;^j(RN~?Xw~5n
ziBPKoyi5X19GEA^cR(-9l1WKO>&h3`++8Crh;^e40aD@9VN#_SeDK+FCE01~YuwMe
zvSw#^+3nIgUc6qO2U}F0IJ_TjWsq3tbFb_4#>Ms_8pac`LRo{N^zcWOiH?CJ@O0_W
zFOcTL3VOHb-L(o#bEhvGh}9M{DtY46F0Y%5V2P7(@HdJ-cOVY<1=O@*V*TY#C>5vV
zna$F6f=J1y8sJSq{c8mP{@(TG^7=ko6{MZl;$$?dIU?lwg`P9lB&|HHOmY%EhobVr
z8TXN?!J_DrJx6YB^aZVW!gQXwq3Wm%!lFEb*$W<)WR*3MAjI1)WqG$YgP!fUl*JRm
zn`D<b3nmaE-r|t;mHcP>RiPm&G5vkO5j;oD&9btxX4?LWN^ukzccD)fcAXgvjyv<}
zk1hnW@H%c6?P^45`RLN{YM)q?)X*b2cmuvcm1k`PtxuPsudM6LF4vk6&ecn<1X!0a
zD6$7)n?!rPNase$FlD4e{z3o>QWIX;v<R+VQsp#Q;AveeWJA0OGL-bFyo;p^Y_zos
zVnV>#+tpWnu6S=)7*KWg&Wm4gbTyexpkzraBkOo~%^BqBc@+mGjlC`*y4vllFL50w
z8W)18O0#K_Mqxt~i_&l&wlYFFew~bGF}aIQB@!MVF9r2k7vM*dyTyVJi~KMM577);
zFA~e_zxn)oJ5AV?(^HS-Hd|{=PwZG-o_8rJ0riyKy2iCj7JdepaOw`bz$XQ!pmRGm
zfR+t)s9bFNyAI^_meod~s5?5IUq<X02-lt7g=m+b?l4%rF$>Cfh&t`VD(YasK8Tw5
zZaY25&I(ZhQRN#~D8Wlo6>Yv|n0r1X2PwJlx?qU^3r{@J@DH9?t-aB-IX-nXZ0qg}
z*eDBj<xCb67H+z_!R$Rbc5C}gP<{G1Q&6515{F3T?NB<g^M-*o<x4tE#m&_WA10Y>
z|8%9Jc)~pg6^OuW>+DqU^Cf2hM4TQJeZQ8SL>?sg0co+d0TPAk2!j2?dTafN^{Yi9
zZ7T4o<Sq!!mV3&I>gQXveJQrG2wx;<PNQiJrg3Jd6^XN^TtTHH&Hm$wBl?idQW9(R
z`3i?^Y8^seQngLAaky1*B2*w<$fq4(uY;6-btg*qzgcB9T&8Ac9{`$QE6MDhD)@V^
zQj&@(NZowsY7_U&XYPS_vNJKSogwdh$CHD%KAgSXpaggPj>!YM;k8`S!rX8uYhY=N
z*mS$e>jG~9o4mYw-hG+9F^ULM<a%qG<se!FkeUUGruQ6KW7!*RE6=82=_=aB;e|qP
zY?2hE%_XoqeZ0k={3Q-H8I?X*>fD*h-e5EwSOn_4q3U6LyvQ|3cDh|@y{wB992{|z
zttgF5R#_(XtVrJr?Fb3cmZOfW1p!TLJ*g>&x~hao1_7xpsW81S*QmOer6}RIWoP<t
zQxa2)5J?QSX!V81U07D}f%jXeAW&RTPnj7Mv-xnKfp8LH1ib@5umo4y1OOx-lC6(M
zk<`#cumW;$1C)Zi0qmKVKL33zNJ<=)omOv_!Lm;Gt5>g{?DQ6~IY6}MbKQrP19cW-
z_aF}0PD_BSPzX7brOzP_+MtwAe=M~mr6rUAoz`ij$t7I`2_L$UJq5Q101=%zHW{r1
z+=VD+9E;@FGvzOS!72;NaYt!GX9@j{vs@?XV@V*&*h|7|NR&o*ac2RJ9eP*?OePk-
zO60i6DQ~iGcoS2NU9Tn{G)o_50{GMKsZG(3`o4O@zDs8yEzWrNpwYUsoFB;5vDhii
z^8XG~W_P6R)oF7D@jS;66Dr2C9Z#o`0Gtp2FxKr|gZ{_@QPArn!w*@SGtVn>_zIV7
z9=8jYq+D=*J$oh2IpI3`U=c%>X5o;wz>($|g$8=_kn&ZgbWM)>C=yu9zkk@S2Q|L~
z01t`U7NCUhQ7F@^5v7Sgzw9vH)Mp0i9E+%SlZ!nt@<~4h-%SPS_;jyp!{-Sk7+a2P
zw0I*CbzS5$6)KblP#MxKuuP3_?I83Izfiu8<<@+<<oBpuYdr``BFx46Z9WU$z2auK
zKos7Pgl#D>tpR9@sKtPYVzO_lJE|Zx&+myi36|XS?+4}BE`hA60UNEKUZm<ENZMu<
z%^|Y~U#9)OVb8f=Gm*u6r!8ZX*aOqGJ%9Nz5clGPRw}?q;Uq7Ll|z=c_Y@0`1gE-x
zp+Cxf{~A=R<uZ$bAFX78rcji4dE+1s;!(41lwoy`8xel|3aQX3jm*aDyjo`n6BlK@
zsIfR52%W$jZ}bZbc%}vgZwTu=d62j8%@Lbp!mPKbsIr7Ynh6%zakwLa)sgccb`HW`
zZtuf~<{z-wkX{i{%j)mh-%Pi9Wo#BblK0Doodx?w*I&Ks{L?S&2iE?!I-C9G)nws$
z(Kj`2u8zE_>i_G?@NKH4??oy^{jwu1(%zoF;kQjd<fhTh6r+zHcW^dxCS1QhkH`Jz
zUw=H{J<s62@8sg5fWnu{Jo0y2U9ISJun%e&j|>)n8DLf4Iyvd&ztn!i;XZ(L{8tV>
z)b7B-J@`U*vp5FS^%@^B(+>}%iPR_{aCh;C<b|)nyE3PWj@G|@>_`yu??9U+{#4km
zUi0GYA%L+=;Z1c!X|ZGNiVp}cx2TBL)!mrkzmA~%{e!GmO(b&s@G&;G$gfvj-1~Ii
z94;?>1kmxOU%&jAl$u)oJ%Q+n#LRB`)xp8Rf(0wTOeCga@rQkHBE!PwIIT$59e+q(
z9Q_kfASO$(6V^Yn$cIFN<t6@nyV(rX%k%JXkJ||s@GB3*@owACiNb#*^-~Ygg`WYB
zFF#kxDCfNWX^w7cA<o6U4SfI>4yc$}Y{&~0*2i*lbFW9P6W#Qc@0eSglR7#<Z+M7S
zGuc4j|1<j*8)${&U@MK6S&Z6+TXZjXuKuj%K|O4V#Gy&t@zxNYTN^7{o8N`MOTnG0
zr`dKJbkh0~6-L*x_tYEiD9hHzkKcBqB)KB$AIk`u1wC4hblVGzHugnnBs@>%qu1sh
zc0a3=>EG`+i7vM%#pcKkIv2~)t;DwJ=$D_;H;9_-5D+t~!?(=s;qSr~*B_>*OX=(D
z#Nvi31C%`IHo0Ea?4Q-fFKH0Hv;0Yl>N3`XfBVs`;yYT4c%q{=9krSl{R5F<7yhDW
zo+B75xo9!>`z^Ja+mfxUA8-8QDPSF-oTTU0h2tv+C(rtkwPY_xSf)^-b-zZdjf{?U
z^?L&Y0|n(qENX7<m)lr0(zjpsSj1XGewOW(XuULMQpovs-yc#Bv#22Q6xKf%TuJ%o
zk8?&acSHYm(b@}j*yjY6<l)|Y)<>J0n|H2WqVQigf+V&;W3M$jE(1x)@Zd2qc#GrO
z|J44+EfK!9?KX?dn_InOAb$L%=GbJ`ZSD3)4Vr3W97_70wl-RjS0<dK$+3M-WX~<u
zZstGrB%*IWYYkpOOqys%s=RyO)9vkE%w>PZ!U4c@H%k_^iP&w4`=ov%y(j8#Pt?}Y
z8F_S@3wX_Y=d-yzO||XzgWzj340I>&pf(?_#y=SL6Pv^i%39epzj$uJA8h;o!g-{>
z+}pu=tp9o8|IGDxHZVeF!RDs@8oDnlIGbiFw9T(nz~#)%1pqkb3FTsf9pm@_%ozW5
ztwu(Bgxc!1F$kIHb)B8u$!p~<-u3;gJN?21E945)OQP+Rjx3VO=I+x!SubaKAGYdR
zfGWx=DrI_gD4aejmhEqylO{bVfWF<>KhhE9Q1!!=ZCZ0*7oP&Lp;@9CdZ&OOn_s*5
zU=Th9UsK$I&PdtSKG&A5XY#10Sn89YZDiLE5)0>S+VnZ#Y()QKOhP2rOWK-P)DZq0
z&eyj3>!rgW1eOF=z+h6>%bP!c$~GmLt)th%z(2@uBf%)!_hu8a8c#PnfYIG2R$6m@
zeT-9I-0Z*s75068?Ezo#;{3O3BK9;6{Q1+Cy^u3ocUOxru4g#>zJjC~IC%aSrS4XP
zWc7-TjinNs!K8%KxIDrry5vC<au8ZSA*`PRDqBw;<mGbkEuO#nFS<b97ynfs)7RIh
zFwXP%@XcLOVjno;mw$a=1Ejmz&2Q)R;tN!?%%pzU@WT1)UHO7s=kS3*P#do!r1F@&
z9AP<cwqgVPc0%L8fyJ9nX>p-VHaA!Fu=0GJF|>iw_r?E(QY5L;<t)m!D%2SsHM@50
zy1pusA2<)MT9nD{l@Gt&_bO`Wv~WPC>~GqDy6k!$2-B*aA_ok?RpU|*%E~G#T0vL|
z$OP>^`?ta5jrBrLlcn(VfhRevMdzMRqmIB<B!F8P)SR&qy@USUrODfS0l24@UUV;{
zc!iY=2Zl9yJ<oNz!c{xqcw`I+$`E2`L>1pA!qW$;a$B$cdc`ti*4MXK$__nT(cI*z
zWM-R%)pzPT&n?fSULOFK4Mrjd4ZoV9%5)<~)(!zB6hCX$tl{sGM1j?4`6VdhW#ZQr
zyQmti69{49UdG82C+<>zkY??y((I5DtLHZdgxHx_kxIqpZf!Q<(8NShLWBuo#n^`}
zcvwb5rn8V`(|f8OiT@#i^N4`eHEm#b)Lk~<&OJp0ym`W4Io1oR-=p{hmAvMO-As9b
z9FQ;6s`m8wj`(UDSj;D}tZ@a@6iglv6Gwr|Cnmf8_S<hW9nmwplIkxTGn-Ju6o4r5
zL9b1sLuB##b+6>F#aTAzPIWiTm_}tzbWnGZdZv>D5-<X8>BsT<(M%r@*m|g{y0tr&
zQ~C&kO>;V3NoHJ+n$z&6Nn+&j9eN-WNgz*jn`#ziYiWr@e<nMjW!bcJ4}rL7sq^F^
zJaQ0iBjPCj%seL^#O?XRXL$$+c0j}G0E+^bZ4l+-(7JShps(xKuM==v;3pCm5s?p6
z=cbhD0q6xu9li|RaIjjnxJiOC*Q2~6$@xa~9tX+e>ou|gbbF$eX&Zyogh1lX`QHM~
ztr>+LBx0f+aB^dv#s>)IBUV7jr1C`~-+H;ApVF;&5CsNDp^S8qg{{GoW07lz0BDr=
z5g=5Mx|4~J8h?L8rlh{v_9%<GK(*frGW+75Wd7Als3n&`5u(5(Qx4*$B*Nc%J_)pM
z{A(v~%}AROyv;b6mHSw;?TnKuDqhA>DNieV=3T%&q&<V8RTdLS3aOD69mB04Pv7^t
zjHVGAn1b@hviQK_X2F?8K_(CN%fz+XgIXCtwIoYr_AMagYm2gQ45o%cWeehpWc<B&
z)kxS*bShY+v?_59QjLm#a$8UWn7n%?1zRhG<2nWSjyUZ&2okX7+rdp-qV9wn>-L-r
zFM5~j#XDFUJR?lqgIG@=DEf9A23Q>KU-(qywy7fWjzX!yd~$oU6VbDN0&o&(4RoN*
zojWzf0exFxo7Yr83sT2m0$p+Ps4yCnOv6)Ifg3lsa>hax2f)2afjr562jnQ%HuyY5
z-PC|PgG)-#s!&h-)H~OyBh*%2&X=H)4yM`x2Sgd?5iL4TATjFJUYWLIUa!D&96-2u
zf=2)omFsE3ge-#j!70hljdszLnx*zxO|z5``$8UHaVrCA*29`f0>aVo{K9##4lW%K
zfVo|#8|+Gm5DLEt&9BX>)V5Ke6K9b@z#f5P+iK9=nIZ*MKFy&B#<devurxT3)S5yq
zASCmTJ7CZ1vqi^Z!4KO5j%y{s9I2s_(moIsN&GsYaMvWTFenNNR0&rl?^bh)8w(`d
zicX97*VpheO*UG8EcrP+k_vWK%OuEH7iJUb5~wi)Xqd^M_}oy0EbT3SCaF+sCC)rQ
zkvbOE7C<=&D-8e<r#UUPJW}ChnukADn(mZVfK*M3o)=t6<bFfwJ0+hu%!+U^)H(|k
zj~W_tsR@hTLJKutPeZZL2%7GSFg?K%IQ~-*DK$<Rgw8;fLbx9_Wx@YOXs3x%|Ko6l
zZZI`>eR+G+I40vY&<1Zp&dCdq799c)uoA1r(eUq2iW6!l*XS}%+w#t=1HQbFT2qh@
zXH~h`jSn;vT{rM>h4V9p;3HP|;lPL6F4q20LT0te`GJyR(K!DGk6*xNo^g!O$w*Ch
zsyutGyEx_6zQ#YDI+TNR3GjpUv2+1DLj#z25?L6hd2XKAVU2LrNI3-G@@U;)u;t{L
z9H1Q;#1w^)=g)Uh-wd$!0S?!Gj)JQhys8oJ$&09}dZUv*6w3hrW=C)pkrt$Z10N1?
z4z;6#`kO+~Aq2=Tz_7bIftX|#bR0Zc+d2>mQqy#$KvF2~EZ<*D5S}a#Iw2xoY$FY`
zudg8)&<Hjitmk8GV<YqV_1d-60gFiVqF`H<1R{)o{EBDKp1lX_If!<`W{7-P-xj~J
z4VeMKaxAe9IT6Ta^k2HoBe@H}!uvP4q$L2H9`UG{gfGdcN(J=KLJmp75bfM;%o_Lr
zWHYns%eCd3{Gi>67;`#x>Qob&%x1lSw$MBRn$lo}*p@gE<6n&q&cW@j1KVl0V$Wq?
z$E=0uoDo2uQVhZjg~!v3@1Ko7+wvjO==*EPE*s4Ed~ycyLhVFiX_4q39DlQ2CPty6
z#tWXNtjWfD-%ZWTdp<sCS8gbV8&^biYtN4li;dlz^}7Aek}H3Vu1iZsG-<Rv#Fn;&
zx}8~#jx%HLWDUSi)2(?_&<PSXsl2m>SJlMgPBj`6EeQnnGmgr6`TGL$D^TR0Qd2X;
zuB)SSaCzC7q3ejDdX|ebg?1jAeiDTQmc}tqnk1Qhp&sh_Col{Z6%~b`PTT;rSjM5V
zH^926XI!=+Gvc9f5H16r!NTkJCx`#k*VBlGpxURKKZ4+`@0x&d?^TlO{}saF^r}SC
z%p>Px9XIeW;?xID=1(pq=WX90ukN@USwZ)6eRz6?=tz|5&QH>4Zt{1Qp#u$yTLdTW
z7@V{xM`ZavTV+7V!Q=UcluWdc5CXyvj~ciOmWyXt!nno6qdsqV!++(|P2vD^-n!Yz
zlXt6y%A~+-1F}U@`kAKIN9lnq?_4d|WUZr}X1nH(4@N4dPMMT;^51)ov#vdHh;Ngo
zy1FlR;zjw)vX7wdSOW=YGjJ}Gk~33NB@kTeI{K;U7D}e;w<$)bh9fwcrF1+GCo$T#
zD*Gca8J!wEy<X({6&EWy|3tzf`KBoO@^A;8Z|>Q#cQtAEY`1)>%Vy`2N>(V!!s!)D
z%Y<H*adIkAUm456;!^0uQm#*vE^ycc$S1FbVh^!uA9Zsr$25n6TiQ5S(2LB3kMZjs
zk`1_O?n@lgxeWe<<89Rm##dKx2O&1_>1L*{yOCHPQ?MQrE=Zh2vKFE_JwErtl45ap
z8?Ey9_}Be$9E-qHQ>&B&BW5s><(ihxz@xtqLWiiRZ-t@bZBuywJQreX%PZWyzqhxu
zh_uLi<1L)IDLKY;lSnO)hb|N?Rme#!p;9y~d#*B0?je0@Adl4rhg?%wyaG$-LvG8X
zwYFsGYj3ly)p$i&UA?NYdxNB<i50bGU}}H>O&y@{f_KFj198FAd<?(9_(*I<PU38a
zXjhVLcQt>{Xv}Trh;_N<&k=zXCiuEL00L99+;C?FN0ye`tw-3o>;@H7E&MJyi$Wp{
zi-OPxy)$#$LBkJyTwAKfToRu@e;&yDN5L=#*&rC#pYQtHHoorFQ*S}tAWC&G;=m|2
z6Hbv*Ib^bwIjwasyZS=|^JcKC?9}HKp#ZorBBBaOsA{B!7t$F~BgtaOsZ5t#tUw|X
zrW|Sj5nxzUR3WMgKa=QZ&vqpK>q3C$JvrX2tBelj&1gyF-Z@BM{GAHBk;GwB%35U4
zf5{5Q$%ap6p?6tnGR_lv!OkN)$4wrg1x1s#Kyo0s$KAy}aSMg6Xs2f0asRAw{|JqQ
zT}UH!D8uBTa?J`B8}jJUFS^qmYC_;saol+kE`Tv1$eWgQJP?O#GVQlby2^3Y+f2Ob
z?s?6Db|qH{+5`x_f}7FuIva?UBIPV(?*$Oxiq&M{itrc;nZMOCuTCeHZwcS5XU;IL
zZ#mh)DdSZ3@nigi|AxvFC{?m&dto|#=JynHde2CND=g5NecS#gy6t<U-wH%oj3q?|
z%O^ESHjFFosoex60_hx~RyTy2AbXxZ?yb&*nLue9Yn%ui6ZOp%P$OyX&d?_j5qwmZ
zLgC?2aXR6m3LO@n^?UV>(xPA=3{WOu^%YUBQb?nOB{MT~UrsfWNlzXc`r_%g7(|bt
z)90n*aJ}{I+tkCu!g3HJ;}A`qnznFPnPff)G=)j4qcf6t7oc&>*}5&CLKNtDNUpSl
z<hA>F!EF4Rqkf6@_DJ|Vc}R^WGZJ&#)VB7yN_gVs9vnym!{2LRC^0t=U~G)$bfq^5
z=O~ELRFtPU`9)HL$Cyomvjc+u?5)>uiV+0qEFlW?supFTGUDM+^1;D$Gc8Nj{zUy{
zbl==#S|lUD7zy32Rr^lEEk)%Xkr}vKFwzNn`0#D6I=#Aii#Oi!5J=qtW5v!RzQ}u=
zHU2umO-wxrb&-jMnLiO)qYbXEG%c)A`%uQ^;auW~7@pf~s%1XHg__#FP<L#vOB*$C
z9Zh>)IIJ>167`-Q%?dj?wJkgbW->KVkyT%%ztyN(Fwx?6!&T7*{0ZAHTrhpH6}hKU
zT%0-5c25QqTb#6SiQSP~rVNOO!6>Us6=5XB7oME$GW~xUjGC|0Piyby<u|H~K099m
z&ph^kTCy5nKE<@9{sl#>LdgjWCm<t(qQ;NiF6n7x_2zqCUd^N(qzc6x3YruQ>0pUs
zF`K6={<FX)e>b1CM1C@u)PlNCQs3Fxc^5D5Ae5mA3VzqJP<B!1rJ{q<cZH~!i)E}S
z7vtAl$FA|@Py;I}m^Dda|GI~bnk8?!=5;v?tC>c3psRT1e0W7L>Q0&z)Y#1l<9`p~
zo-f)2cQP;&?VhN(tvrpa;K@NUN{sLxQnq+_J`<5zTqBX1dRCSNoJc;0?TaJH^tjCp
z!mC^69h!rzp!l4lhx>1pwzNEG|DNoeGE<c7ZMG}g{wRO)+X8+Gd~6UD?h|AiWQm@p
zlm=4fNW-QgFBee$3IV>R^H1xL%Fp!l&P3&;()bZ#?9dkL4O_VocYbEXd{z{)teQp#
zYG-FD+)0*Xl4uGcE)k*CO_lsKO_F9$bahhJzIId^<r|X+#+<J@$-X^@$`~I1x?6;f
zalfH?3dHy_D5GnJ?9Iet{)j>b(bUK_PjB3Z6COPkJUi%G?evHJ4x5>4#gCArP##Fa
zUt*pyIf;)b#DD#;%98rG0o|Al8{eY-e25AHP++$}YLW{nlrQ>uP#`y<1K<(}L{GBs
z#TzfZG)U6GAtvi#W$HAhfs$348AQkqs~G^{*G%M-G>StZ$d)K{Qjb9#LJvz+4Rvo~
zb%p+sTXlZa^j<`Rk&uI$^lMS7KzT6ydW2?DRL=LU3dpsH^hMd!Bs$y)(8(>5L{}8e
zF#$k`{57EzP5>erIxMj71r4*)4IHoQfQ+<kXc#+ub=mro@~ww*ct46&2(Px?LpMir
za455A|M2*oqNJ_nG!%_FcA0~L0l1WScYJALaxaNe!=qAn(R{c>y9?ASZiRs)O*G;N
z+_>GdOu%S%*KYM5Bf=XY1FP>dwJgY5hCKIQyDFlzF-NLA9<r$kdK>5aoViOIA7}k}
zo(-D6N1l#(#0sG3l|VG8KhG*80VthkHWkfu6;VsK+3<Y~kCP^OJ(JZ34Yt=WKP_%c
zs&)IxE_N-bp_bi!xCJ5h(_$Ib6*_I7km`9PC1K2hOZX(W#haf9+n_dfgYC}7`&vmd
zFmJ)4CYX<D*=5vBl?j_Lohj_uv!hHY>=rQB=Cz{`bU9NOed8B|;6#n&rD<SStj{3I
zliHm|q=KXVsb*-t8Y&@zlb|o)FFY*)bq`6!Aci-JQ~!;wE`q@vLgXKrPmq8B)_A#g
z)D>oTeRCfSF9`aO=mO(^5&pNS7_Oe8k{%{EjlWvqa!2nAk3~%ffV6_(shr@|=H^Bb
zfRzaRA`sEE!<{DZ)pVj_6cCt2LT8M;QL}awYFrzOapmBNQ~u=(1Mf1*!KBO$phI&`
zk*S)XHz9>of%FGZFN?7k%>(qO@c<s4k^DFwvXfOs13s?fKK|V_$P5Qnlu?pkT(J?|
z`I2d;b~wz}IqJ8mJyi(q7g1o4#7<>JgkVbMVfNc#LbNk~&f62|gf-;m`qdws!Sp^E
znh(H49<vQ3Qcb(f!LU+luw!#mwZkMeJy@|KlXRRkSk7RQMq4CHp;QZ0-C(`#E9Ob#
zPD+$Vfqd7P1Xt)dMkY%kCHV)=TKxtHfr=I?|4^}-87TEu`;5|Pzmu!JA+gnOu&+S?
z|F<$I*yDt+Nm5Z$3x%s=P?5a@q`TQa;9nse`27~+)E+H>&3;N9SbHIU4b<WI2E|tA
zL&S1D4?=)l0!98}(aF>X9;m+3de>KQJ0CfZM;_oo0Rge)_K0Ieff}p6Jydl*&8JjN
z-1Eo9jeVj+4GK(*Yv}2UeNg(95ilVB3!n!N!y7C<PS|!M&OXjpOSuESqWL1g`5FhH
z(Kv~{f~1m%L7s@p<_tM^=Jq+1TEUdFPXO|Hq&tk08|#)j?It9b{tMLn;x<Y$N}SN+
zP7$*LUF*Ja4Hkb&n%Adqn-@Kk_8%&M3V-Crgz=D%gpXO9wEopn$-}I6CbsHRIQ1SG
zwj?=#1pHxgq{ZZ-Gm~Fg04oAGEgUejlEI|mGC&ZzRaROr`MOfISSf#Pj4Y^r$k`_K
z6{}A{WHB$5I5Wv=VgnO{dQlGQY`=&N;6Ef)76e}Vx)N$YH!;_OgL?%aS0t~xlB(YY
z9&&Ib)`RV3)LB5f3?{a`8AX9fZte}~e{I6x0=#!7lyC&X%<J{{hc+5h=PhI#b8oWk
zI5Bs7|MLQCvHoX0{>L2tH#{C^La9y$jZ~estW9{%cXX|$s)6**PAx*gL}C@!H}4OH
z>;?}2ZWe-hLM3TzB&FcQC;>&NPUE8xM?Rps7GywiQ$H$fnoVW<e38V8uio(@L8Zv)
z#58A79mF3JDNgz7pp*%a0wT|B@(iBR%gxKfNYBiPi%3jr;)MTw<&QKXMoXsS8A|}U
zjiDNG`YugDQByD8`Rzg|BX$b(X3fs5hcvmrAO5zAYwLl~ucV$r0nI8T1rSC9vFdxX
z@Q4VL(B}`VzQqo7)_NkYi!t^>$4CqYGI>5T^Y2~%L*7!A?~#X$ScwTf<WaI2IAO^&
zp0-opSXK<mS?hBRy-_@rN)djocb-j4SGPeR;eg?{r9^Aiv*dKFC4LovDGy<Y7_yRB
z+WAKzjSV20*l6b`)AU+<?tgo7A<Yf1(%NPI!`gEGU4!a~klO5F8s;GYjL<%`VqqYO
zmf(>^V=ycyLMvr&&fEP-$9Jqw65%SC#FVDqq3Q>5U!QrK#>JvL!Yq}P)1|Mw_{MvF
zJuncO(w05l(V154`}~_0pwsT6X|ej(1z&~X3p)f5Jxt=d*&p`87Kl&_ROR@0w0=`Q
zzy59K(B~`vCcp2?y1H06U}kHziQ^ZyB#U*+OeE(_Br*!a)7Dk@AYaLLNxR9;Et3vu
z)W1I&?vJsAV5mHKC{)EwvQ~cz<0ZQ9r9O!JO3QZWfIVdq*&n-TZkPfTX-t~PD;Xlt
zX7Y85o&)ST7A%u7Hg=wS@^sv!myZvJ<Dh_Urc|DfV8i<61EDOH7TW_AX#n709*@OZ
z{E`5CEGgT0Lz4LzX)nt|4umS#*Du_^GY!?$AbO9@ErJ1|3qf8!sQCQCV|)kAQY1{7
z$kl>GZOfdk1YHp}h<YecgC^Jn)_~IXBi?Cj%nCsNT3JULBg;|=^AI+!v9WupaMAZV
zmGh8I*9;jNI#OdY33D;3E6x<?Ewz#&n;GbeE9x1l#x>j0Nn2WHQ?W*NDL01h(TK7{
z`+T%8s%p8QRiP$6v%IgSVMdfDDbp}b)Q&>k+RLHViv0Y@Lm(FfK~35X9>#x&*5t}}
z7nBK1q<_bN3Jl=QPfn0FApSZrKs4TxsYfj-;3X3$e1Yzd=7a7f8F~nEMI6i=G)BY;
z<<^;zK!B9&Yq)hGIrYUX+-o`CRf-@dihQtT;gkQOQDdSopo;YH$e>L3hVR4|cKr#U
zX|2z809U9~Be|fp9%yx)Y0ZAuFx_Jp2SiUa%=`!$J(BfOgHAL`zNPk}?&Ck1@O%n6
zBsn`ZNMnCbp-M!1OA31L2CfJ?Z6*P8o?uph-g%{}zgBHGKvlsX)QWjo)C%cu$|xS=
zq{=aKcDffueqj#K`5@|1r?HsfGsgh+u!b9=V^oN^ZW<Jb%SmNTcbj9JRzssj&43n9
z4%JgUM5`p7euwd4ReJO<a$Hcrbk2qrXW&Xhq>N@T0Y$Z9&ZWTBY%edb6w>pf)iF&$
z^A`|S64bRAghsF}@SC+JlfYN*0t(&6SS~$OOzlRDnNa?0a_kYJl`Hu<+lo0aSAwpS
z2DD%t8yOW9mE)pGE*!Ls!Ge!hjimrmW2~{JUw|Eua3Vjc&y-qhB*#j52vGselUcPi
z8(N)v)Z$6-295GXZ$)IbMADEJz;7uU77m8GvP+z^6B+1Xw5fap)PP_upRvY+_PwH+
zM-^ID3Iu`(Q0a|M=e?LTNS(>Rv>i#oU}7?F4^5Xw7m5;+W6kafy0$E4!>vBl96U1=
zI{q<K#*6x5ck=S)p+aJ@WUPv7iU1*Uxk82ew!mfz&{0LU*hCp@c-eO~axdi>_y3#R
zH1_oV9*5S7I{!)S4$Omrc`BsMCltx9&PI*if&$;_TGSp4+$9i!W+PB{CXg>loI$!l
z6*X-y$dp>3y{E>_#B8hhAuSqP%V6F{<8jTYqGXFN^e-Kd7%Kax46H{n7j<(IpG=?#
z(e3C8+lPv^(_m5>lcdM1id?*V>e5)ljS=}QF^Sm#vh$sSf<zk*V&aq`>Bqr~;TW*=
z0MvJkQpR&iDLmVzv%HyDV*G;!_Yki|QVNOMB>XF`Qwi-G<R6(%XbUT%?ns^-LIg=0
zzdg|UBq8GutHoV@B%N!_x!=gI5++U40ceOEwY_YejkAGbNAB-$Jfni^?__{9ck=pe
z-d(%)LN=3*1|@0*C4^>e4OcCuA*aA&qHh<+i8ZMiSMRUMr1^$4PNGK$<znrn*X*R8
zAg4yn2;)$B)pm_14r$R0bjEc-VNH^2@Z^y2jDJx&Z4w%AdqvVrF_dsT$#pV`2&pHa
ze286h^qI%sQJ*nsTeo*{$P-W#l2ajk)_M?sD%8s#Jklt=NOFQx>f*;|1borm`wb0B
zQ>Fn7jOKqU#l}pZ1+1il=6BZa4f7K`P5L*p;R))?K--HCO~xpw=a<6xVAj+y87?qE
zGcD#b+hiv`U@RO_7&I5@lBUZ?W;{R<nlnKY`)krxbAE95VeUq|wCJpD8-@ay+d3cW
zOTi#_ey)b%gJ(-+**PbgxWFG3WNz+-w;(wN&xY-C#yQ?W9_n!h5J~~~;4iN}1M^^r
zE=t!dDQ-MxEK5?)u4mL^qS-K^+HZk6&=Qb&D;fhD*S0nou~RcwAqGKE?^?A<4XDFF
zLUMQV8gqs^W0r<Vsh@=3t$p+AqYA&rPSZrNDNgo+TCuX-OROLb;OBCrVeOz@G^Jd@
zbTFIHRA+1vO#h6*3YoDp_2}|iJ6cs@_2JS@)9%okP^DlV3Rj8Ax&(#Xz3ud*$`lf}
zl>=r(AwiN+)A+=pm<tG5f58R7#hyW5iq$~{WoMa7L_$l#(=QotE3fwqGd{Or7IU_l
zM8;&l16opu+|-}?j%KH`W|LmGHb5tp!UIh~9>RmGSz<SJQ!V7sy3_02>zM6ejs%mM
zeBzB1HYs2XY7?93gU>cI?mzaOT$!6!GZTDKe)evnep6p!S7~ux=;l+V|5tlg8rAf@
zrej@bwc6WSwN;SYHfj~Yx}Ze_>{Z)Ft5S+mfq+U8S)w8cB5N0GtrazjC|(4sfM`KQ
z2s^R1HWyG)6cA#dE&&k&2pE>I%<}}iGxyBQIrCx8nJ;7eNdx(>zu)^l+q-JX2g(mj
zJpAw&f(72XBd+#lhMpW-03;SRs^n`L=+T<yi?=$U*kZsrHyOY;*j(p*jz2v{>8RJh
z@FxPtE3as=-06KmV(X>H_nHlz@FG|`3_H<KBM%2Om9MyWP4mRNg~}bC2r1<9k{I*z
zJ1e}>p4B)(4@Q3vVayVM;kP#iA0$1!b}f*}u9Zwl#NLURTp<%5z)}`edNgrgp%aMG
zD7wEySq#5Rz@G47p^_HVI(V<QFd}}cXfjCsjZf}lx5Wi^EksId;f;^Gj-JD|fh~0s
zBQ7bvMBc@iLV`r<;^@pk&)A!h>8@|!5(cDOE%~H#MD_+@3n1-D-aFq*Pk24cX_pzK
zgLYzZEQoQRFIJA$)gMB0u{apZhGILCW5?V4b^u8?+(MxWaqXEY24qOyg)p&GPlz)v
zEX~w!&3?D=k3Ta2h=}Z7^_nyZu_6?9aW$J&{lF1FB=0^iFQUy256xo)I*936p~;VV
zv!twFMGP)bG0C$<ETMpJJ2t;kF91*~Wo##btl;RY<6qucZ;z_;gwbA0%u)zAo0yEZ
zX<FsIeI4vpF6$`n5=l|cHVk!jL-@`*@M7IsO>SB5q=Jl|_zL(kBFZmfX{W*F+O`g2
z{^J~+4}mQ5Kn#Bu6E~=Xd89Z{f^GNC(+XqopV;Di<w3yZE<+nZ$Jt89aZIn!Q*q$m
zk^rnRC4%;pjV+4?$A6@_OrWh|%7p=zh`u6<O*S5^iU;8(N1ll?DeGf~F$*Wj5A`&9
z*qi|EW?tz6=@~>`ABZ@{Aw1rj^zw0da)_W&KeVGCotCh&Q!)k1NqVAbhVlKoD2E+<
zinl>KqXxeHDK_fEK0d$t!>^Vp+P{AXdMCoT`;^m=Zi*P!cR}wKLiM|^0Ss>}81}3i
z)AsB+>L2<2UmIItY0eTBj$Uw|DQ8o=;!awUKeb%i){K<Fy0LMcOJT=B*#q|aIEh*)
zA{paZZh`Y2KW&~e=KJ^nr%z@sXw;*3@#gdT7U_6*_DM!g;)k(e&9JuSuJmBbC#8Ta
z8CfZkVExa=?$k!HkdcX-gU7I0u~(srIR^dJUr~5L_v{zrLwn;Y3-ys6D7m&KJ?&FO
z7qllZS_25>O>D1Om0CcM4CwB?Qn&TVUi}OZRF6)o{^X>S%cXQ18+XQVyI&7kEOSuj
z04C&bVB87fQ%WEHUMp%vti=G?<SQtZ-G@S$915ch^y47#E`u0+d)~6#u`Q+3DXn3s
zKcq=<=rTPpcj0XaF{to>(Q!6YoPwi|edQKQhh{7KZG9>(79Q4pB>L$wj6@)bQH-Y_
z+DCG@Trp5FBtUFo+T?^M0>#WGl=RBG7uZb762boRfR{}WA7N!dkN;}uRtOb%K0AhH
z;?jVW>Zdt@|3C>s`de!@^v97Xjh}18!*1<jA~eHjSo4#U8{csPio7jZ55DKS0d;oc
z7;zgaWYwuhmWOe|ARb7_Luw-qjvl%aC(VvjBn7mNgEiM+%vu3HLwyZONlCL^{2KE-
z#ivp%Enh(qg@5-+#R37G13~M;LS$sc(ylF%(l$p0pQGWecE&Jqe86ELB_HSa<cV|q
zMWBoi3kRugr{_JlCZg}@{Q5xIw-Xi~C;x{UQD4d*Es#^f^&eOgRpB^LZ?`J<W8B2O
zG275#u%O<jE5MR@&0RBw(C(?jEL>p|PQwNdG1rC4P)blppaXK<)HO`d*{@NBA6)Z1
z)TXaawOcjZe(B=R-}#Ghyu|Kri9h_l_>80H8}H2CyYy!_pBraZCax+=JZ&^KN#d2h
zaDLI|bF0qJj@f+m_p_UmZ0y`tIKJ5<dSjXByLg|)VSgC++;i^;>u)LUow*^f>ynLG
zUF**4jn}g~@=A0CH-olTTUcA@3Iy$Yky?_H)(fg6peivTVG31mtv3%BuyzbTGQxb>
zl~LxK@q9APiK<&GZB(&31Y9LbM7LIuF&9Ucfvni}{xN7x-F1uwNs>^rq19@Ig{XtO
zPtQDSQSjAhFWbQ;b5%#w@`U@C&C#o$*t92x4ogN0OC&1E+uuj7O?6&ZIT-+Uc+?2<
zYh(4Y^!6N5Q{v`b4$`mC+*E?$oh-prnS(-LPxTPfCl25gPsSQWi=Q+GZzO5IMm9We
z+xh_{pnY%TDGdtgj(auF@(~AJ;Og~_<DgD`HAHVYX{u+Qs>iXjhJ@>N2J}92FhW?$
zXs<zt#o8D4;x)3s-PyTV<MmGD1aY$c(~7G2wJ*M1g~UZyCNcG{9ebCFaQGAYGt@E2
z7G5F6TRByu(L{xEj*WY++h1O7G#>oSriwe+rlzI@ht$21t6N6yd~jue8F(_6Z!kW6
z*5uFK)q?<6iGRkJW}J?R(rC&0_5FiiKR0%Go_c%s8!cm?-a@UKvRq<2YViDzqf>K`
zIpD?!EAxW<(0HT>3vmnWW<;I)eB#|<gU6Bhs1=l6HM-2T)srCK438ehM*yd>IF4hb
zW-AO2>(PodB;QfRc(!pu%il@{58w-mmJbVYj$P1wABVcmUc2Aerohqtqm_doR1#J$
z+sZz+J0w9`yD`||?8hS8xr4`<4BcCDS}#x*hohUJLSBk`#w8^!e(H;_kn3{gC%~;t
z)w4eOrwpQ%bL+<c^Bm6MdQ(a~xf#J#DI?1kO9mmMIw~KJEKewEAYqLaE@k5~+ew38
zoDC>IKYM{vTE@LV?s0CWQec=hq-?-8v*b72KsGmC(L-K5n1bOisE`-ls*eV4n_%RN
z1Y6rdQ2d)g*U_6VCO^gxeDnYR3Lwt^e=P+6`MEHUuNp46=HO6Q^W(x9(Z$oFmZy#~
zh(#c%&;6`RtyRtMXZ*wPGt;?y{4SbA7mIqrwzo`=F0Qrpa!k(i9x+1TRmF%sM0*KE
zi~<Tl51YDunV|8!L@9HJqfjK}Bac2X1%G?(HgX{?{A!Wr2vUHG&o!wp?}`p*)F8qJ
z|L0G)eX_Gz^m*|x!L1H<gw)N6=Ae7w@U1e_mm@;UtE*2^E+djsCVut1SMWvgA38cZ
zwxY0_BBB~853X_?>aE-tZ;1O|I&=jQ|NT3pzg7YRB$}EMr(-=!kqWY6s2hRi{vgpy
z_T<Qy%h9+&P8)<-8~S-hKsp#7)z|V`{|7W15XD>V@sjO{BK3AeHA=CL>7%XW{W`gi
zD#x5H3sABA)uHWQot&>2x@@f`SnT)yvd?jo)Ls*Y2n>(0`rtHy`|3WF)sGou28ENV
zz7M?<`?Al1gLRX2tm054&fmX(fBxE8S58r&%wZxjh0eJ1ib3mw@U{iPRt3}A(k&*x
zpW{{C5VK@D9j9;JqYeu(oHLN)RRn5yF2v#K85wHGW%Vh26bv)Pf<yDp*w4@eG?I~K
zb{{MOT~Kz%qU6JjlGzlOLl*Izw+S0-G4o*@>Kg?=HUriyWH-GDLVn`PXnroxS)@ve
z$kZRu+2Lh573O>F=9R;6{8BC=ASl0|$K48-82bv_Vg|wx1%#G84cY=Mpqmc38AC0B
zqvbQmWF{t|*d-s~7!M$7D}V%0f>Yrdo<GuKgzY5jmakva3u}nOa^e2g;I-;Hx8OOe
zPdxhC(KbbY=^2kYu?*=-*yg5*m`AT-p+Yg1NQ!f9@@)j5K@dVHN&_Hco%&`(P5I%V
zD#O<dM|MATsW4prIu`Euk27SlwI?^d1ugs*9q+x7Wq?qm{-ZN<W$Z$8Q7h^s?S?3U
zf1>^peE4R2S$?X;k#{HBYGBB%USk$be>tO3JiWb(ldPX7TlEQd0+<j<!9@9YjlQB+
zygY_#qF(vUkb6s3*LwObvMuiJ{^t>l;H{uT3J;(9L>OElvCO2q0ij5rRTXt~qQ06N
zo*s<R5PEugiV#d3y&y}_UU2gIo!Xu8<?VI)LrPmZRB-nPBxe+j-YO-OMZ=e)lT*j$
z+=hIFU7P6|cAl=f<SOWP890c2-3|JK2>ND%!0QSsw^rnvObe>3&BRMY*p0d$%$T&H
zh%+$vb)ygdjCG0-H|*Pyn39r`CdM-L-H`4t27Z}$cboo~b(s!}n8+a$oZ*CY951F$
z5h_IjY*?Dp3G`D_MdhhtrMrz%CRx#Xb7iWH<<KXtR>zZWJX>tq(4fD&49lIkW=7O@
z^qkW~90{q?f;3trN<&ktI=T4CW<W0@DHIMJ|HaK=8o}rWuay@muty4&NQ&u(tCUQJ
z6m7?BJnc$ffq(MA4YhB-roVdpnf-X3qcV!Bb=4hBvJ3Vz+|l!kq|lKx`ZmfW%2tF~
zO%++gX87DULC_=9U;aOBZ#i`2)Y|^|DT=1y84xjYNC?eJcS=j`g5(()=BTiXh^eCq
znF92gLxv_R!XJ)}`!w0Uh@<wvzFa6((DuifbagjlyLus8;{?E(o3BLlC;iKE5M!6^
zZ03;rvS59`<le=1;IjxXm7L(31IV}p1>QncKP_C0bWG9_gr;L0-)ogR?oZnc2#?v8
zwC3dJTjlnR{pHbwHinA9Kh-^1(eW><8*>F-@<yN#`p!e_VAMrXl`@GF@&AszIx9-!
zcsVzr%T#GEp{nPI#C3T)edkN6jxflO*{^_U9yonHWY=Z{eIK_BY;sYCVqa=%e<l5l
zhklt$^Qz%8$0&6CS^Uwu!NJ-8+ZDo}2h)Lk&zm<7BQoB>Zg4)pqPK}r!|EQ(K3I=z
z&IWjjwAR+TIbr#?4a%U_`vo6<JU6cEvkM*ho_7*g^)L7Uf9cXcQ}ATP&~wxbHSWXv
zhWVjq6}=P-a)Yrmi@pOKZB*2_FM@-Elj>@!t6g?`Y11>GRYpH+*l&F{75jO*OtuO4
zIkddLAkRQxFqh}P`}U+?SGW#ehA<R;Zub}wD(;op&Zz-!8|>C7Gm|i3s>;1F(dG|D
z^OaK?43n&q^bl8~MQsgop~w7W4ukRg-vLHBQmi%$uPupl^~?opYo;XGxn|M6Z@);k
zKNs+dEr9>NwzncSw)U?V2=0Ct`QWGc5rT*%yDeP>f-xeh(=hP~YdQnAt+X~^_o8}n
z<|3)y#^l<kcBL4fDZF#gj#1JX95TC-;4*f>nt!1Eor4)~`gx)qf#K1lVB#8-2uy&e
zviULfc#-JSId_%RoE~%$=cIc~3m0_jb~Q>ucKucm@-gyZ!a9F5-UNo1fl3JYVFD3p
z98Ne&qxBq`K~>nQt^3t@JBfzTJ%zP*1>#?bJ5h9A2oT2Z?dVjphI-{b7%7hALVs3_
z&4~lD=E@OqoMR=s{LJ<8rj;DV_jlMhsVikR5qz6dF=)0jsSUMOS6qKr9KVG;8Awpj
zbbk+N4z5{&ta$6^w<kRsD%kCSOF3~HcI6|7@t-SRu6&s*ki&#t=*;DcK4f|4r1fJ#
zFs;E#5mh1Yor`PRdNbCt7`<E%qTr%W#KbI=EI<vl9Vyo?-cBGSf?$#qhq*{wL3<Sg
z$5Hi<dfu-?;;C!Bw<6LgSW;6{;kNbB<;>pv+n*qAj>!_L$R*NN=+4X4I~r0hg~0u#
zty(Q@g;6L&aS3&SnmFoHZ5EoVqZ4L+nQ!X%sX#C)NVN1L`&;U|er=}>f~Le;b2!~w
zDHzdoA>jLT2<C<PS1O=xECvBSs|51Izn|&B>c0Je^)9!Ng)u-EEkr?pJ#=>`3&xB;
z35iKWKFBLej!X2<?#J3z(?&nFejp`YEO#sou`*?ruN%xNDvbvd!iZ7A=v^hrh3JAQ
zw)Y-+R`%VBt2ph5eUdIzAgkI4-IJ6{q=4H{dWBr*^;dI*AkfBJk`s8%=mmxz`BxX6
z-hztFBrrQ16_5q|&zJq`Zkj3i3Ho&Jsr#ty1esmm8kn@;4+Fsf3ZQ-H?W%L6Uzs*6
zI*O{4N|SIixp#hzNov-+WWS->_JhT06DSiD0W0v_QiGh=AZ}hRePO&#ii0RJ#yRBK
zHIbCN64k<pnZ`=t*mWIG#6k*Xq@W0Q#=VqXUG7?OItH7hev-7P74+9EQtK?Q(=#(2
zZU(~3p!X04%R7RD4Qrg-pw)$`Ij%oQ{hW~w9Er&76)R9c)BWQ>tn}H_xUQLsXSAG=
zrC;NM8_^$1=BwySI*jPsaH$N(U25Dh{Se6)i-$kF#|eJ3@CJzR%TP|VLA0s1Y7Sa0
zeGC+za_#c^XYf>?3#r@yz=sxDpXxjdKmw>$3OUV@F>Z)<>pz67{lKbERx5CEmuDEi
zgZHHRR`iN_aC>Lvp;la_<5aXjonH%GDo@i@t?FVE$rg+YTBUUMzr9Q?RO3@o3$|9@
zch~Uog?ivT9z9EWn`N+7#m)Aw_I}@&HMz{?=M8`eqe2P&S!(2#`Q}G)uT$<mJJ8>G
zKqpqDP8lOT!jg2A5?3>tg8V?s$wmsq_npcE>9b1b2v{;f#dT?wl4_FDO#u61WtG`}
z9zg*1qXOjlOP8!xB+Wg#SrC7A5z}pgeWA*52SXX#)X>@f2>Z&fJ--_Fe3YbN0Yr_o
zNbea4Zrm!7+(62(W&jn(!;n_I3)M2YyJ;e3ZiV%~k<9a$51twj_cY=jX|4)w`Ik-i
zF1?4Uu;Z|h<ftDwl_m!xECx3Q3?C4s=Q@u$Z9hW|+wQqVp`lRI1S*@E41h8^@lGhx
zMEx12Ck%r9VPT4=UVO@c1I~s^>Ls4e1uplOQ`4}LW#~gKg27~Z_-9{myVZ5PBqXso
zAi-`rS+>J@ZFLQD4JFp@qswO@oS#TDid6e9W^-Z*#s<Ri@@T{CBUO<0hQTb|u6TY{
zhGG;Ck(J&c2Or0DlE!T)v^QbxR!vqqxmDi3?_61cd6Z}>v<-#}C->8gsgfDt8>R?c
z5S5(iZ5g+te4R>b4q{c<?7XMRs~ZA=M_Y7kT`<G!ZhgVYpI#FTU~-MupfImi7l>C|
zO)W%rM~()Y>1@%$(IpGN-+)@S5?hmOYmG}6%j*>@R^Tx=X~=3SEiFwesVkk-{b-lr
ztF;L5=3w^;V*k(eXgBT@i7Rs2q^#KVU?s>@_~Lmp(w+R!_hDECxX11yZEA*Z?q!o>
zE2ZeecT{evI9wsF99JaE$k^d@2W2;o#(u_*+_0vvBmez?3#=`&<*UC8xV!GHfQwHL
zH27Zp5GTfo)CmA>k<4+OBST3ZViDFyT%%r`fn}v*i=w^!%cI5WqJVJc;>ZUffxGNh
zDEyx8FIjl-*^mf51{(406*Q;fN>EXK&8QVZ`zoW8=GK#dVgE+9o72U~1#%jpRLGMi
zOzX*;;~T4F3W40asdLzK(6nh4TBC~Vu$<ir3$HGq55`^RAn#t2DZz#o)_o_g=?l4T
zuC>6{o8A4kzq|znpSuoh>iLaR8u!5*%>p#4f|Rm*Zss2x_He-mQmY%f-ga-rqg>t9
zG+|{yXfTX9r-X$l&I>}rVgkcyK8n(aU6)owv0ulsZqRO_+W+Db8V-XuhnfswAx-BP
z5VR)y=Os!SInc4@DQz>{I<xS$7X@tH?~bRHP@z!ov4IFGR32%JIU4Fw_sPtu;k9W_
z(}3Fhe4XRNQ7Au#l;zHXl|+ioYl^4|EYKoIk78xsTep^Qu7jDbzOYlUnCWVQD<3^X
zVSd2}rTKg8^7w{x8*L3kAIF9hm31vMEoswTd-`rn2Zujk_vAZCq84I^Q*TV=t4~KT
z4PKr$Af;82&Di>2qcc&0@SorZ^;aF@tv|!|BMm${8(NG7t#aTO*~qhRqcAX_hkbvM
z7KT0ET$FYR+ZzEX^Cvyvyc0=L2h_qybYH$NBQ6zo(;_YB7(kw6$&F7Mm|DoF2XHO<
z%Jq3kTp{#fEe3muo>GLMeoj3KKqe-zW1YbN1Y_{-)6MaiFbGPM=Oe9$abM&X&t3vg
z05DqSIhLLgaX80eA;}21*H$+MIHS4)m~+vo1rSl#0>rXx9{?LjcafZ1V%%-pgdA$C
zNE`Gcq~D4f<emHE{MH$$j4bC<JZ7S62Gqc1*dQcddyU9~P4{V=<x*7QwNU3d6Z<bR
z*N+(oT0l7r;9`yJ!cZe)!C8Ga&xb!y`cX{54{}U?sY);`r8OKRyD0#QCYg$kHWU%E
zI{{b$D869-jG>7w5s`ApL7@?h_E%z(w>6nN&KNho<<WQmKnVF+W8FPPj*lPCzODul
z4mo*!`LprLnCM@bgPehohKIF>-WIdGmaCs(DmA#bxABm&#XAt%=teEsWlX1cM(bW?
zmN}JSJgl-VqcLNNPRer*9ww%)pnC5jMa(&^<R++s!b6*KX%U#zH2wffFo-f)`EXUh
z0*3)2n<#K(B@IPteY-Onj5xe{ysy_$4w{%#KE-06-gk<)^0#ZW%kXZ&RecON$jmsP
z34Xu7{>d3U9>}F#%lpP84c6MxijH!;Gp6rHb#$Ena)=<}suP@K>U~#01Rp{;=?sTq
zW#)7h!5D*L={gFJwVLgM^>KtR1Y>*wsYUJBfHKxrvpwF`rz$<UHQ04V6#E3P=wSkb
wfm|Vh?LX1^H2ruiG@SZh@P7_T-|aW}xxgWOoL%i1_Q9fMtCptzZT&a@2A>z92LJ#7

diff --git a/articles/rxode2-speed_files/figure-html/unnamed-chunk-12-1.png b/articles/rxode2-speed_files/figure-html/unnamed-chunk-12-1.png
index 4ff952965fb6cdfa43965ede33b7ddcae64a65ac..20db0a4816d9dc69e854603cdaccbe72c2e10378 100644
GIT binary patch
literal 64330
zcmd?RcTm)6)HT><#XO2ADi}bLf+C<I$pkhC0+J-E2oeO8EO|^gBBBk5f&?XJ0g)U`
zAW1SvMiGz<lJj?-#+mwd_rI;(t+#5enHomAf8h!Do_o%@&p+~VQfnEQ87LIWTI$);
ziWJH+TMA{#<(13vH(HA;weg?jS7fA4<G=Vbc`Cyh|5<h8>_rO-Wv46ocacZ?kRFAy
zi$XnpLdo`JZ__PXrS_)0A^UwV*gT(X*tc8a<cG(J(Xp?0J@|1yd2{UMkoc;_+_$fs
zU@+ndQsP!vs=%|$aanCj{g%_qPV>kMoe{eD*HQ+yr=oXz?e&|*Q@?hxv0e4M%3m;U
zp_N8EY|&BdN!9;)Z2VS0i|)-dP2NTHC!<gt9#=Q4{qr-WYRw|PKR-`!E+%h5{tPNC
z&0P42(i*(d_RlwSr8W*OUifPlk3Fr+{(OS<Oz!GGKZ-qcVfyo<!{+~+7ydnNy~HJt
zS5UCJL6%as^5~9r7h_fM`2OO~%?%}sy0wNoO4==jy}iBL`@$%rg;x?d_v|@IQ_|Fo
z85yzUOIax)du+$L_Rdb8BS&f*WEaUVnVXwqSijzfjdAm4+nH1We*VNfkI)tW!(RLU
zwekG_&xPljO(_)HQ_|886%`e|Uc89>^5v{mU%m3>%a@x?x6s$6MDm8g#RmK@<-+Gi
z<DI)a&x@5m-7CeLmZOv7WY_xlimF$qR`9)DNB?Zl6jr{#Dh|1qWjyhjua~S?r<D4&
zCRQWaq(Mk@Fn3A%{M@u&f?9XNy@e%TEtR`EsJtP|K4oqsT%3!8BR})NzI~EMnk+p(
zeEjI`<rT7ayAUS_hZL`m%s}Sw<HtLXhZ$Ce73e43N@!oWgbyx}8*JU*7-`j0!*%G8
zLRVLp`OzKSRgs?e*KB)j9G51kVm$k`FZE`d<mdeS=#Y@pw4XPQ3Rrb3HQDvVhSt;Y
z)@R(dbOn2e(b!Dt-%?egl`F%~$Bzpwq2ED?5xdgksQp;X7GBMWoyV=c(q~7yDuo@V
z-oCqD_Zp96gqwALOVG`Fi)WMdJ#G_+!D5TF8EE!BViutO*5GDAItJ8N$iz3@swb>2
z=~_{jrBPd+JGNLYZmZ7Mw-6X2c+*1{MI?uWu*n99q$#BuC#$#LI4V8rBx*Mj!Qwn~
zGgd9(B(G1jT*F)4{B>CqKiu?F?T1T_7?;03(|$jr-)U|*&^YhWwvOK%fex?Tw;!o7
z)K~~aMu;*Ju{*?_G%tk9RZk9gvTOCOI*u`j)ydKBsfn%3bKfqgo3|#}uu_tyq^S7%
zvl~-Sa;^5^qJ6gY@9(S-v>&@}F^&1GdU?!Z^M(!bq4_P2<4wK?FW*j^>5h)?#)w!|
zsL%(@rG8iIGI`yL7cbiNr*(RHd$Us?y17-_4`hX7;c&CDou*w)X%MvhB|qC|=3H6!
z>QqU7oOYHvqOXF2!c(aR%O9Usl{`OmYT2@7v=HO?EET@&i7Uw*GWLV-S6F>qd*m$5
z+`@v9mDOx)IY}iHiUB8<UDboiuByf~3#;xbE~;X~JCkI~F2yraQa%|@H*UNMa2&gc
zLA|)e?F{X$UeT740MQ2B-DY-n=>fMow(05XOViS=`=WewoYD$KZJH22MaLU0u#jjW
zckeEZ)kyiew{I!^W`mm%^92*F`_z*<;+jlrGj9zSAAGw97tX;3Ap?~pG%(iJFxqHQ
zS}pvuq@<*pb$&w1|Cog&ttw1*`vJ|=vR<m@rgg5TFc*S8^Ha4s+@hjcD*s-e4cD)J
z=Ia|3D&yl@R4Ow3NEj2<p5Jn-)8B6BGZW%8N7s+L28XYG6)56XQpK3*m-vS6m3n+O
zU2Vk#X6?e&t5@CMb@sn&FujoDG|Q&7pPJw>`CHv=6a5WxFJQQo8eX4q^YZiKp)w!5
z<kFux{ME>;?U{yYxeQ+xo3pbswIWU{W2E?SDYjp@dIJ|HXIb%!!^t)S+Ehh(`6pZU
zs|DdDRWoGU6V&$miaDlV{5bB$mtB^in=g`RLcfnoJp!rxffeg^ly#K&v1`RdPgh+y
z|3c6z+PF3M?@*(kEJtrhoEMu6xv2k&pDdTLx@*bCwY*eXPEL-tT^IXke^Z8JU*y?y
z=N|Rv%sWw;jMwvBOgzo&Qxsgfu&(7dA?rpJx0`!&mmIY2RZBLij#6pJ9Ln3CeChpS
zEFdcFF{^kS)pMWW^jM#dqo$%F&1pQN<%Ot?iICrqwl=b!s-i?0w;k4hgpk8d&9E5|
zI;BT{eu^1+em#v8Q{^jHUbAs=b4z28N4{9cXY>mnxS04U=7rzWr?2o;gZ{=eX*oIG
zWK(_puxAG}!tz<{&mjsvu}s+i5er1%MyrI!+i;N*U%|?c>&370@@c$#q`iO_a(ia7
znkba)wg$5tL1(9*Po*|*-TEuR1mQ!A#JsTcfpL!0V^msAyiWXRPi?eThV{dL{!zf4
z{@S!jSyAzd7_YFfI#T=vA6_kynr9EO@I>tfvc@fc|D^5|b5J8MFh4ijGjnTU(6~>o
z*QO|Uu*6t*R~581cszVqQ4uO5c=N~UB;I(-gdHDA#CD!{N?N<c?eiMVp1qCUywCo&
z#gi4cw-@;!c-0hHECfd{SHzV&KRP?ZTjs`6rh1b~1@w!ze*W~yD?R68l0LJrWyCW+
zojCEC@9P;>ua2%ATf2?_G>M^X>>M0DE%S3WSz_V~i{-eq0L9?Rj&-j4&OHxWwP_#c
zfdjIUk&*ZsmUGC=Y}R<WUb6aV*{StiUEzVx*s_pIn2+8FicyL4p}&{Ql+>O@LGLXP
z{OgmW)l*FSs&CXDv>m*Z#AJp9CuG|At9{GDC&LKniYoZ^*?3~`<NC<Y3t%ksHvZNq
zH;c*4+&odpthx8Uk3XRQ!YJ9O_gg~8?-i^L5_}7|PksPfDd@gwMxW5<&z~0-n@PjF
zTWM$LFBKHIyLFihGk1gAIHujJsQvd9H~KFmE}Q}aOuo_SZ^Mm*N3;C-Wn<sBgpWlB
z7SN*XDkPn%+kt_BPSXu7zD2AKgNw2jK&3<$xu*J5U8z$TCQ62eRgi2yZVDJjzq9}&
zwuHQlke}>&v5BCxnP&Tea<>JDS+yFWWzgl-DOYZ+Wt!KMCqV-_^UeR=bP;Z9D<L64
z`R`VH0a@2?6T1F<&)&W4)amho+5l1e%*H+RJC%I{xOH(|x^$^Q@$*9+zmE@Op89uO
z!Z9wnckZv<u3z&*_T0HD1e5>1_MCie#WH-2KAeeqML#Wtv%dYgDE+stYj+AOm6w;(
zLO!~#H!ga1K-+X9{hkIf<ia1X3;E@Fu$qy5bPfZref@v?ZP2<#M9S{Qv?R8!PycLE
z&#;d5(O8Jx%zH5a5lfKaJ;diy(`LRNAK*NGyut0ysJm|@qyE%sb}G`WC7TvfTzqb}
zWqvmBzprm6U*C&Re<AGL(`vh+VC(imZ)qxZ$UJ&sT6IriTCE}w*we?G?b)e#HoG?W
zBR|d4G8?%U22BlcN@AxUGABLW*pKy=W0*1<b}jtcfd1>%t5;WsP(6#=flRm#AFllG
z373&4ypu7Itx>CK;m*djckd~)?8&hIo^T0y!WDQy=b1KE#tj>GQPp#tvH(<mw5u%a
zejNZN$|xHzuiPdMIRPIg1oEmXCBAnHt7k%(TtI5P#X09<)NURgnW3Q}b76mNQ%3sZ
z4Jwnx<q@M2tzQ}t`{vC-YDIIFeX_%(IbRAX^eB`GZgPQZvFgbckG35yZqGrIuYYUk
z$@FIm90;7UeIIn<4v7>Nlc7G^iiocN{hCaHZNbD)TXhbSkmBXbuk^pZjKrhbrYX~J
zkn4jHa`91%RhgYEb*^Dl)!a=jEH17B0$H^X;3T}s0%KaXVukVd5BDBEe*BCrG%U<$
zYNQJ>dd$3tei;`&pcD$9+k*#XNQ5LYQ<+XR12-vRm5=}Vki;o6SZgILE#Ge8(~f!m
zd@EH^Mdew5T{o8_l9hRyrnc!$`sEBFNFez8!Q;m@h~u$HTrxDEp3$kEc;l}xkKqAF
z%zNmclsNTspl`W=J|HcYh?zUfbh#8O5G|EP0v}dR*X($UDj*Ao-?v==9&%)U!_K{U
zy=cjDuZ+h<?G;$m)tR>}$m_>h`Z4ipmH~XWo2OO9Ts)jL-Jj8JIy{~;pM~2CI?p*=
z-2Y5{o>#&O72W~uOf{gaa$iAXHmaYH37I=V)5fD~Swz%ScxpQ+L9Z;*G&@$Var^w?
zCV`~gYZjU9<l*8P`XVHfuv6!|c^a^xdEMo%`ML3&(KZk9?gEa$n2IqZQe{oe{gFR%
z{`q{R98g=pq+Y;rx-TVtZmP$443Q3qZ**q3v?nS$u7Z>Xw{I`jyL#22IrG-kNagua
zzm^$EEa<hD7g22g#w!A$67-L0|5X)vK^Wk>qN2hJ$+oxmdV70&WJE+nRU3l9;w4KM
zw{GRkXo`FCWP_k_P4SGW!@c|WxdjBOGMgrjv9aA<?SS{-WWswSs_j<?K44E_+OnlP
zd%E9~o<bPbu8kcCn-mDIj0E1|MkYmiim{xP^%LsK{d);BVxLN!=B+xb_|zW^XCx)N
zV^f+HTe2Njm9+{I%s11{XVqj?YdF?hSD$R`!`1~nJN1!uE(j6N=-Zpi1YOlNu)3IG
zWVv~HRj{CM-K<UhW(aH#Am>*S=Hf!Roo)BSJk7GBc-u}9TY2%B!8?z49DA*s|CsCe
z@r?QgL~BJrV|NxFtwz4ByKf-ZGj7=uHaK_#lZt=NoIV|)SM&_n$NR&F4@d7~qZ^!D
zX`7Wh>(WqKTIzOSNcOmOue_w>eFRONt^yfr>vsgeX_#dxr!-k>gf5E5v@KaL5`=L#
z9%#<$t_=6!@LQ*(as7G>S@?W8Gm5L3c&oZ9!`Zdg*tNB_jdu8%8I=W{Ag^aP)=T}h
z#P%Y9sJXehm%o2(Ow4}0Yu9)VAHD!s7n*(T+BH9&w6wIsqM{2YPMq-OP#zf{KVpA-
z`HB@hM~|i|otF=0GEh@{K3V<x%xXciW=-pitj9XbTD*OH!V&*@cJDs1Vm*rz!UR7*
zf7Fca-n+>SeGF^Xcp*P(^7^3Kdie06lDvFSW1}V*3#AMbY%binu&}Tq@<KRohA9v$
zf>1CvFtAAkGPF^{yVxtby45iklZ?cVxP*nKrhaSW!>C3@M4XiiPg1Je7F`{MsX?0J
z<m5bSo#EdmPZ@kI$vBu^8&&f4Yt2JO9#Vyo$Wjw?5hHd!GCh??C5F@HxN%KPIgMRr
z1ZbBel6F&Zad92z`RSI?zLXYjj0zxJHJ|gOjFv@-pw-mzu+UKD$y?FC?NVi=r7OCt
zqX`PuO1*Ic+g}C9&1$^K#;2hu6}2l+Y$X8ZTeB81ushl)E6H4u!Y}%i>pI3c7It=a
zlJRQdw4(7yJVHV$M_L?`2)>urSrY8jf8KejQo2^_&85GweOR0(F5?AM(zDry&F1<u
z2L2j+ed3MU{`A?g20??eU5Mh+MLv8-&i5xJCXN7hbyuDjS2i$+h)eJ09_^}>LSZVe
zskwvcslx+>fUwlm)W!fmj9$UZqZmqVB&QcIUKD=$a`03v*|<M{avnLNIypH>7J`#g
zHUk5Ll9E!$hYu$inV6zeQ`PX*Dxc%g)!fn5Mgc+0@d<%}I+1#ApRX*p1QeYyo$*Id
zE9pGI$tka`9dL17MBKL0Ku^yLefi6mFXt8!(cHLkqY;YXNS2_y@dpo9A`}R9YX{xB
zbxYW3CUqUFSPTICFYHhxg6yC3q5Bay<ot!PM&yiIva{;nnM4bj)N=wHegFPl*)H&z
zm)CQq#r%g4mz1?kr1dUa)OM|HR=)0@%gl@e!o<1!$2&8Ain)G#m701KsXh`5s{{9a
zAUQj3diGKC)Ujj7I<C2He9g$n_(F8TU83Yl!t?wen4-dx62TO`nUY8ItJ)6b2|X%h
z&~3SLy=C@r@kFse&CQnk3Nq=h=hG4cL;Hi*HcIW>x;tp&Qf8*G#-b+&_eAy|^_L3`
z5AT_o92Rz-%S33j{QhAXSuh|!I>c(fwLCj;=z=n~T1U&<Tf-e4J_&(0TUzd7G@{y=
z+fD`?VhHaYazT}aM1K4B?c#A2EvJrj4+ie(=67oQ9Xgh{%Lg;qx-WSF`o#d_8QO0l
zt`;8?5L?-`KV^AL!YmeM_h_vSUvddb6GMbUW>LGh7oFZ$MonibJBngcHakuC$=Idr
z<KW;qcI+an^GpJgM0J`)r{efLc%<rUfsQKO?<_jMc16c$SE%YZDt|Xtz!t__aC31L
zWcGb|x>wlz=Ywpp97XqkBFo96cC#Dp4$aNYWvKNG!5$(to$n#t`*;Q3jJS-})=O6X
zjTc?l?L?B^!Y)vD3DEQxB7G%l`C!l?MnG#*2wcVE7Rch5=@3Nu7d<IvE!utc$rn&&
zz77t?T!mtXVKu!ROj)-z>qyak!TzHD6QTBM7v<&U4d4767|`inS{G4NhkSeQ-aV}f
z<4QFlJ1e_3vwkE<dHG8L0?ZswF%{`IzTZZ~D)y9VJAIgsFTS~<OIEV%^2pShl`EfM
z^tbWL%!>+f?ccA2tg!7yXGcfWn>TONOI0xrsP2EE;P%lf)lYiC@#BE{+k-hN2IVKQ
zF!leT?0soyz|oB;S65djEh}3MPRR<{!(Zf<qWIiIFvvXa+1XIHfi&b=MN#`PB@Y(S
zsw{h(&v|(@Q#PZ+!-reK0k5m=Vk*0oFc!R;sX=ew3LqA6gGHY43&4d?zec?2ys574
ziv{-fT4{h{%8F&nxP*o48&%&LRZAbxc&CKQ3>6!IZ2Le<gxpS}l6LdRH*d;j9C)@D
zZV$BSfB#}=Qb&jBj>A7qrmAXcYLS1jwa;Gf?aWNpG3G?vzEym=SH|7Z7&qDWGXMb%
z^Bgg<_2Z`B!^7U#{GpEpK}Gv$YFZxM(P7{T8o{SNze;MWp;tMVeWQMrD&KL2{G*SH
zN=jm_IZYGPKuulhA78R$Ny+LxLVhKgXMZ>;)NkE*Mf1raUWRQ<zD+r*H32N5cE12S
zjc!bvQUq%`Ju0>wP!9pgWjyk$9F*Mn#Meu}jm-R4A4s}0hXUkcNf;vMsv~GN-|E~+
zYS`Voce6Nttrs!Vi_gbW4)f+w<wmUa+$XDu^>;O?&I`GVU<mgk&1DFhu{JG^DN6G7
zQOZ2{TUktrKzwdTdpp@Xge4@`ZMyMsEh>AiQHS`PxpW$hP8ZF7U+*jrHVgQU)N4%I
zw%-%=HjfRo@KMvp>W%rP35+lX|Byb_9UWa&nG&<zH1&kk;~feoPo9j_ZfvhhG1Iws
zid`ewNI6vaefI4oF8AM;mYz>(G<P#iYv1C3+(xdaHokUhv<FpWzaqyYENNlOUr*NU
z6t*l7+P_66@>K)gmSN>e&hg1c*S0kA?92wf{MkcRuU_|Mq$IuWw3*+@t<mzU0O>gX
z<;xwwXD>vDjQD%gg-NvbeCWH9fk6>TV;d()_y-S`iQGJ>G-_q8Q+g!<K<+FOLPx>l
zq4X-7Czq4eO9J@YXHD~J+wCS*yl>vtA32-U(5D|^x^!b6HR|REqorH(?2>1%3)(GC
z4|GUh(Isiy?^L1V?K%7U^=s~B?<+pnHfEi^JEZQ#wDMWsO|>zxahA&d6pdJw;Q^Ir
zLq`O8Uv9lDWpzHa0eAo_Awo4?8yVLpF7tOGgZTEW&qm3<^D=!6De*V50AoT!LO8j&
zsQDhOwL@+BR=wX2fw=IQ4k#`ytt}6hvLZ_+#jt8#aPDi~&C6EftJ-HBc{FlFlAFSn
z-j&`^)zsAN@6?dB7yhA$EQGulU2*O4d(#v@Pe$gz<A~7hbEQgS&gyk@IZ^q?eWr6_
zBG&rYo3>e$&N}m<{6Yp{S#Q!bemSqlaOq?0g-~8$=0(99yE{5Md|w$T^NMU?V(RIw
zO9Do|9@@h^y;WDd&FRG3M8}9}W20W4!gmd}TG96J-wHii-D>67tM(KmF2U69-I|OM
z>-(9u%+Abk_t~liGVA3wgr8@X^%aQx8uxqcj-&D(Vp9=mJ2q?x(lAXs1755KB(eAN
z=W4<732_1A=J8TU;m>%rqRPtTO9I7ns;K8swe+X9tvlu~mo?epSD$(_nEdbXoX*S-
zCZFqkM~|u@T~s5ijRN%d5W4XBA>HC>K13gc=R%|8?sgv?-s<D;Fe!xU6f~>|VbEE!
zHv$VO`o_+#D-r%!V_X-)G_A89zBR6uqm6)9QDPo+8Y!1jPtuo78z;34vP!JNcIG|z
zzY2;VN>#->jq=QPtL7*IG*}Nb$71&x09Fax4hA8&7dgHxYFCt$luR~~`7peHi`%v}
z%a<*CA$r99M#a&Mhwgvdh956yRqMSh7EqGBq35;s#(K$*A3v7J7aV<YMDFJuB+vl9
zJY!X@yj5?f%M99zM2oFiZ}$!i@cET=jP540rr)ZnoOeSjcszW_3YCIl%Y_RUqBrl_
znk5ZjWi%XJwH;{lok>F!9-XjPL?<w9N-y|+!#2Z^WfL=V6e`MS?9WhBJ}$2Ee>QPg
zj{H(MJRvA)0wADc?U$UEd)Ey+UV52<p8de~KwZhY=MQh69WQ0q-Px8@`fK!>gWR@2
zWxbs3;@frlLyaDF3h~mm>GYdCQR+TiG0|(uxB7ebfH~`}R&CeDqb!GbrIk6}3Y`3o
z=c%4BtN2D>Fh9Q{iTO%OUSNYt_<}pH+uEig_KRAd-~qrT*yF*XV1Z9a@|OLL$1KDx
z=4K|PrD~<u+OX;-s1-9$w7Y*e)a;oz{V6Noq;6c&<f9&5J2zKi*p{D!1tEUBCDW6W
zU7$A(vu@hD^^1XW(3Y9m+3i*K5emyQgC*UgDm`Pre{}V+w@2JezS)*1&5<23(3#cP
zlK^HKlIQGipPZN*uFdf};$#U<jWw9bEX(w%Og2`fG8M6<uE@k7db1&fhcq5J3H$&^
za&6}^^N_~Aef#!F(?}3;=)bdKeRofMP7NyUXsg7YIOS*+8t?K<m-vI-C-wF9eLub{
zTKE3U$IB@V1TA3wAC3@M_o?YT!y{qSyrr^v1=mq)q3pj9x@oV@#_ijqA*&^=ahX0?
zlE}{SNWSsH-qnxnGGz5q<ExFlKYj<WceO9o)m=H#Zf7}#3WO7ZrhKNm?A0BpN8kJO
zt5>6fu&`9FOOIT&YL#}6Y|S?)6ewk$pZZ#qA+IddYPDW9R7Oh9tDWUb`iggq%QE}F
z$Ju|;c(>)&nMG!|_U<XHnH4;_e|*-*o=<7~w`iWh9RkU;?G=Ueu{;U*#hdf4ve}E7
ztvDXQ61idX<`B?}e(!JA&8AL^<*SY#0(*S{vme>~wS8G;o_)=%?blhy%U)glg%?rn
zsrFpSoW5t)rdBZ_awxme<3Q6AZHl{0@EY4_jiQ)z?(U|`#nPIpEAj*#n~SnoT-UQc
zV>_Ue-S{#yX1{K@7ma3K;K;?rHTAoQFZo(2i@%s-{f;nwk&b*;=l#iZ2)^%&*MPcN
z@B*+?v+T#?-pJN!+nd~56hAcaFpjpFk+JI@leT<@b)O}u2i1uk-m6xx_6*_Pd1DQ;
zz%LsA=KB!qp82l)SgXJn_Z{~t$EHelnzaLLmn|qLXg2*TiW4wR>43TB%ig5)_<zbz
zv2$`pD0v8s2N+0le%H>4yT0ptvwf{&TBhRWe$SQLHmz8;eEBnG?&~i<rTbhxXt&4p
z-aX}LAqT6?TekQLiWk)g2}uiV&5BgKaDk_{z2zI;N<I91U=0|*;s)gXIFf+?tw;v4
z8~JsaW&K1{W~K;o3|N<fmzE`*l@f-3PZ9B+MpNt`5|CF#DCF5++s@0z7NWGFNBj5d
z-p`$h!ETKo?SoI>QP>xLsJ6<Cb1$<|<zv8NNvk-9sujA4r@MZnrG^jZw6?Y~i@(EO
zvJzc%`b?I6x7`DM*+9f{*F&32x2jkj({q(^ocQiK1zJ3MbgOQsMVe`^Pmh(cW=Z@n
zo2k}2>j?=nF)^V&EBdHUwLR$SVC#Ek5!)O0##Qxbi*(AL+c8X~gSapR)J8yd2Xy4t
zEHzCbg=j)TLdX7r{!JaE5P8AkX7xG3?ON4badEMLvCFd~LyvFn3h?$;?qQLZ3^FKI
z>uVma$XXG$QHXi8N2G2zc@HWqpOFgg?IQYDuR<Lt$vgmQs=icr#BL}#h=13vpvD|G
z`^d+Mo-t|TKdr@I2^syZf9+Z=H9iI8kj0PfZT5BA7bQzuo;-OX%+py33eDnqe|xGX
zFF`CH@013L>lo3J!8v(pv>it63f-g{TcD%?a41MS%kG6|48+CA_vZvUFdt8!Dj0nm
z1|wnxnS}svEU0Rhccw{oRRDR(h~ACVnn$ye7x?O~I%KCg2W$^r{-`;kKk!zbs_wcB
z5h!ZEXuj=wJw6$=w}rY-*k@?UIdUC;6`de1$oq4fj^xF=QFqtp9V$k!@BV&-^8g#$
zC+LROthzHLP4hJYb2D7Lyu8VLg5BGF{CAbBS&PRFjojmJ`*E`l_hlLBZ<lYhFJNM1
zRL<CU3pgG7d!_B)M*RD-p3T{dv{IL5>Qz~HYO@}GR1=Q<m=|&#e1R&`VsvH+;oV4L
zm5^`=lk<<lyS&)+S#AFoaB{x_?E51JCwXnWhs-&;7A;!Dy?=kq8;&k)C3fnjedkht
zinR^+`um%_GePukmy<_XYnr0^ZA<QU>%RKXlolrqpjn>rIs9Mb`9QJm@?ck@ysE^i
zA3CPJGf6$RgDPdUje{|7P^4`*7PO4WlOy*Fk)C}${WovkTqSc>V>?1tN7+0IIHj&b
zXU?4IphgWfiEce2m;Gp!LZho7Gsn^7CfDs*7EP{k8S7=O1GDOXb!lt@u<a;FNm;sL
zMbY;gUQ_({dzr<L?|8gi@YZSNt%4ix+_}TE<MY*#>!Q}Z&mXghMuAZ25}%(eW#{OT
z^r%G+pr+gI0HMTT)m3o@xLHM0^CjVskgS<=<Z$&+-lfWxH@{4nJ9|`(%*TVf$7Vg;
z(DKeigV0W(=+;hIguc2l?-0FieAPzx)dG29xmy!W9NukaV)F4-Dm!{#i1$*e$GL;1
zHhey<Vtd|y(+8=R<^k!E-^@ZMKkHfC^inS$Kg(jrt9Nc*F3kEGf9P~$>I1!$hLY79
z@87=<<vj=CDziXGBH~er{^2EMg=BembZ{?-X`zxgw+0IX^H_l|K+Lx975N5&aLqM$
znvA5aaj)*4J$p89+^BGEuX<Er#H)**!oFm9ht2{x-P(U;hev8{flxDRUPR-MA6G0k
z(-i7oh57S7VyW&AWbZdQ)I2Wf2Sy>?X5g#+T7#=Vlp4w)JKCKd2!-xW!iYU%gK5B~
zn!$FQ)Gy2+TH|u&reyl#BW&C^hV{4P<TU0}D-fl*9h#vo_kgJZhnIf+d`qF7uZnff
zM*hRmoXig~F@|Ju)s%EMWuz=$vo(m+JWroK)i&5^6eY~TxX$(C4(6^B`8T|E*M*XJ
zFUfkVnx9Cw{oMwJV&(3}oLaVT=7DD#v)+YlJe#q?sy}P$(X?oFaOui(jrNwRN_+N*
zh-Atc@lO2(euyV7^`xri&3m^rFcm~hCg7m;{T=DEXH%~_-2r*UzHgte#o|<GHWyBx
zzFWuBz{#|yKqO6QU**jKE)~^eNl49*BV)lnTCllp><|iBb;6hCd+q_J)*4rgIGfgt
zPSkaNk=#oFNe4xRgvJtkp6z`4(`o4a$`xP)izQmN8eUoIm3J>0LVtm2%8`Fqs)hVw
z>jw-$J%Nhy7@JB`+uyN_ks_<=x!U%?7PlWiet5ln8RO;UC1YoXQlVRCe#*y&xpNta
zmSnarf^ERBkd@0V`t|7UScB(33cY!Xva$fYX(6TJ^O^u#=^@j%q@@e>j8ALbYWF5J
z;nx7sXrNSW^ZDD$R&zUChpf5}ECB+lw2aJ4QG1mNXAZX|nMhw!@mXWOb3-1^UfrUb
zz*T4Tr^}qWzvd9fvZaG9)ZlKnLx-L=nZ}p1X<>orAM9-&&RZGoKAv?*fp(MMB&iO1
zLCMsEk@Of*Fyy>i=@BwMyi*W7c)PCD-T(CI^ehq-c_PA-KYH}YoYidPP3(H>3@tFC
zp?04^)JI{vffiPTyM=Mvwu1R(F5J;257;O!<AP}}sx;HAQI*+&`MG%QTkVXZ)4fSX
zb%~+itx$GvW@T*&>3($o?%iKUawgr0fK3bp6(Q8Ag7Z#Sp6%n@zrX6m;cFvcbEA;F
z(|;H8u+JWuZ~*G4CgI^%*$GGkAfTr@#pffPLC&^|>0oJ0jSs{FK^uUIw(EFtt@u?2
zP*4rjdYig2=2AZh9<VEgsplBiy}ItBJMv>+>%AjTg-?MC+ON#R4}Np6<fAezo2GV)
zk70{63rkBqHgzwL0mDahmJ$VGlQJ|kBwQ!W)03{x5T_E<lVhLKq8jbBGR_xwY8zK%
zNxhpE9;txT?2@r>+hDU*t(JPim3w4y!@lt>Jszl^NcR_wW4TfxQX$v9A_Rw+SoVie
z2!AGXjj<X>Mt_x`BHoSx7%F6Hc8WV?f=5y6w+^uVvC3k3ClH%YIv@0uJImKrga<mQ
z_v_k^{Mt<=1YBY4P!Ct$Ql6#^oA{(o3il67;pFRe{2;l$#N6u`7>Kv;Pm{t5d@UqE
zvHj<ek{IlY;*$mEdDB|5Sc1zpX{3KU{1v>|sTY0L$Fq!~ceEao<!@cPa%I_%&u*hb
z`K-nt@2@4?Xnl%VJgA+XxQzY^=rdVOpX^{jBhq6NL}Khpe_|c;^4bg>N0fNSsSyQO
zW{T~y@^pzOt&0??Kxcqod$8G}<I<&Z<qOxZU%$7=461^*eV<?~%uB|#ahUq>DujAc
zI^#2yU{rXn8)MoDt&AlLo~D+nFb9nS^~PyL!#HbqfuMIVI0#yFoO7IP_vtVJsa!SS
zJa3OB^xo~j5{gpTzR|6CJ(v2#)Iz_%7!7i>Cp)&6yf`ccr&%SqrC&pRX5ZeNh|m1}
z5V1kIkAFiuRaSOXBAWTWkG5Swj6xR1b+osxvgm-3Ce#^$TfeV@>nzGN0S#{V>y_))
z$8P%$6)@j8R^oTWr?M;N#vaw6gex?`prHsiY7M1|6<(lwuC6jPko6)?n(BT9VcWJZ
z543>h1>*ADK=yQ;h@<0&cJr<#n-*of?Xg8W<ymp3k#D*lFTTe?q9NiX3>_Q#vEcA3
z3>ybObV9yG#V;2w=U)LDMKrw+I!_PWTg*pHHNE%gE)m&{ijs)*NCO;HXgyWKrQ-3B
zR-1}2EacDHos0M!Dj?AziWx)M{HEJrnWi~aq%-&W+cz~BWlD7$mhpx|Ga(kLvHm8W
zt}Ch2;o`HAjRp4fD^LoJXr3|e$5BcVH&1ZGTPTF!_@q)+Dd`hs5-RF}t{*=|5DzoK
zJh%TRp`U<(W{?K!(rH%5k*6VUOk2Zb{ldf?r~Z#3H4_^f+po_|Is!iVkG2IvzqL(R
z@P*_uL7|oqEh;D|zi?hO^!4jK@HR=efFt}*Xg@=Q_R@~8{?VQs=cQ=~=mL6$n=kI)
z@}JPYhpyD?I8BDC#B1X@LyXRgj(P#~|7$g9jV79OfF^VSY-m9rs1?!3t#<!DL@oUx
zZgqwFcm)LTQb`SRf8EaRKdu+bf>GcS&>Gn$Y+So`N#IeQRToMfqz_-j&WdG}2wmWw
zhaMhBe55>BVt@TI_}54vu!X!pJ_(-KT~tMs+Je`=9sJLwLjScZNhpZZFN(-iqtef+
z-9^84Sp?k%VK+Nkt1K_Si<;j4Ort7TYW;s--$$-4_19m2y|o)PqB3!)Ov6jq^RHni
z?>-DW-7XN4Kipeg(hg1hJmT)izvdg=)5N8536E5tpzF%Ug}{pCH*W4mzqbVa*DS#J
zM7{SbLL8zsnU4SWW#Zjtd3X8aa^!t!8YJb=-KXx`{_FTjAm-uquuQ0D+9srl*o_zf
zg8Vlq`grWt6FQEg-bf1bXK5rf_?czb{cFK--$v{Y#K{P4R*zR4p4mk8$>&6ir%<e1
ze=eek{3Ln_!ZaHd`?%6&(USjsHFP`93TQ(Iw9>0ISp<zXXW1ZjP$&}qdmSm1^=fc}
z0M~L8J7(}WFZm56f{6^I3^=1^=lR)@=7l?r-d?(xVnp1m!VVK}*tifFi6{48mx(U}
zS$fNSij<?q`++06vR2bC*1eW%DW&QI5Q5k9=bTicN1J9(>i;)s?BrgjkfhAAMxUjL
zT=}$ea(deS&vN|Zm70PINz_fMB3wUK4JIl7Coh%xK-O`Gi5v7mU2uO2(6`Ex<2L<O
z^dbbtN+6A)e`oUonQXx~wnJBjuh|XILlDIb7OJvV&_&Q4?K8H8As8Fg#%a-b$C_v1
ztLpf79fdvYYF;F8mT0Qj0;mPU!C!KTifRF9fF3Uc06=0%ms#f;+}r?${;fgz9*bL_
z9&7|#9x-lB9~8=2yfri7H;oWegb{xu8-{=XCS*RGr3gHy+wwhXA8g$Fv$Ipx%w!tr
zG2+j7GOvLrE`mRLG>3H?aK`Aw7``ZAT(e(v@~3C$2Q5G&j5P|d;>-gOR!I_oz#k7G
z$0H?Fgp70M(ET(KqbkmSUp)Tx3O*O^9@JdfGrwi16;P=K49bEOUzwQ@_y${3IkF`x
zUYq_Hn02F|UMJRFhtGv@rDAgvx+YEO%Cxtzn-M#=U0-r-Rf4XR$J@|7I?l7nb=j!m
zx*i|zkErZi2+|R2L}4|pgJ&X&m}e?*&5_B@K>Mp1R4UaDe%_M7D*YtG#?+hYsW)Dk
zW)4YWVyU#xR}@lAn*?A;2InLL^??>b^sw))tTiyNiBtsqCS(usCzG&a@@`@I<!T__
zDZPNO1`o6@Zw_3bg<rqMRST_$@axKA_htWN8H#?mz<VGHW?Ua=1rFARuR#g^9yT(q
zWF`n)Q%?@xL-;YLnIS2vBI2eMT-LP(!wVCtv|<sTi&0l?d?4`iJ}UIL1CZ>~ohH7o
z5dHmSum0L5tdyQ=RUN?;ofmlRteM|M*m9n1uo)Zu`1Fs@xvwf3+qP|k{XC_(ojjOs
zDA&TJjqcoCERl>#TXg8trr+Z^Y9FRy9(F~otlZbfwr^iIyh6%wezKwVmxC)K5$jSb
zwhtX2cq_lqz2KRMj}ubxt9p~DF9;P;5HYC4uwpe@t!jV0Vc|Z>Ct$p<C*Ww~UdiWl
zlQF&;Ju?AX&1hy~NI$h9PU|?X$%EDp$-Y%Es>AqQoR4&B@cU;0kY{}PSR(?EOteZY
z%`!m<a9muo$7Er9*)m{&n74sd6}Il>?79+P2bv@r(q=`zNpt2!unS6Za*spYE)OB&
zGqSK~M~@LfkuP)bHmphesP#F{IRQ>nhM*g&Z2ce)`5wOZm>w^Z!WVLn#3>29fFvvo
z`o+&*fq<n}z_J$45ny`~K9lyi0R->4=>g+HFD_^<o@}IIhm$RNoy}SjHK9Q0CDko>
zpjPu`?m{FB?gIci51q%yGNC!%K3<PLP%igfU~qK@o1=bSeD>|k%nTb9#&>@fhrtsj
zE!j<)`R(R(`5WlBqco5&ypoHW-W?S^^23df?8{&il2u{h<#qeKhl*U`jZf+HJ)v8r
z;UHn7IypJ{3R^{O+p(}=1U7F`mCbm!D1z2q6*j62+fEsPk51I-Q}AnT-n1zgDY*M*
zfoEt<PTS+-1Y=bjUP?Er=ALcmb3Q~*!XQi=*EsQA>-_3(H#3x9(sB(pOJ*?#TmNLn
z2nN@WyWy|!DGD9#EVa)#CXr>_oS6nC_E;BtnGzNryhLY-bIm&-pdVo{f{R{a&M-)c
zE;jgv3|+GE3vAUa;=Hk4ndqcA0<X)(0AmK1MBfizj|1(iYFT*$vt4t_1;5Zko>Xj*
z5S;+WMhJ}h@h$1*pBYuZw6y5J`uNl`;qIpbPY#&<=8MF}$sD%K_S<wUH6cO~>5T}z
zv^7G;I-_YuT%hnJMV1Q8t;wtE-JP|>9bPj9C{7`6?C~Op{@G6%c+=2FELNz~Q~VYO
zEnhG<uu&1d@w|d#<APZARS97m2BqoArB=&^Kv+~v#G*3bh9(7~6p=Oot#r%C6d~Z|
z3$(Yy@?HVaAsKB0sCn9GJi2n>7v?AO#fd#WA}C0LMx=E_Dz}O?1{+q|mRD^Yd>siK
zz`TC3c5aH`PZyuR)=KO&W?*M7^z`$K#*Tz(D}F^tpz~Z>vu)ejq&esS$|yT5D`k&B
z^C603pySvJBx@tFFjT%|{U7^hN2`gKfy4!=f0hd<dl^ue*CRV*$L{tq7#zQJoRYhF
z#}3Wt=<@+1NrAKHAnx0>d#e&125KDHb$&<8cx`4++*xjj4GcgX)A?*+w<@j2Zr%4C
zT^u#Uv_!TI-ledhAnmF@khD9CsQQs++r5GdSi<Eze3NBYzx7!Y{2najQ3KQ(4u}Uk
zA7QxY>Xpp^Z*+BcpZ=d40Mxkb;{bk~ogN>zKKv)txyKnT{Ph2Q!yLhA3xEC}?!@~a
zcVT*i7bPM9eSQ#IG_eR&pxP3`6{)zp-Z*~Lx{^K;qmNk7k1+8Bn1m!T!rdn(CK7uV
z5sd4u1y+MXQUzo(J3{wb%X=W%QU*6dH?iwY$cHSjo0_N|3oqP@Tk{A(F>s&fYLw8U
znk_MwQ^bO>Rp7KzhTnuO5*3Mhnt3ScDI@8>!;Vyv5NG>h7gkL-LBbTVO9eD`VRS#S
zk@hZnafccM8Nm=qY6{;qiJQj4g>9xxLzX_LRm9>=VTPnew|CNbpXoiv?%A<UD$<I@
zzV{Gu2+=I3MhqQ)7jTR=%aNCtm?b{<puaV+>q+aD;l$u~FpJ~tJ>RYi$p(rwY0mu$
z_JE@U_>68Ow*>+~1*!%|B9O#!sv^vbEVq6>WvI=93uD2#jJFh<pBc6@+)HkeYfpyD
zAQx_Y(cw;X5Xgh5^3if84){{I2IX(wyy>4z?=fw41Yo2H$P;eptN(F!Dl9)Fe202;
zf$faTo-(2me=Ne$Ghvg3i)YX;9yAA6uAg`dfP7*@pR!#ejKpShnMe*EDRhKXb^}!v
zrHn#)k&sSkY;}EO+@sBcjfgv-%4lpm(U5})wW^RL&+Iae0udi4E<i*)+K8LZ%<GE!
zjDDm+0km&f+(Is8F=Hj44Z2M1)Ion26crslb<yqfYb-An`|bUU25NPjV`;R%Jd)Rw
z&;5CXTnC^LgqWvnYquVhfc_RcCIs)*Xl3A>5)IX;say>m+#6`p-%J9;oY?Hh1&cbz
z4{y8E0p3aiK^*s}CojC$LHdKbzy^{3a&s)?Cw;y~-&E+obO$_&EWh8se~T=<mBdG6
z0gA*DOn=~WagQCx6t}pA<~1;^T2(H4q=}8njZP<6CNK1i(Q7B>E%f#DfEcquKX|jj
zQ)u+%$?oLl4xoS2xlD-A>Shb?_J%&Z5)QaDATurG^2hsNYisTN5ci0qo6Qh8nLyG}
zA^Nq~E-WJn*l}TVTTP`^8&wZ}|0tT+1mo#iW`VOb?7!157DnPKeIzKT^-bh%3!$wA
zy=}mADb04#lf$4NMBZtU0bZE2Rz2|cgwso#5uYCxXK9*)TQ43m7`?5_%hR*1ypO&$
zGtt09q3~(IG71maE;_^c8lt4F<utS%(qF??C_t_&p-VbeT%ad0l2%E6LTOy!C_VM(
z>0Z*Cm?#eaz)yDa<ZW0j_XA@PA_gu|TuHFLnb@ST>!~!m@%~s4a$h}51ht%JM;X?w
zQ!zD-BRm|WPb^8{TlCsgVir{zCFmiF{1NPNxB?@q`d+<yMLH|20K(vzSH5KrgDH_E
zi`&64eYuig8<W)!Q@9nHE{vMet>B!l02;AEUac;$pbvWYdUj^iy|-b*G6WIpb2o#A
z9>q{JHW>q(Bi0yrr%b{&zytRUtqMMZS<rx^YW79{zN!P6h=@-FtH9|H1=}hO@=Z9w
zo;mZg1J(G7Mb^)sR|$lN71t!Rp)uPr12AbX_4oLAEgEPq9&xv+#K!=nUX+3$)9EZ-
zMVW9O>lcYB6rP`-=%KjoRFc@TY^j1pZPn(_O^pW*6y?ReG>nRjF?%;tub=(StkP62
z@`9(V+r<l!p)VeWr4GD^;Pl<Y%NZ1tXz_Jmz3Vj3s)s8Zci4)>MMqCJ_Dt{Cur7W$
ztp$74rJfKNx$Ni>{6#9L4^4kBS<zj@ms91%b-@w^w_!)|iz&pr&CJZ)T{9tuhxLog
z(ApCWYdJ&d3#bY}10QW2N=3&bpot~VX*M=BF>czlmr4d3^{COTFMT2YxsL;?JhVS}
zu|Yef5)-t$!W+G~tTA=q?Rd@AcNuw9e!XA8S_RLWx@ikZHj&{0idYagdjscZI@O)F
z%wL0#K9Ao7(}qAg1h$eQ^%wW-zg_@}H`*Uu(H+M}{XIBXhKA97s;adN<D<WSSKxn$
zkH!!MAtbPBaMGj1l#PHx%oT8G1Y;S{n=ekDIz{N7qPn`YuK^bzgF-b6gScZ-mv{yg
z7dlR(lX~b|#%Rl4vH)C4^ebIioeV?*0N?3NzPj)Z?K>Y311?JqF*|RHC`I^48Pxtq
zH;%wnjOL4{nC?E2zol*M?3hiM3s2Jf$Xwb8a0psVSY*(l?EZmDCXgGH#fU)`86}<o
zeXY$585tR*q?!8Zd7hxXU_n(9FWlSV7xz`9KU|Q0+3Q6VmzIRO>%QX7*+aGvtj%wx
zCF;D0i;JV%x<Nn?{uydDMoS(XSbJC3#Y%Rar*pH@MNLgwSd{gNdTt1q=i*n95tkQz
zgokUjvF`nL`BAg`g9n^P4%K<1n*tjw`O1|m;N?8*Hip>Q+8Us(lNJJL5N*^2KOd4W
zQ>y8!ZD7CJn_RU3e!Yh;Uotnx7byfwZYzpbjTcJzi+m`z+Q8O1VKL<fBlyWMP?~3G
zq?rZKv;OBpBPmp`?L9p_)cQ1waEO<(ar9m#`JE(BY7&Hy_nU-nRaMnZ(UGDm(bdnd
zm*^)QdPwhj+d)57VO|+XhDf-ymoHxmphWiMiPF-|dvxfTYMiFw6?XCzF5P4nT5|}`
z1D2ku2t*FW`KtMu<Y>$cIz_zMV2n2-w&VgvdN7GwxOnoF&kVYgVSc9k3AFUf32K(;
zD&6Yg!>40<MR4KDBs)q}pbQ=b=O9f(21PNaYktN`7vKkDNh1#z3AOvru&I(>M2h#o
ziQ|we(o$2OtI0O4f{hYG!A^xKu>Q^EJO4hzQh<@5a=0Yh7dNkN*dnkTGW@}3KKn0T
z`SV^pWDX2Q5J+LO*+Wh5@HGa1qKCVaMqJQRX>5k20ix2<()=mA$n(-Ws3^=?)zPZk
zi;zwQ6aKP*x!=oA$rR4meJb>&miiaQ-$gPO3_HLJjg5nY!+q<a4?iCkAHw*8RHs&e
zIFCUSohJ{y5Ktn`Ye5PP*sY04N$gbAU;c|1FFqIdWd{U(W<et_HjIE8@JrU8SBoN}
zP!=(Z&Y`^DzsiGUV?H_9G~RSg$ON57TpMb1#1zC$Lb5wSw(tb7QO&Kas-NsWg$rw<
zZHBa{uyKJOb7SHYO4~!;FSng!YUWtj+lc{{1`cc&?3ph#pHB8?Pn+sa!hxo{84!;`
zsj{HIrSWKYb)LDdS*bk9Vcp<<*wCA!v3kq?gMTg=O21@i(zViVbdf)0o1B^=^w~aD
zCFgVg!qsYXSUC{VF8^5(G9<YPNMqzcu+c(5!+e4BR8wKL9Oo%u9C17cT;L>Lc5I<z
z&Iq#9f)pOQy9+);?$y1*K3aKay_mt>W&$69ME{H#H)%mMu!N3jIWc&dN{qMZmR(oO
z&y1cYjB5T)o3pfD>kQIWS5cwp8BL#>85a5)nw<Tmg>H*0f}@z5rcfB_sTCmHp5LGk
zbN6mQA&Fo`b@eDL)@>g7?%S{7qkgWgu5)N&XZ8@E?mfIy)UN8&!q(3r&2t7yNC{zB
z1XRU%?W`&^_+aO?AZ!SrU7M<S^Jap!^Q;wkSU*jAi(wEsT_s3i63r%$+7Yk))c$F+
z1&x8e9yTgrqs69v?ftW$yy#=di~^oOJhp=Td?{E11k?89x3;wfL;s@6$#u2}&48RD
zMvOFWGNCSn4>mXr<0`jb$v(OXp|P<MYvm@qyr+OI;3o68X)Z^_FB?M-G`k48Wc~*d
zjCtQS=-mca1Rr;S_f<reV5cHEkj8cJC8p-|4#l!m`seSV9c`;;G{({D;iE?ac&N6Y
zKSQuI>TLTeDk}vYr!0|Td^NX`hZ4{xt86)pQpWFZFUxBKCE};%t~)$7R)a1|FvA>F
z?9PhAr6R%5`#pKc26V~$hwP5lzQ(i*@OwXHgG{&GWVyU3vAp861&;IQ&r=lv#>m+R
z3wh=sDa3;6G5vu1<hT#w`4120!AcGd4b}J22ZKV~=tH*1GUnz#YFnyFo@{m;*GfxI
zhYE$publbm93b~YY5e5D4Ky{2f!5@YkPFgiuTf{}>FIIt@tp?%ZU3N#)-PyQ_@}PV
z8q_Z7^Fb8|7<>T?AfkXb9s;dF_zt%I#A{=G64f1{Wsswl!O2%cY2LdgMh33FifqSe
zHtOGh{|)2Fb^qEjdBV>Sbh6h&1n#CiFm0dr5FL9?jt)Tcr=J!Jc|rk(Eg4OwM>19h
z?bf}2>p8sFq#=b|4z0|Vzsh#|3tLI+&wa76u($vom{|THBNSapsG1V&v(6wbQgoq*
z$l6WT4JOG>O?mo+TSGrQN<t3m@6pj};_wG*ey+KVnYK^X_Y1O*^bYsOpRYUx8_gCO
z67uQ!!ObphL&L*pfctwWV3mt5cD@1x7Wx7boXZw~f2}_MuffI1k6f6nK4T$Zbz#-w
zBYr4XyVSD+qNAg6h6uNgjt*i~L3On<R0ym-()>!gW_C33P|^S7D_}@lzMe%hy6-Qr
z@u8%Yp>6)_uh*!rzj`TrazBAkfm(GVdOf6tS<g0th4w{W23b~6{c88F)-$@u76Ln#
zX1eI2DrP>kSi9;SmI*pXj6OePBu7!)n+8uvx-6la!E*#q>ywg_f=%@3-5Ru=<EgNi
zP=O+?nOj>w!()qxi2Q7CXUCJn(R~{1lBcU{KD*TS`s+d<=L|8;q(xZh`nMp=qxtk*
zSJxi2W-B4t89SUd!%CTht#_;Q%AAo38FU(1x@#|KCk;?2;fUtrZbixhoCi>i=qtP}
z+5Nto3zIt<jp^0qV|N!OgHJ1M6}E#ecA9KwB5knn_Ia1+XBY&c%^?CO*>u+>g+uh%
zk3F-AdsZCtTZz7O0rbz}0lgb)LS>j_{e_j$_Y{}VL!8Bg%Vp~Vfl>{GcQ#jT1?CDn
ze!n9*p+o&Ph_{QR#R9cX37bs>ibw&KrJ3~F^Qh2b?Vyc&ThGXnHaK)x65+uA=*=Im
z`>=v*;fNQ2P)ce8`S0!6GsIp1AS`Z$aITx>>FEg@emFevU-I)iqvD9Wn3(D+CRP3-
zS}^=XaosOd`H36``rl6B`2hSBP$7t21NH*1`1tr^%RYS9+&YN%6C!$mFb7rbg|&)6
z?%?Q%npr|)g8=9M<#F+dceCf`?6DNM#l>?Ph0M&%+;<*V0o))RDzZ_7g?L=z%mIu4
z4AQP~%sBQ2f@vi1p0aVsH4=;*ep7?z2Xs9|!%$9V<#kI|gXT9t)Wg9!rlI1iK~cq_
zvlV_2MQn9a9A3H<1ZR7k@Gq&4Xf}+2M8S=7en96#lxo#m!n@#5fKs6d+*vMIs~HR<
z%E5=M>kuCejshCwOXfD5MF47Y7%r(69)qXP<IvR4zgMux!@z`|qH9wJYOF3}b^D6i
zzau0ONQs~ijtG;4=O;@RZYK3K6D(`EzRG|P$$<;R^9ctg0v~Z9pj|SEf#V{Uz1@$0
zILHw?q~j-VZY5jm=jyXmYB|UWatzY!T)q{8kAPtXn>(|hbbo*UK4sfc@I2vYw>S$@
zUe(YrGC>S4gYS<sj+!<>3?XNL5JXx5R5Oyd-E2em^BCY3B!1Gojz)+$Zr0A89%wP@
z<dhQ7GN3E`yZgRaR>h#xn$$4lkRO0+-l<E`!VV4&mL-C<5lEKww#Q`SeAksXR{lHT
z@1}MPrdM6s8IgPM^5^Q8!TY|paq_<W8Ocw|6H-FW&$Y~F7&MUcU|@@ItC>gPGzEe|
z(57EA3mFvb^8S>jD+$CaT8?@*5*dk8Sa_;CaLo;G0z9x#Kil>q9g?mq(9zv!M8C6=
zH!Cx9<jY<U18{qG5ZscjdVnE8Bnco_5>plFt_l=$<lb3soK*J~$NS7eL>G|ke}@J>
zh({;y*t+lVN4Oyy1hfp7prA5V6Cu%(p(tY`D}OtsPD9qg;X&~e6SlfqP!rI-9EsaU
zLr7$Fy0ju6y<0GV#l!DJClKKOz9Ejv0)dtbM;D=7^hOniNP+^1c%fmM?LnJBUg>Rq
z0?Se9<7rG|Rs*YJPcUxW_=-#l@)^b;<TJ_JI~exb%F3f0A#F7@0oXN*9<!t&Jd^wx
zpFI_&@TgTQ+aXPCqIHSr+(a8lnAS55@DLI|68JS41i5U^%t=U000!YGZzJIV1NZuG
zg$o9s2O9*e%hW`h2cK|_X-W@N^=IhuCPx%8ue!hWy<7B6EUzwLYFfzg{?t>>6CYVo
zQeHP#PR>V$gt~|WWCE~YR7^GC0wx|y+=#pf3kwS}j*^2zCLokirW}ndrC&gz2gS6`
z%#t#TxQ8($kh(+Ah)7Ozf;mhRyrTDg2ii}R)fjH$T$e?}ym`vws#&pIpja%ZBE%JP
zjuexLjg^%$y3iSS?j%W-wBvsL`qf9|8g>SJZC0(hi`d!N3`oUt_N=^uLb+ZK&LD#1
zE)S0F_MJPV1-w58`}(lLuI|lyZU4T1XCtzi?oz(Ow1WbfjjWuh-#`lCIK0s5ft;KV
zlrc!wsRsTKfDk=L3VGsxk&}~fHk91UqYxL%%%=fMWMA-~`NL7iVkoNXhbeJYr`C&B
zwL0q~ACkT!(0sLjXd&jm&*S^##ISp-nY4fJz*eh<(T@Y|1?imIR7v)RUlqVG)4&fz
z0(KQSaR-8{G1O-1bLRx!;gkZh0EzJcskI8NS%eYHvKyuO`Dtj0t<`*mUOSV<R0ST@
zxcx^?W;_CU2sem;`OmX38*ARQ9RRL!&>=ylV)D{{EC_Pc2rv(ER{*7tygq&(5ssXz
z@a&n&WfKfB3AS!q4uJRpp?(l&B}jw%rKO~dGgcEOpz7{&#zF0G(L$)D$Y}s%3<Cn<
zIs=`x@t{1lOeu*^TH*AoKt_xpjz5^`A_r%X$w(Uxk<TCrlH-efkKTNZZU=?v(ji1e
z<!n=BZI9J>Bqc~))!41XZ;09huVH}dnDH?`?|}o#Sa1X&VA~S<Iy*bNyC&ApH1+#^
zhIzG|!Ea#04eJv1fKR+YR0&-DvZ2l6jm#t}NA!x=ji&cz=b*nf3{o*Ud5GxF{TIdr
zLC}}3vR&kf;?@A{Upg!Uj7X?vIkPHgc+38I2!M(la!0OL&o8h+PEHOTqN?b7l6<sv
zch}6od(t6|cuRWS;N)ZXUvCdILmK!CF~zvCA}}R_Bs&^-4hJru<3YkDeFX?nq{$zt
zPxI}0VlW_tCG=5%$=LFqjMUT!l-)Q~E*NN%jS8SpiDwFHahjHeETE#ONI%w!2<i1K
z(Qn^YZMDh<UN)z<lpsAtLo7jFqI+L=>}U*<nyK%z26M{=Qyp-*`OTZ(TXvw#-)Str
z$n0W#J%Eb=$guWzTHjg*%xABg2x;6Cy;Sry2Z!d7bC%k_UaZ+W=;1sQT<-KLOsmT(
zdddz8AdyhOd?}<kVq1EH8N^W~lEiZYG<U&tny5hRGS9p;e4afc4YH*D+B<s-DVR{u
z6*b8@0)4|qa~4Mo<Xk-)fLRBX6h#&Rvp>!giNX7wrC~O?8}lnGFCe)q0K39k_}pdL
z!KYa61S)|ZApo8TTR<0xHz=ET0hr)5$=M6^%(715MgVA5ESBY@#}@7d((Xzo8aBc0
zDJ>wi4RIVFnZuCb@6M#^KoO4lk%RTbA-;+qMjI%cC<Bl#9qS?XrnO8np+&h8Pe8aH
zgs3RFo{^CeB^<$3g>yvAdl0WlSi{$V6B0G2pWtOU1qG8oq%AI#oHMOU{H)gb{rf8%
zQUI^=v!2lgoJ3(af7uJc0M=qJ9RG$C&b46Mhq;>mm-Jex*X_d;goUnuI06nvLP2%O
z3qP?eQxBwVF!^gTNvIOk+sp#4A=83Q?7`;q0qO%lCCyIcWIZA-OM0+q@uVzQ9Z&bM
z7k#$>B7npe65mm9kOQ|+58wr1d=Pd7fjiZsGap54kc|guXj-+y#FNIk$%%<C1qD=8
zY_I^<l7nwb#iq|ex>p9(P^2{g0g80VGNSyL$4N@$<OqZu#Eo!O*sfm(O`aqk05(!1
ztB%wIp^^S5R;P(qcw_ijKFN$7?nfssFnS0s>b?E*+FTJ-9OT$R9F}v*nk#|>rG)_e
zU~5^$G*K0i!;WxVOC)yX*1h)ZAUWYPqE5j$#6|_wB767j*#|(AXyT8DVKx%R?%zX0
z1o@I0rN5t-iHYe$>pVD_KZmuDMs&1tBi0f>1HzT``SVBi9fdJNEiL~20Pbl>TA)B)
z0qScF)R!JTc5GV<E2HC}CQ>0bG0Cxnw;-oaAPXoTHM|6#ig0mHp2(YK0Zf#4rXYc%
zj8ete6QYxqmGzdreG%5Ysi`TLXHQ@n(3Vl7+3&6|Zb*#R!EpfuF+wgOy&b6D;WZ~V
zZ(^r}QHmU$hHlJi-(NP>QOZ~i%&|Cqm(qRq&Klbcbe6){S6v_R0K7jrx-w4}@P?eX
zl{533-Z+5giXB^&Y}F)_#X^F2ZO$A!9*KUmkpABBk&!BR{@|YxK+7?DvJ8KId73wd
z45ni#hVqFBKF|hU5=zKu@^^bWq@G88usdP>p?L9ZM!{(wteauk^_s{fWUYqApSU>i
zd2n!0dYsD>E_f+cM<rN3CUkiO1eAy?2puI(1-fWqXQv9^XwB7I(oD))@W*jCG*Bj!
zZepExDnb0C2SzB}Ja`B~pusB-k@Xd!pGc5K13#1OH*&PXr%#{EXB`x#(Yi_?27I}s
z2^;`Q-$%&=G&zZ!XmW)Y)+mM#ng)rygvtlX$K74lA=6*hhybo``?x}?y}v&WC3+#a
zr+|O}f`Z*m&_ivw*>g#)p!y}s;}QU$i4EIabS<o`VpVfIw_nY*N25S(_W?6#cgFX6
zrboJhaefFo&8wn&w7W;Y;W$+65)e&lyp37%Afd@X1H*@E=Gt>1*Lbpl@=4On)u=gT
z0nUo?ofwWBk%yoQHscjTsfN%L`@w^6q^)M0s;EXrQ8+6g9*3MLPdS7R2YZGNFB)jS
zio<zGEHc5mc{H=wMHnq2Vc`Fxo#*H1N3QiGm5OCV#BS)$!%19%FdoAe^ju3i7Vn^p
z^9{(Eue$y?{1W(`aNj^~%(1U`xMLe3ufHN^b|TFyqt4;=RWb#alxHspdboK>^<Ubm
zxd}Lu%C6=0irn_<6z8I~{H-Q9Vk9;oC-H7}uPi#8sRC{3K6auQXHF_P9m>6+$SFQ1
zzi~iH!aLT@-Mtz|D_P={PmoDI$2;(V?PGk~Ln{!krH#}3(|fqr+7>;Im)2=SaBZ`0
z1j|WEwLpj8|G<1jj!-m4u=IZXc)#DbcE~3_S&yrYVn?=|j;>a}>>);x=!1AtpLETG
z9JmOPiV!hOO+p^&!w6@_Lq9%a&v;Ia^+l5b!;T*9UVBqEp`lN5;5%IYx=RN@-j(q7
zMUitE&<8-6`jGkURP`E_50#Zkn%fyNn!efxF<XAmVZvbC^q$q|XgSoHYBX%~yE;?c
z^vR-lH<>=M*us<V>*r2|rq^aw^*nv?f~#xZE=j9!VE!nM$;ifwIG@P}mX=PplxV<7
zandQ?vOyN1I=Ud-kYKxBTni}QbYA2Yo`@XriXM!p8}o_7ot>RJV~q&XF(3u&q*KDt
zJgyL%doQYTdV|3uP>gRKr>%+J4|a-vqQJEJ{w<_60vi!=9~k=y5^4E#RZ&sVFMAw9
zun+Mvn90n_K1ld=;ua~rESsXaj5jOHAZZhJ4=EjtAW6_1=L5*`QAPem`bpb1eFW#>
zN?iR!mIgU+?CLk;6grxOZ-dBFKsiF{hrP0DBOgKwCuc%|3_vW-_%&UIxQJLTxxP7Z
z;LNpaoia}1+7rp>msOxr|3Tt#pUXhhf{ZT@5fRbT+tV``{4>m#%(GtTyXTt}i-2Z<
zvn9QoKcj?G2pv~Q6@fnwJOZ^-XFPFyfU?5_uZfNa5Pa9HxhPTZKYu<$N<T}dCwjnu
z?g8R-x=!ee%ec|}Sp$P$2i{RTFWJH|sEcq*hZS~;fGi=7K*Uy3#1R@bGb_toTmVl<
z<Y;BPEScFMPk(={E+Kpd;7m(E!$KK_2Jux5;NVYE)f(n%SUas>?_s31LZXDU3lPi=
z=P+D=QQ%R<#KedQ2QU+krtu`1(>r;fa-wS63~U2;8MgOD#{U1s-kHbcn74obGRAC{
zX2uv|hB3krS(0eVGG>gjFBNUZRw<Ge5~Ug2*v7al5iJNUDkR!eG=w%<Clwhgm4sx8
zc;3f(nfw0zUa#lR=g+6t{l|UZ>RQh8{2t5a_<TM`HMx!>BYH+y&Vh*c_rDJRgkj~u
zz`)7}57u02aid)o3rcYxLmyuVrY56`B~!7M0kIAj|1V*o7`hy!aJ8)ba*}6HAYQqr
zJnZ7Q3x4$j*;b@S$cXGJN9P0{JXmmg(L@4=n2qyH$s|IVldhO<dDkgrMQ1TV`k_o3
zXu3XFwB~e7GLq$H-F+C<7WdZ!Z`9;1&#<u{B%7OCSy|l)`<n_eVA}SPRG7U?LH$dS
zx+^tAIm$w;{!$!QFAAQ&yXkgmti#j|hzYr1y60WcrC^iEli+@JXNv!fJ*w0J3u)p!
z<Dy%I*hBJ>K1z6<n1s%*`MnAZW2$&`wbnA8rn}4L@S7`?jJ%o5`DqnC`S~=xKA7ti
z-Vs8^`7gS6f9Trl$FBn=n{RyvLbIFTNaZ6<4|S_JXBs}6JY@<jfLMv<k|Y_U9}%F|
z`A+o)@B~p`affQe_Xxt1G`ip*PTSQ_8Pigb(FjIS+UwE{)0A|0_3Lkdt~^q@KBE!K
zZQq$j6$VFQXKbwdo<8Ytp~hq{_7G#Y>2xddYEX@~_IZ7UR)X<OR==G05zj~I)X=^a
z^?38>LlASzn)uV!5BodsIf?1hC&eov&8nY@%O4eFs^f}Esu^21i$=Mmw3A}aih16Q
zsLp_1<LBO@BC$-ITT;Jq(XLCa!mL4p!K3mdmmjtJJMX3GS9p#33Ir$Bf1S=(Kh$U9
z&qfdxFinOSq>BjGze1=H#QuWnBm!Wz@Xe>6#--0c)uq?$Ww@-0LlNKvuROH5uC#sy
z%*IcCtX=jm7eH8!d$}&iO1-y3?GfYbGm{JVFxL<`g879w7}u94Sz;)6!lGL%>KZ{B
z!H@Ya20_-p{PN3@02S!}mGty=-v{yZO)9$oo}H-(>=bybG3R^H?iD~JU$F3Mw`IvU
z!o2OgSFNg<y0&?jaFhrWlck0Rp4e_!OB-VJ&MBPh6G;`TelzIBYZ>6<9Dywa2OeKE
z!9S56jonb(NEwI`m9)BIM!6RISJ3-!YhJHkMkREyciq6Ghdfw?oP(`r9pG@#Wfg>s
zT_=O;_l+sXj0il!)EDPp!X)U?FGq{}b-Oe<ZEN#A<`jhk&y~%e^xogo<dOquxFlTl
z4s9GA1N};7E#Ugs2(Zv<Bn66m2k_rC54|U1(onpPZJzf0ytW6z*lpX|39iGbHv`_3
zI3Vsgy?blVE#5*A91dD2qgO_K+k9t-y%kmtg!fwUm=-oIo}nPyH{)B2Cmg6lp4f2G
zH`4Wk6<Ue295-Xn_=Uj;6VMq>qrJmA<p}02Hm406Hvo#rSOc6u@)H()GzOR7X7-5}
z8er|Z@!F66(AFO>?0&jsrn(Q^N(mZA<c@WzEa_6lgDC!hysIFpvl4G(e7$hMN?=CP
zl{jNI0Y%!Jksc;yq<hsYuOMKJ*$mpI;|`1+^&7WRqS+xqFsTVTbuUHvX_^P>r2F8p
zH6v1i?3;7fzL{Cm_y{Hn@4Ni!jXvEBU0}j+6!U!i5?$x*!z*xr59gsP{?l(BZ5ll+
zJ}6cUBAEIS!>wI>U=|`(EizkH^6=LABrzHmU(IH43o7t4?$uL-t?B)?<+=7%>)eBw
zIe|Ln^QZ~pNpF3BOP`}OIMbg$EH6x2xs14T{KDFb8`pX4cr<Ikz^~7J^Q5!LIs{E7
z#%<ecCY&lXlp-KCQ-kFn<2fZfMB?Zy-AK0mO!VNQA}8>JrdWnw`pw_pUjh`rC;?vs
zcg~8n&vvX6p2Vgb-?g?Q%L!gvK<R!2z(L&2w&1>ca$dSL@xY^54?eGB3~@t7&(s8q
z0pvxWMLO$aXlN+Ml!z^c=G%PFFUm+pRbd58Dh@;hrmJ{bAtx<f|8YmWH!K#X)lry>
z39c2SPO;>s6Bp_pga>IKiQ2Y{qo5sd?9F+y$n|;wX}#bmi$N-PQLuU`M2M<FkRa7K
zZ&O8_h+FTGGC6~pxH!hTJl;^#-K(Ni_Q8?)U9zoa&la%H35JA_{sC@6VOtkosn4_r
zl+Wutcg-Q;i-m-w&F+D@%IKDwsW@jmTC{A{S*xf3K!WIms;=iBU2MxVrhq%x1n;Ld
zp*pBn*pU680wno3<V9qvQ`b5!N{o;Hj+zX5gQRq17VsFt#(TD1vtWLl6>KcAx@4%j
z_#J{s;LB{v&AoEvJHvU_=I}`bRAt&^*r%-R#~<HO)H+`UF?B4>)_X7hb3x5TJ0cMX
zP&4LnW6Ye`@luoot$p+IlhQFZ;eZ`o!#>$=3nC--X%ydfU;M#Ibh^0yV}a2k%YIH8
zMcB1E?;R(kR|<rSvm3Z41d2K|PSHgV8bsZ|NPEHiKZrzAyh#^$9;qc4;;PU@&u(G#
zYd9?h2++3DJJ>7{<3iyuQr(Swx~X%RHc40-E-8o@3e_?Oi#(~-xWUHOR_sv)2Llun
zr%%D7qJsPRl~DAcVdyQAfWiRPr(;?wK+x&rgiB)kiJz0}^5sKrUgWR4)?acm!wNfY
zNAHRa=`;k*cYD0W9A*&O+vDSZz6JVK4xS56AY-fT2M_plHmCs|P@c8{bKGw9@InI4
zrz)cT7b{Jml6|u(kAiWCqMMVECsXF7-Hg3XvLzuXviVXU)6p2e0!s9*S>yMD?-dCm
zW9d1?W<N!|@QMk@Qxlr#G6gu7u>~iCbqec0O9*>6bnXXdQ3T-bRoke(!Qypj2bk|&
zEiSeco1dZ++P5to2jc^>ah+dh*CkAR+Us^(1fRB{$gxKr?<2LUcWwA_*hw>cRqIsz
zwDdbYy&`$`7GC|Fy!rfaJu@$PzNP5I6op6o+S8qXl6ZtNE`7=LlcgwfWm+mKik;4U
zWm#3hJ=moqkX*n>_de(57q<dbLsK)0^VZLB@nR}8u@d0K+~}-bIOEGlMNhje4oQ&_
zbO%%vr{Gi=WZB2oFx(KBWYEYVWs=psG5!*$5ZFES@^9bBu$s(>mVc29Xt@ZwWf_p#
z#7LEQU^pPAyTF576m64m=KD*ATPN801Af#0Lx}s6K*kKTkOu~6$GqnD6%qQw*>a+l
zdq;}Dwp0tDIc!OQr$=3TEcK_$fp1a#;^-?1C1+f54ZPd$_Qwa8Xxpu$xmk7UWf=y3
z(7x+mznj@~`oGPBZ4q=8aX-%hxXHMV&lk61AnsUT3lZQo;KIOA{x?cbQ}z>|CFNrJ
zsB4@e>VMR+#L#_K^_yQxxh9e^Mb&zHP2={OUi-UAfUJ=NMd~``0%jl_+R8h3=FMLu
zyEAUg=IRO=x#I6y_q4hh>pUrY77|B5N1)rwe3l+q?2VC8OUqk$e^4+qKjxtzlu_3*
zNcYMpTs$S<wJx~lx|A`!;+ai9M~ZNUq&F$vqAjqOSVN-~nP7DL`3ux+g`g^vquY@r
zR9#1n7y83z$zfXH9~fuKK#b}`2qgg59Q*MTKcG+JKY$4we(n0MU^T*%f|Df-FWMb)
z`V4#fs@@K#Wr{@0LFA5<85t0dfa{URm;9Z)Dwa65%5(;0$C+O){ATpAf48j+7Fu%O
zzTQFP)d<c}{)d4exVIg*F8w3M_pl(lcHk5KvHdf6hcn5b<Ob4C%=lbuJ_7Suo-Nrg
z!?>jwIEdz~Y5t5@RN-PcECR|YLB`?GUBxZm-oNmh2SA%P@!c2^f`(rDNMWX_;&zmv
zNhqu$QquU!)R9pzqav}V;K8i>DPA{F>l~fe65}^uBbTyuEgQJ(l^4sQzTM9@ysc9v
z1(2W7c%eBx5<^pXyr#{N)z2_b-KFcc1ZYr{{}?q7G-U!+AdFe+;D1>~!U(@!!R4(X
zOb_qA7Gr`w7I*(9Lq3HS4EQlqEm8QQLN^~Qe(fRCyI_<3*$*ysQMiV}h|5@j-?wPs
zux?i0V=^t@aG#^xW!1Fj3R~~D0|yR_|KqEEHEX85#5w@(-QAaDyfeT_rqvFA`VbWr
zwwOXt^Sca*Lv`J8+t=S0&9*~X4VXCj#N|_5!l^Q-eEb3m@hoQEll~`VYbPI5Xb)-`
z)RLhBY<4b?PL6hyy&2z{qN<38@3Tn5+fpfXsw*~KCizlM^|AP8gp8^oaw}5gCP?Ni
z&f2G8sf`h)=&+%b^9FsaOL?A9Pfm^w^DTv;Dh>^%>e$#=v58MeO8Vc#wo_J7;w-8I
z>3DHX-cg|k3Z)+3!{Obh1`HEXxBe*&NwGoUkp)sLcgtibvIiE^seop&O&n~cFB}=7
zjo~}?;k)1E8TbxXjzaybznyyf-j>?nf`VDuf!`N<y=HaBRXB#T75z@gw+gQ@WBw^e
zVOQS`_HDlUms9O=N?k?LHVlOUs>w_5_^6#Ubj3wJUcF_^#;DnV&m!=d=-fjaO|~{9
z*6ICA^c4C1_3x+#x*nN99~4RW6>8%Uhn%mD40<djT9+=4Za>M>;CD^oO%A*{%;o{E
zI?bqqdz+HugvRua(WtjUx%W*Cw{Pzou>|(e_}r4q4+wO__x7E%?;9T7j*$u**QSHQ
zVtn%KnFZtH-lj4TK-h4#(|)0RiBeT(m6z9j;px7+Y#JVD9Q9n+%<bnV%dVDfx02um
z&wHw95JspJ%3~3wGcRVU1|JAh2(3eZM&HBK<XclHPWJD9`6K0*?c=O83LwVS6f%>c
zw-RS$lY!1+aswFJJId%7bLFd_XKbULDYm_v_v}k2#fkl4_o@2^U-*mh4nGPP`1*ZO
zE&+n6L+OCN8#%Isds`&rFzm-n^Mv(+N_fAG=H^ByM=S_y?1ZaD*Sv2G|B^;%*j@&W
z&i|7W>=&wexb_a%iCfgEfJBn9c9d<Bh5BAiO*kERfadio%;b}P6E6BbfGnMN%!=U&
zhZ7%jAG*$a1~G-{iOaitsw)zkw6%2Ld9&@#IPToMr=WdOxn1<vmp0@oa-oVY@&RAO
zknXRhr;k$v6bvr**-Wn&h7F&px2ZrzLMzg+y<Lwge=$f)OH2HB)w8UuZdUyKSYqJi
zC-e{tPtBfmLQ%^6%L`|8eEA7^_c-b!;f0f}q+igv$|Da@-p5iytEIY7sr0T>vWP$-
z@8#7Z21twHd@q19BSRG(>16-MUGtP{@#uRoL^$;tVH^^l1G6XB&px|k|0$ayji|E1
zOZl{czAeqd8d2ki5XMZ_w(z9@T#{-yyBwmZ&7a4Zp;uMZA&h0G)A}d0Q6A=q>`M>e
z=7ud3!aAh=^B@oEP;$P2^loku&DDp2SY%j=4Y+V37WhzA%5Ro0&-fY&bkyUUk@G2)
zptT*;3CBNxK5)p<4)R2$6+(tksXBGR<cB7<r>SHos9mr6_gD=KDAqXMZ(desZa^DD
z=%nwfp@G)-t|*Ele|5=3r7`TNa<Ru`E4r`uC<>K9PX>2!mbZ8Fq8%cp6XzFVq;?&J
zmjNSm2`IBW!>;+TyVn>pjDhuwMRaRj2li#$2&8U(y!b=;d<6rn5FJh1EVQ-S!ilBl
zmRUjq6c!Y9-dsJs&TpFs&{E03QD}{`J7ZD{94CD1@vAq%a=VX-P7;%1;ddh7v*ouQ
zc8XgY-<AG*7R{?p4WD+`@#7AGHi}4K+V@>9APVbEf6(3geov-v6V>3a(bz4IrB|qY
z_B|I!O!VFExG(~iEnnaiyW1`UdkiPH{luV79m<&LzO+s29Nce#v2Cja+$t{dRBk62
z);WBPJGUCP&^+O2w=peeS2fxz-QKG`C3q;5yJZff1eF;ypo<>V<eJy-FBLNa&&+}6
zP&Uo-&S6S)9?xgBC>;a^z{~I^ld4d$F(5qg=gly&v@f*XbR*QXW66@UM$4Sv+lT4A
zsIAq)NmF9id(&)DodcN+et7J?irH8c3+`fCZGF4&>8m_HKR>4)nIFla=mid=HTw0*
zA)l}0ON$|I6N=~iP&<ZRET>1Vqe&^o+rj&)tEHOiR_I(4jd~>K>>F~#B6yw0s4%@(
z>kdP#TAw6$^}0$<&a^fds^m#=#VgD`lkF5&OF+)UUf|l~&=n)`hYMyDt7AI)?dw?D
zR_7D9w%Gg~G!MQ|ocBO^nK9+Rs8Ll!$k0xf?a6?Tnv{ep<7*KWP$$K_SVN-7UwW?m
zm~iOTp*+K~qr<?aX_fRUq=nGNb6*M`ViB##8N#edh$@>_)ZCofaG*y6w-iKW;(sp`
zTiynuot`I|Wqgj$Axb)Jlfw)q^?LLO?wYC`hLDFh5Z|aelJx6}td|*d&kg+koJDl3
zTcz28RR9lhZVHcnOu5KS(-u;c8(0}Hp=jE?d9#7{8{$+HZMOOz#JKzS?{_s(He;n3
z%jepcMHeH8@nbJsMEGS=stkVB9(ZtAM9x)g5kh<w8BupgG2JNNdz(V6umm>W9{B3w
zvv#dNXcu;>QK9>12{V6;$#tF*8Jy?6-1`aClBL@Zo<o7+m|OASfz5(l>B@6k`=P|g
z+Re;&yJaGG5WbapusiepGY)b${62uMtn%|dj3hw39F>(sKuPhn?si^(kV5R5Udr#O
zn%a__RZUkm9PvWBNmuyHDOkAROmX+>pcc8Q=)%Xe7JB{|6F2df!Y^7`Qib2BQ?N6Y
z5zGyL9l_3T-MVEkqnEsF$Qva9bzko0*57c;b2)y`5=hMR2WWLSe&KXY`Mw1+<r)^~
z<MQJVM6V<6q`<OII|cXBa0K`2(4U84M`>WCc~xKKRf8Twz{i?J$COZu!%0)8;01It
ziedxP9~2u!j@#u^_%;8kTm_*nDr%lKZNC&N*=re4oW<763m2BKt}$Bj%hN_H=Rivk
zEoH|Qt;4|CCqu4fxM1Z)X}?>CVs(mejrVj_ZbFqjFd;#SO-0t*E7^wl1K6FpkqaJ(
zm!GsZESq*$6Ju)e8{l3EE+8x#b*QMmZQ>uD&UkP;b*@mJb%e~yEIT^;bLEBmC@<VS
ziFCBR65ZO1)91b5K06QC(PMPb)k68SA-`&Vrw;fL-FSafnF}VUo<A2Fhs>DN_wL*o
zyILbseP7aESO#HK7~4dx32H>p?l5QU&IAKu5vg2^gP?ywL9IPOqr;^esr*f~^l4Zj
z&5_I=A?z_);L2w&As-wCMnN*w(jt@)qnM8JSDGfmRi$O}^iL&&vC*T5{hJ=o9?Su7
zz>j<`Jd$|{J|@P-S^YaHf3iuDDyVFNn0mrO7|0`LD{5bW7=w8i-<L}Rn0?t7fHIc8
zY4V9Llt1&A)Zn+y5uA-4d)~I3?95Ev&JS@OoWY979djy`D>(^65|v-;8hGSF#Ub)%
zUaEY(eFRlZo(y7t?!M~BDN|lMNq+OYy#7lf!H9h!A>T-LA|O!<2(PrHx8}t}!{fJ(
zfX;vHHS6sLfA)O2@-XxF>*Snkem)d!m*%I#e^&`>L~qLfJ>tJRie0D4F;IRgLgcC|
zKWSs7&rp8;{|o<*OW`(ztQ9{4Im0_hjT%KBHRvU=nVJh0CgQNK#xrM@bxKfJsWG+d
z@ZS_?9T><$4Wd67e#%U9c@xo4r}bQd-LyDh;jf@~)1I-n#tO;@$}?=30Ki<wANFW&
zi04*Z8EVh-D34`mza+u+d4O~D4dK0^zsHlkOWiKD`Y9~mI$UD>so{^<Xx_;P$U@b0
zQ`o4KjNNt!?MDt3Y0IEUL<J-ou`uyqX2bs!?;G)QWQ(I=vnEIQeUSN|iUq{L-oN6c
z45b(c&sjA3m`h?!9}~2+Zp*`8#D*ij5t^WQcW?I;$|Tk!vL>ak!Nff3vNI&W7XFPh
zYC#<**YWI<a7NO#lq;5u{c2BiAJ&eVgo+KTwPGN*EWHoI+~LvCmxxlR<fzJ%?z>CF
z0yqi?m#EKG5t%^#fI^TX^^?9(6;}N7&tdu@Gc`O+=hr?p8&>w6+YhP5-p?Lc<O*$_
z{#S6MyqB)em#uwuV8EA|i+0$85{f<w$k$AbcivMJ9kLEM&aRQm6#-eK&TX_6Gw~=X
zSs5|KS#=aZa@dOZlVum-J0P`bWCiQ7{k)kD&lU>zXAj5!{Bs=|w&)r0x~Ad=$s9Bk
zY}ZsRtUy^0$|Na#k2iyN5X>GmK;+`adTm3Jot+&twN%IV9|K?;Ah#hp6q-#|ZfnT$
zU)~=yEwjso)s>^r+8xRM0RlT^l*1ac6}_493P{nvckiFn_v-7T7&uI)6h3jv>MpV^
z>X907kGH|+zt`Fkk$ii!Z?4WRpSd5eH#za!fN$SBRB-fRUgEmh;pTeNcTWE7eXFsN
zMQuL&^!*>V{Qk-Bvmz#3s<#dcc@Xu@ci;T^{zoN-SB76bc)Q2zs$n}-Jtqy-zxvA4
zHlgC+{#7}B`<7;z<i0enN+GtDvjZp*tqcI-!VmI6Ev4*Y7GqH#m?A4d@fdZCYQ1FG
z<gALHW@far8)lv*raAVH-n+G3TWskF$6^wa&_dZ+(64MyC|vr*Oxv9Dk8Rhf!CXkj
zCKP5XZg@g)6)!bdmc!C*so_Y^_hnZ#%jjS$ZE`xa6t`v>)`*-2N><eyXqrwSFr96c
zgraivCL(&ZOCK!bz_8P&XQ(~UN=i-8*}$H^l_`gwZQ=x@7Clyn<e#D2HrlZto+gi$
zGxdc1_Z!KG&b*?qBXPCz&B^$g5gQu&e6jR&zbzZXpaL5eD~+crpby!)`vQ1r&HTG=
zjmJkuIJNn2>!z(8PWcc<Go#kQeTkt_tqAoy&yU=G?V3=s&7^{+*6tmcvg8S!jYyTU
zK1~?T!McpcqgZ;ql8?(My3kvJR#k#ybVvIpv&yPxVbhP$c2%8zjOVrVGBTB~N|A?s
z=g-}5P(I|moSi;c%3YIZe=iV(2S7G7KXzSp_JQMg4^oFU{N_xd(-n^|QOvsEW?%`A
ze8q(YN>=eZw)v+bKkv)mI>$B2+Atage;tu<2v~Q*qB!-#MO`-HIOxMTC1XjOo~F#H
zAjOo2X6hwn*<9(1`??62;3h-^OoiVZA>B5V&kFUYH}AMQf?$hAT}Ooc0Vb<g1pcsK
znB9QW8;_0Z62wW-><IYdr=sh#PF?AG-RtC)t`+6Rt9BX?4T3&<6ckL15y}JPNi(%%
zYbbyPbwk#Qfm>siK2y`{+UEdw!->#Vst3}%W5S-8oWi7t9xu!0Wc|6+mZ<JNPpA8L
zZ+#1Bax!M50c^euc}(4ye4y`#$!8o@6)Q*_IH`$OZo}==TRT==?d^XbflR(2U`%)G
z&mtnhfPuKM81}jNSk(4j(*d(JqWQ8MD`qh};u!Jz_fWcG)iAx3hrm)18^dWBxccqg
zY#5^Obk`zaxK8Xl17Ev^KN8O~7{9bng)F`hLt3HY`|thCTqKiP#*h4@{xQvijRO8H
zdy)(5mTPX@b4rWr_ecri`7F-H<+fYkx*w7*s+n6Us!^7M%dVt{dYa0^>pQx2zYiay
z^&-Ml-o4wa=IKA~GR!&<smYT;QkfV}f;s2`w0Zz;PVU}vmz8tJDp4(M0a@4YL$L#d
zuBvM~1JWY%+4`z2nq1absvRxSx-dIDPzsGCh2cP%2FFxA`jH%5aY!%qwOhH@-R<ho
zCr{j{rqi-D3Vqd7@h$1TgpoV)1`NpQn<+I}eEsgDt(T8B8{7pjda;X`1`dc<idoep
zdEq;<MV^B_QH^MKjz$Qr+DE8JOQ&UU#HFEXc-v^@+Nt_-$R52%wHkf4;wH6h&40du
zbQ6Pyz!~OJgCeqdjNxmN282{qxX2szhA8nA?c9OVf8Sr;WNSOXE-1|EP`n8x%(!Ua
zFjag<+9|Kk06!%SS3SeNTlg1-7MkEvCSZ2*B*53E(%Ww=l#S~gbnT6Isv<O3^?RBS
zzs};pE*9jbqPmCh*zoU1*8DS>sF><?p35MsGZWY~=AwbuRq?Im-NU<KyG|@sheCzA
z5utg2`pB~6WSQeLAwyC$KxkUsq7?HEmE(NnOQCbIbS^x}qlbz!>D0H%C%XQp{9C37
zX`m~rb@uGtTZ?(_te<|`@b7cfUb?L0gS>X|+t&^>L8fZo=9<0){=c%#HE+LB>hSGj
zAWTUIvagi(Qc%)=pB6_{ZmR1OY?fc@q}Wf^3wQ(BaK!=zjaG%E@GOq$H53&MXLfF0
zjc>9T|3P~w*C79@utD+)zhhu5=#6-s>d4qtkQxS&52~&yUuKLCx7=@hA7=DC>=sb4
zG)yZp2m98l-u;$er1EaUlLM`oh+<0}3W>oPy)O-cr~4x1wbJ%*VM2ByWk{j{=_#sp
znZ*gXM&V^>sd-v$1SRi!3hYAMPM}Ic6NcwB)%fvxWK4JdeMenx`<Pz-h&4b&*g-Kt
zEn0&i3obhX4!dXuaVnbuJjssl3CR_7XQJ3U$d*PJ3u_h7ZOfKV3@tS=saB45!0+UY
zT2aljSFOM4(01d5Oae2Pkl*t^zbR%80e^z&(UVRSp?yLF!#{BkJL7ui2Us`Pwijw{
zVq0lyCbPjHU+BlnZW+feT9<z3y3|mkF_mx582zSaQu)KG{>rb9St{SDqbIh4jsy^K
zE<aI}zqW%oYa`$lr&4iD`N&)O?MJS1|JK82afYg(>`CVj_q}2f6IwiDMo~UOr(yeC
z!)49ur<J@&O!B*&HMpG=G@jlMu<!at?t%=TMFAyrLD;d;18(&40X7xmmF{a_FPD{5
z5syc0oaa~&HeF5<4r;|t=a>@K!F-RwPvsEcsA%QkR?6KK)*29L<%tEOFnDDTkYZ-D
z_0m62a@#+e1)B%^(ngu$NqdVvX|G3fb&Xv0CzYAPF%S%qz1RAqd#J9DX_tPTa4F<h
zvX<!X&zzaXq)DxV{Uc0QzP~PT8Sl&paQBWKI}Ye_L&uEkSrO%cGEPrBz0+N<w&{as
z&6(3LeQ=I<FtpTKxD29hji?_4MTOZ<Zv0xtASu8(J~%bkGS9^SX3gccJum>H3K2<_
z<Y}k^)d5pKdh{({w~$F!G--P=IEut!9pLjiP#akRd6bH`4q!GP?^K}@8CteeZ6C+^
zmu(!f2$s@TdNKBMh%yi9THCIfPI0o#oZT`))6vK(LMVxaJi!AHc|7;x`F82kR;@~)
zbKvM5<JlegRR^ZG<T(A0R!%w08FL))kyqVL!U|-%HJ9IzK?DL<hsuqG8ra@8z>>A{
z;@2v?Kfsyjn@gd05c<UIvmnK1HLNGlAa?iI09F@M49RVZntkdAUZO#J!AeF8Lxv2Y
zecTz%9%1`_hfT@fa>3+`L9ua$?9}`qHWzSO&s2u{qU*)vMBr~i)KOU-z%<j)vV-at
zWn3K>UoiU)UyTS6X#x{qy}H#8T}R0FU37nBmxgVn<Hn7X-jed7A9GkvG3FuWWID}I
z*y)-5^Di(<28YN0QnmwA?nru<RjtTMXqCuTaYmbH%y>p**CQz^t5H03-$7T#AChhC
zHux2{j~F%Y-jjs3=b(=xzZ9q5QFpJdWc&^d%H#3^z#Gwr`T41&z>-3T2KIVahsU?%
zxB+24g=Qxb+@|IPRZuj@s5ixi?537!l$flGqgc(;t@U&?5~&#Fo0HrOeo0d`)t6dw
zZC#?jQkCTnZih@qo)g+Dp&2$&Ng;JG$D!F`!H?s|%Rri67KsNR4o+fa-RkDWYZfg!
zD!gn)X`D`7L&M^BqXU;^waj2N!>^uU;?i~5Ijbnq4-v|L7&lHNO+r0nAto`Dv0ULO
ztZYP8#CN20-2WEfPxt{40Wc(5bE_)ZDw9ExzROfd0EBJXkITsi<ibp2y1o4|?R$x}
zTPO(K^uLQXMe_cR?<;!AL~IWV*(~7yRX6asQNtqq&N;>06en1m*o5J{N5eId3s~6v
zk1WG8_@hy{hq61IF?6`o)(+-hrxF1~T*Wr{!$fn+h3R6ZBf6!S-~~)DqIOsxXq;}2
zD4v6OJn;LK@(D>yGog*JD{^mF2W?9W3n#dQLgtZ8ddQ-nD1`IALS-8n{1N|Ni*_Vm
z%--n5#1T}Eh=}A4KL2Wf_1Q^kq^!Y>x1RgMVCq-%wCz?d2B+uX@*dU2YZxuRpPcp<
z?dvfSvx$Bj#j3J-m>X%y3^g?0|4TX}TaKB(0+agNuGlYRDBy(tERhSWOmB@HG7lm6
ztH1-|N(KgzH8C8z@jcfO3%hKjPXd3;3SqJH{cUd!Bl5EwogU5E<-0%=zNcMHV!gP;
zoSEW1s#&6?{*P_Z#v&bqi^ffEXsW(=KoD&GiTU~Y@U)i;6PBxBPN^1kc{6exxCk&M
z`a;9BBiaFuN7b(JT<`v61MMP}jlFLuOlEXibobS3mN1f{6{z)i5Q~zzddUUPAqQjK
z(l1`5ev(5&r~)A;YyoZRPgwCr&&c{08^$?!R8N<|al$exhBIPr-aT@9o8%n3I5JJD
zP+OeNnFsb>p2CXG4nJ=<`6ci!a`QQ`Wuk^x0ecGbngKxBlGF;P&faA=gS6=SAbNog
z9lPlRr(qAZ0iRvy35ZT-#Fcht8#wXLqSPPtw8#RG@>kBuW!cGoSziyy()~7_spq4L
zvFJ|T9PGJ@uf|`BLoR0T!A-snGn*PmmEMKUCxl``Uu0O=?_YKu(}{pe?PJN))$~`#
zsg6t-nqiQf!Dt`(PG0hD`WzYON&jmRDHobl7prZ9D|SF^Zbb`09%~g?soDG2nZIQ#
z^pWJ`WUJiSC)(2^p^gw<ai2bylmAMt{h3X)>pl$LI8SsSRJCGf^VTzGd=^Mj=gh?G
zDG?5L+nuE>7LGOqsKyzqt6$P`-PyjkEh)r4=XcMIft=i`O^tO2f`ivySdhW;XRd1D
zC<gwv%`HxFrxdBch{pCVebWcaW>leIQO$^au(bP#R0piwoxYw@Y+<Xw6^w*qqkaSn
zU?aIc?AWZ=C8^g9eORt50z|YJ8!O=3i}M&_$XWHv0(4r~PSIOnU~YLzq8)oHgg+|d
z8~`}IwaX15TS!AB8|-;%(Fiob4svAD>|GtW(gVG5{}ChpKAMrd?`}}TniGUstemHF
z)eT_d!hW@2<iY4umS9A|F%#EdP`mv^xvUJ)sgv79fn5?Q24J!sR7Ni=Z>{>7TPl3+
zh_9Eim6lY-BHG_%ix})e?Tv?os0qhobT7wfCjz)S6z9R@Db?3UcI-7|$m4^9JdjGn
zkkXu(vlZGeZ~dYNw~sNB5O>iT2q9QV7(g2}*#1?uaD@=K&8?sD3+^(^86wF0tN`C2
zv}FKUOfp3=f!S*J;<u;z@d{N-&QTEQKXCJ1xM<P-U;m8s-efmnX?5aXR?Txidhj|$
zQUfubONI{K$<FOz^1(8IM%h1p%RS|ewVhq`>C;0w!s7fW`%5IaG^){v;>@J4@-TO`
zwysVpD<dylY3CbQovz&u4y!I27>&rl99HR({hgL{5%oQ~yhS-*t<3qBf=ZS^m!!tk
z4sf?ze}Larh^q`mi0_D-Oa;n&{K?^s1f{I*XTVFouWO;t&vuur<$iV*PFj$n;58kF
znI|7FG&J<)fTt0$H()vT*VWB#lgn6!(I~v}tU1lH(ig?U?XcfBZ=Q|`4Ojt!=(;Co
z?Q=HVC;W+u+36fN47cj@jE5v=+jxU6`!!$xuJtW^8HC2x^z7n7I4$CZl(iIzyRA=C
zq4Xw#6icxe?lZ>c>*qGrM@T&ad6PMtKv-eT<tDSmeFEmN=edK>Mq^}n4jgQ_>}`32
z@5;+w6x50TY|$0Epg3~Nj^fr@#=qPO;!5+oR9jB;IfeEa<RM#TVu$%(?nUku?S6;(
z=~?vZ;ACn(>6c8ajW_CugS#58$ml@BigGDnO0Jko5hK<B6;=lV60<IEbb#h7&Tb=<
zM0YJJ{dY581RArF)=C}9wkbBq7mNtHG=IqIQ5NNW14p(#accdvYJ%E{>;dr+{?Ci|
zHV(-doV7=PQef<aBS(&8PrLEbphLTMyR+>Dli?olg-J?OdM+PC(8JpQUVCul{=h(=
zoH-ZbKK`}*0u}AI-yXJ#@A~=Y^EYP0+ojksXdMOi*SKcHIC}@JLLOlEjWsW$r=f)z
z9+NW?6rhK`YNegU+@`rJN8GM#F;A^-pFY@OChELMZl;}+xMKifjyy1o0*Ha67{+m&
z=M3uDzN;Fo+sIxi{YX7LL6YpXA2l~OCocVRO<@CW@z<8DN!vQ|cEvx{T^>y_@7ojH
z!>b_Ijm@5$+IO>vnR*QBS9I6sbLpu?><CU`QX5}w)I%Tzx{$9|mQ5d67UjeQUpgbI
zj~N3iO#JB>&D5B_&KX$I0;#vRDbwu*+oQB7Eslr?5MuP@<$=0m39W?@9r^-<3K;{M
z1C_=}I4x^(US(utOf>rXQ4qrPms3SsaN_Y<HfW89o{}75b>Q2n%||~fc7!-7;|44~
z#%-MU9sUdAgTbsb3}jJZ8Nd?GZT4~qe!!Lb;6zu?N}CNSPFE^luf=i#jc{FM!|C)p
zVMVW=Vro+O#`ElE#|iN6?Z2^@S^7KOEX7OwnvIF+cSLPGGBVLAvR(Dj=b+up7Q+g?
zXr`(j6dk)VGB$tYt~0f71ABNbZmVkG6xLRLa=z`NI4hIG9aNsx$xa{nt!!*+Y6?5M
zE@<ZFwzR#;b{RXiY>8N&>i*d_V$KeuhY2|u#6lfYrn?562fgZDy=~u_GiUT&CgKnv
z0s|49hyZBS=IF*D)FV+V)>bb$kFw?I2?JX)wL^Qaz#XZ7ro_{fa$asiS7dBpF#oUL
z8%On_9n7m4by%fm$D_V#e{^H(?LQaB)D23_S&?-9mc5UKg+-Cqg9w9RXCgB67;ak}
zXgos8VE3L~_2Y=5N14GU1yc$ejz&~B`S*Bu@P+`-BMh$jolEKXG`2|<+0gferwQL+
zll#0+(+4}ny|6KdpHNKiI_Un8oHn{Sovv@(v?;hrYGBj2w_Uq!Jj5WV794Fl?2{%s
z_@dgndtapvZ+VVv?gSC2z{{FMlQTQCwS!Ty*{w&k+V^RUq~(<6S`>WVMImzPG!?ts
z<%X2GXZwdG+UEBzYdYoh>e8E(n2BN-`r4-U^z1p4tMB;xpZT8P8E}5M^UTQDl{F>N
z-Y2c19L_j?q@`u}s={b>#hcSVjf2@I?Sg2K+4EZek*G{5!I31j9i`QjJ+iK8&9YB7
zZrHHD`jFGsjVw5pPJn%7a~mVZo^4!nkg3JZkqOs>Pc=S%{@l=d*oe1%g6bS{3!R#M
zqi%M4Y2!KKIh7t`lY;Q_-y7?^&VD&*3hV42bgVmgK4{-QN8fnwJHz64#P&`29_tzR
zqUiKH6{|W{T!LnmZTM>|uO19Qn7^0~of{Lcc7M;ocGxw)=ktRnB)RQbHq|w{@y?pl
zBQT(Xm|g6hMPafSoj!Gj;L^nX&KuURpS603H(tPjO<vXBzpaVR9Lt(;B9>jGop+Aw
zN(NfB`NwzA?Az37&&+x3oNp9DQBFVGFlX85;zjpIZ0|ebRtyz2QPiOGobTu6&LlgC
zqYPz~xb|Sd5cg9NpZDXRe0!3hspvM&J9r~MK4rvLe>jbuPt}<4=WLn$wKde&Z#=l{
zcdFl5+cD#0bHR#LK9KpPKBG^-sYZWV`*!Vyy}D&YaQyS3&BCbU<T9HduO<AwJ9f~?
z#<CIV)i>9VdFZ(I(=WfY57Mg64RAj+u=*eOoRcN?u?CJy<38)&U2jwR^y5VX<*@$a
zmGf*6oj_GF8%qEVR#pAx%>9cq`h&3n)w{?tyDrnh;0c1Dv}IID`?sE;(pdevVFZW?
zas&c&<E!(>ekzL4J5%-ARqv#Q=d)P%od$HUy@=KBJL7mLxXk;uwv0^uIq=RZ`jK^Z
zIS**y!IMaD`(D-F{mEu&%g_o>%da_+%=xU>2CL#nV0>erhTnr}T3tV~y7R$Kt5Zt4
zZZ}2brn4BjWalp)JuHd+S^boJ9k8{(x#!u=KcPS!(qnG^U7f|){W_SqU==`1>kAqb
z=zeTjq3bzeQsJD_CMxB@nj7VGQdXlsaCD-&h+CZBM;z^BX<v@f5qu39$(f?-m`>#*
z^*ieV{O2C-P|V&|`z`_d_IVY(qGkXN-tWU!mo?e2@wthd58IpHeuKISzKi{!0-`2;
z50ZzRvg+tjdm#8%4Ormt(L>`8bK2bCL9`v-ZLpM2PLUhWnZLSp`LDmc7CM$rlcueq
za{c=CRwkF}rSF_~|8;bA@%5K|HZ?l0K6|wlugTy&IANn7d#ZPnBThi(kL1Hhr+}~J
ztzQns5T?<llYEXrjsblZ^DZ<4Lme`zD=Sycvfe?xPv|XMGxK^$U1>?3k+JdX-`+%1
z_V;J2CG7U=ng;F|wCZ<4^-*`*960{r6I%1d&Tfk0mTpf_RkREGt{&v*BEYT6pnu9j
zxX|iQak_@EW)5A`scA&@q^Q_9v)L@K-~Ng+T2HY&kX31ofm+9T1!&xFPh^PWgBUrm
zEID>#d*U1rnV1Nn1JyEVd79t(i-%@I9rlmxytTD!z`QO}gNYcGa?H};gITmzpoi0>
zq=v~1#CX*7!-<oBMH{SfLlJUY*M<dMHi|+4R}s2r@x|pH&(LV<v*-A3oJYxZ%|148
zND{?r*1-*RSGVdnzj}3`RY3wa_51XD1;+Ly4`u6`N93${O39M@GtOZWl7{zry{1qL
zpR0{`t7pua6As@=G##X3Y(W&sIJX1YPSJ~FE8$YG!S)1%gafaL+J#UfWo}5jCqg$$
zJM-$FKEaz^tOfX!fF86(Xvf7iyW7)RZL}erug1o~?T%PjE-<f5cKU^e(%_a$oK?1V
zb57A9&W{+rq~95Ay<*n+{)f)iyHSz>AgmhOME^W2v`^<pW_ine8@sizQBF1mHL=8V
z4)`LdHV?lya){j0t!IaP+@XV_n`YVCEC}O5%J3*z<iCH=LA(*<)1hb#BmBv;7y7nM
z@Wj0XrgQuxU;sV}=U+Igva+GrE18E*R5Gx2#TyG`e}0N;6ka1itE%?97u6ZNfXRT<
z_*||xPcNr&mkB}VuO1p@#CRKrglA@F7mC@E$Q=#z;vHj4<dgJ+^A1}hfjC(<t+(-h
z?zW@j{Wb&DC4b{h9Pq0_vOlmd4XYR-apo<OvIn;l@J<xynnp8}$Fv{x*KH-IC{(5Q
zBCk$gx|XMA<V*v?wsaAM-gQ3JM^hlK;SupG91ii*KXKxO(Ydc<OO_Pg+|P;KXR@D;
zKTBmWc<?mF9GKm}vU-GbD7ELy82yDUy{ntGcqr^+7CJl%2o7GB^V>?h_J8h&QlNdQ
z!>06uwmzriagf#8^rqqkC6ZPs6I;8(;0;L`cGd5>@JMAwK#4Av@+VeU6BT_?P}f3I
zUoq3OeFLm(OFkn_@DVP}%M;IEu`hD|n^fwl#Cz(APuXH6mSYZO&p|MY5IsA6CEJt4
zzXvp=iSe9el61o|7;O^w(HDPD)JynL{gkmm;oo<k(~>GKQu#~25w>uw?<g4c(o{Pq
zXOBEm<v*yv@0dNuP>ATPeAudQZ~cyVD9cQuV2!PQy3I`G`XK;1+9Rg5u{YaILu^eT
z5OW4p>je~mf`gE%0*71%;?I7SU?%<spMNgkwEM5|5p8>Zz(jZT#P{%Wes;b?hYq6}
zCN0*W5Qw%BU(IrEh01=xvuV-U?<GKL5nPjqSyFBIpYC9Opu=Q%vGEgkweuqP%Y;Ud
z7r~WSA8|y)MnXlBnLW&;tn%K&zvD9tNyf0ub)As@MCS-vFTO@pX|lZ$Tg|iw#dT}4
zP)=fmUhGU4OC2S_S`a)|9-4D#jGfcE!RMwmv=!4zK@thoV!lPEDHO=2wQrxVb%-*6
zolG?$5uX>qwbZrfm65~7gLTgW@M);4D2a2e{Fe*Bx@I#oBZuI*A##jG-YVR23dF$W
ztKlCBJb53cp~2{0s%akre84poI|2_Vf~L*4FWTj3r~G&Du>3+xfRB$4>bRvF&*j{T
zlXi-4nEa&uNV(tvE9|CQ?4nx(0<v`W{KuwXAc>E5_NgBe7EFP|mNxcF8@C3gv-h11
zoyClv3H|lTyKgbs$|K-l4gIt$8rv{{_|CHm<7F!gp>D5^n3s8@iz>63o`0|1ip~RO
zyI%fn4Dh@Hu|g9UM^BfClZ&^q%%#w2&%7=mTgz9D9J8*imX?3|oe_0EL~ox%L?<7!
zRx^am7=GeJR&vBeP-<QViI<?_7=)l1GyA>dcZy5Onu$i`>RzKq-)5p723YqU^>L8r
z^7*Qu?2`GDjn1Du9U1H&)Uf54`(m}e*PU{iXi$`KCnqr=(L29BXK*|DAEf@O_dLeT
ze8gdNf^|Xj;Skrt|JzK%e;j@M&RgNCiaucJRS&G`9V+}k^19;*Rx9)a=>*sYE_S7v
zmnAaNc!-Cpcg!JE<mNp2jZZ<U&RdnL=}sx2CdsDjgg+I+t?H4nO;Vc}ad`(?n}ig`
zqhudBl>NEW1n^>r;J>}CSA^KtVMXRIJ61L(0i*pZVuZFD<f;e-R9f(qyn0h#6TkA0
zF2EiKfIRc)g7<>LD0fBI%XILUhct49vK^8|#u>VuZ!43>5ZMXl*GETRe<qf>*RdGL
z_5oxs0u_`UwuTm(L2j)wC;c6uOg8ax2X}gS_jwI%BoYJ^pQ()OcT>_t88P>%;7KkN
zlU00<Oa-};xiHO9yvCqg<vK)}9+x645P>b@FVh(mfU>*nyHTTVW$)39T(>UrWWkqT
zb1Y(0r#las<3?}=_h`Pg-}v9)V_l!nLMjwP*Tox?ct&pkrwpX5|KAhnLtwz7B4%~~
zSDWz_Ne#;Y_x}vD9dB?7CebSkSBW`<vS3jxygF(Y1YiG18H!7xa6HJ{j4pPM4uE4F
zQX6yTA_o-nbC~491?ajHrj#RLLLEv$Eo;276mWGnSLWs+Dax$-6haGO2$Tu$n5jzr
zWeu-nj1R<SzwzHt*6|Vb&xsd?Xw0^E(p(XRNnpprD=#aI8XZs8FfxW*`S0uZ6bgmE
zHM(cQp-zp*?Q-no&{`M&FEWDN016#hhn~=aev&2HxBl}E!gaU1#yj9cNtw7;hoCq$
z^=0<5vUg+yM_C?PA#>5=e7q@YvE~&1)%;WO8aI2DhBeU^0lVyNsd_d}%Oq4rr~mRm
z6vo7LvDuV8x(K&BM=6f;D#ybL-Qoc*Aj~uO=rCho-$p=@hA%d3y1emE;8e28Frbu<
zZzkw;`8(2$t3q}8R;W1O(K@DrxRUUu7+NwA<VQn4`H`^}JhX)_J1`-oe|`78RV#&2
zK4=Km7qqw&)V@uTbHi6Q-1-l-rJImTer1NKuv&GP8`K@z)ZNWpL(xd<D3^T`k?1<u
z@e&s;4m;)9ztj{jSx)=N6#_!o*^@UpgAf0uv>X=}j@Di#oN{){WK1?kvp_;a=vgNm
z40;H4C@Hq~bf%vBzwfw9c}H1I8!_<2aJV7=eoy1}@=-$oXU;RDt9r&M&AjNOT!M;r
zas_`C9v;ctRGP(q->pC>GR9dDOF{)OR|DzFHUz4&#CLbKVzb<wb+`-&Kb#@0TG3<H
z-z<FhXsM=N+C5k`8QD*QV6DQ=l0?W5Nz}H4KzUZd!DeBX!3=9=L)-vJ7f^ivRL&AE
zkqgfE&=4QJ264e$EDc6Xo;;b9!Cbn@w+sv>$|gb9B;rB%Q2QJ{0(JPGTxS|2TPSb8
zee)d9FG)zggPP<S5=818>oa50($c&~eyLmn$FDf+Y4w7>(earorYg)J3SJFK#YS8#
z2jRwy>Q&YrVis$0;(SU<O2F;+%9Di8#f{1)T;ms`I3{PSWi^qg*z`s#5;isu?7Pr5
ztsb`s*6SQYwmLyAn<T`#+s^9SE*lR4VAm4i2fr#alhqS4@rd9lW6{@65@SwI5jF1_
z3^B#9T#-sh)JjBOg%CgoP|wC=GBuVkO!;BkGT!U}(5^S`(k2Lpe+MDa_#zsD<4iV`
zJsD5Atm%~*3?7Sv-#qOrs@rxet@-w{ieha`-CQAQ(DhGH({fOFds1H1NF%2klYAJ=
z)BS{KJ6K*U1|)+b`oKly@Hx>ns6&Bs#DRGDf%?Oj8T#-uSvXibD!AQ3_$(FQRR|BZ
z!+jZAi-|e&w_9Z^sS1=lx1+mB&Z`&FI-o!l+Ga^gPa>amF0d77Z&zF}pmP_pDMMwo
zhtVu0T%K`U{$EO`UuVCJcrf{bNXtQ{GjOq#3+}%SA$KG$-A)Kc)Kdo44)=ZSCA87b
z!)6vC?!rnrABAEe!%+9PP+&cvfY#!>9Md$L=bH5mVQ0f`Qg@$#x%9u-s+c|rapp;r
zNqP6zGU$dmtml+PEokzhVx8=W)#*n!A!c?EXPS+6l(1Yxd@^L>1pz2^Jj>k<%c?yw
zPSW0bwq*vLKene)c+m#D^5w|XtxSDo8|bXuP}SsJxzHDnC$TAF2nN_IR(@>P8%R`|
zkM>Ep!IZNlk7g&i;*^X5_`>~7Ov=A{6|=S_Qv4}zcv?kSh}xWyf~=OHC)P2QGYMz;
zD3u&Gul=c5X~oqZ42)$PBlTAyuak93Z05yI1TJ#db0o6H)v_`oyHM4B5Tzq?N&v0$
zqpJ4AeP-4Z@@jPbD0$0mKJ!cD6A43(SyIx(R?|KNM-vv=ltFtR)>a?QNqz1)Z%Dgz
z*Led~zCFMHzO=lzS0EL@b}ZVISRk0NELqRqe?IOM!(j2AV`i>Kf99yr+sYH;GV2EF
z>gEH}iX}?~82Qbm%p`;nvV8etZIcbcA?B7VHLSs*$D9A}wddg5(^?07LW{sFn;sAD
z1<ykGkDfHbN2y%;v}$E{4VMy%F7=dZRkW{NR~$&HGx83%^2-orDP-^Ll`-HH$HZ@1
z=o~wc=$#kR=Efufw{+@tz03AY*%U8cAQE;XqZ6+!Dg7q!&C#_CE8*Fq<%c;H{H~I-
z`2wPO(>Cp7Pb%@-HIx?y>&nH>3BO7Q+b&UztfZVH>N!4irtS<{{ct2VarlI6wXkFa
z#i2!*L1zwXg0NsTrsQE%D-uFva`Zvuhud50vki90y3Uh3>3jWpd7XAL5dS^xWPDdl
zdR3esLE<gMiaNU<gNnsaNo(_tZcWI&<9c7#vGEVWEO1{d4xs$2QBy-3$u|K;$PdhU
ziMyz*y$q*z@N&L={c}}W1<qQ+VWJd<=5WR#x4UrJ6oY@Jh@_}`C}uaeeogHPF-9EW
zql@&EQb!eX3VDCUV;7dS^Cj-@v9y~G0IOjvGF(}6*$28s3?#uY9@+M`eU$h!W*up~
zq7$%0ln{cs2Qsgh<&C#Q0uSq1_-Oc#eG|Jvd7Cl2_`qq-@3(2Q7|Q`UrQ)H#YgaFs
z-0J8cQoUVXsrf*|Mv#;2-dSHf)KdhSs<f#1KQy(AN|mfE-cmk?7a|Ykzo=A?W{`6=
zKNX5zT>0;56DcS)|F56_f0a{%{;Qm_`>%59|GzM`iRS13@uj#{T&^T8M=KCsI3#*P
zwZg0d5MU;w$)rr~(IYd>4ea`t$cEs*;Zo}0HsY%)Um^zyLNT#nEZ0q!I`h#eNcQ|p
zYk&?e1(YZzt<c-h+DTIcjwLQ$hURQ>stBZ@;&%%o0UswH95G#KE2y}X`~(IpK|oie
z`XWE12a;u<o(>88iQNVf>!GXM4*pMu0Nd+n3X=||MyybL8b{$Ye4YO$Z^o2D;-+`a
zQU3y{_bd1CrxgZND!f3*tpRMFf#2XS<%J_YVTBiMP2n4e_<0yjsOj->cdCO{PhsZJ
z-VU*>-#KyJ;0q2C<2zc{32NG5yETu0Vv-VZ`4tp1DLcg`J~V&n*LFTl=Gq%?{31A}
zm$C*I4Y%d=>4zwe_LWLh1T}Qe$$0;R4{GIzD;r5EE(l{w&llnFUs12QNbzOhY5>^%
zO@-b}`#n7}1MjgPrkXVEPG~{;Fd3trCeh!HSs-+43<ZvYuRJBB1eDU+;t@MHs>j-Q
zpi10G0n%*jlwTKP2>Epgt{oN4m7oQ{UURSCMK&jnDWsqOGUStyi27Y-2Rnb{mnoL>
z_{LT}>$Sre7a+F_cmHB55%$wn&zKU11G)GYaIdfLqg>|?T=#=^xj$2SSqh)xRVNSy
z33Egp3V6@MlGCX-zc;ME=9#m(Jh)xsTda0E8|-Cd2)ORxl8QAJrkV+1LwI<)cGCo!
z?&2mQ$f{0wNJxm)D<MK_-to~dq<7gWC=-<^_#jFfRxw9&S-sAnAG%N3yi4J>!+2b&
zG#Ej6amn|~SPhLr2xq2rsevMK5cdr~2WeZlIgh1|7K}`d?b{(5hYSE2?(_$mPh~ex
z)(3#7>IfHC0NmAW<wU78-!AzFNFHEE-A&c=x>%iFxLFZExQ>4J#|v0_3?F2Vj+jp|
zAYZP0vWJ=wndZD6OJQ-APiR3$c6D$&&G(X}W&9*x!~KJuDT7SKH82hhiqalcjC%La
zTiAgmHajkZc>KY5l};_q=iNmK6cGVP+#VeqdU@W=l{PTVWKFqZFr>2ij)a8{|2|%Y
zf9gn6Qf%YC9kjEy<!^|f`SivL$rH98z#V{~n3D;&5M91^3>PR(q!3(3OQBJEBC@Ol
zx4=xNm;~EY=AMfmNc`*B48-ZsL~Vx0mGxbpQLi73lLZv4`c14kc0kDEW7dWIq-3HP
z1A+|`LQ}PjnbG@QtIUuUKZFqjkm=ZsNmHbgVUZf1b}<Myl9FkN{N`$MPemLx)G(;6
z?yGVA#9F}P>fcnt%#8A6D-s;I1F!D>^N%QH>gZpt(-ED)EK>~C=p({c1q)pmP+I&s
zCpz&3N4$G3?Uo`pRaO&<G)_Vg56bfC3##xJrKb6hjBDN#f>jf@s_s-q_nFMr_aJbh
zjhC$_`TtznYT#ZqiLX9EO=ud1_qU)v3Gq`u<r)0y-Aywf9E+S3`nWn&)Tv-<CmOxP
z(2q*pYT!PhZvdl3mAWKNPDwEsv}is(F5QpFL<MQrF$J**80C{F<gIFN?oW>B7r08r
ztYOI)PfGV+<ED;!`6G>de8>>$+b^LWk`=`FSiwO!fpY+EW;!}LLb)W0N|yo3Y&wJS
ze2{PRV*j{5c)>7DZXa`GFcB7n<7OEo{&P*ls%wL{HI8})swlq3ROyQzipE`3X+mS_
zY%`Iet9@=z6qja#nw-@HKRcklJFlr|Er@<cWQmRxGgZ$p?TLIByA5uQrn3Sq$(Q?y
zM82eKG%E^#i6?yg@yDLg)H**>S&GNJP7R(%uwr}zGkCJXOtZ6(@bap(wv33RNau;B
zvF23bAuJa+WbC>EF=<`DUR+ZtKcy*(7<lDNJXJ_qVy~wc-gnd}JE)nb#mR(xM4+1@
zBw2_awnInp>a>kq=kb^5uTXJ`mQqp*y)NqjXYp|pK<^v7_CPgygn;Qtd__66VzCLn
zLj~O32zo>_wOkk(eF)jmgpS-XKyW>h+;rjRNY5j#BzT}1-0DjYgEwQ6JP(Yi$|~<f
z=R?zAxhP#EIH2}3;*`w*s&W3_@eJoi*(oC!Dcm<%dj?D-ZkIB{Wd9T9QL`I75#R`y
zJdp^EHMD}=irw7@9ln~~kO7jqT~H+Xo`Wz@6RTruXGatQ(*KF(M=iVEfF~nw*|%@4
z@;mmcCdpqPgMqx5eAudG%<-QnyG$H2_-X$ayfYAq(^gYczjy^F#RI0Rp(O!2Sy@<5
zItihZ7XVsK6)uVrZh__(SV$)?G3?}~?%op=O$`*e8cKZ+3ASn0nkscnYc#=m$X2uJ
zQyD?ZLYr>7T$x+%YSgj{2z2IL1|8YE;Z4~>A4r@ZuZgx5ACQ7LH=zj?H)SRW8moDC
zL8|zPHznhH!T>c>Lm!A^tn*G2=4YrWHdKb*<XH5pddj}tl$K}KN;psc_}vRk$C6eP
z&Owwo#Y#B$>ZnO1uCbD#;41FcVa_CS93`B4wSD)``UG>dlmVq2YmppizM3H3{1}gc
z5h^-RQ`u0;@}&C$-T4`65}-<J2+UMXdBB92$2w62bQLw4n3Oh-y$<)$uMRmUDpkx>
zlq6-0GE-jL>rYN-sf}wrL0V)OO8N^D?-F_A-<gEMQSHz8*i<g5xaTUD&G_9-^V%m5
zJL8o-I^0h|grSnUM1R5yf2;XwkH+jUe)2mHub%9v@WX=SALY~cZC?AJe3e(n%9|OF
z=jIJxJTuY%^td)ZKAUVY_KU}U-J;|DwsgPK6#w<KSvJ$kKAS%6vrdLzOn-c3^%s}g
znCN};$?2HWnKMFm=WRdhF>%+bw-wn3_v)SLz4Y@JzR|jt`uq1gJj@EcdM`)+`MBSd
zuk!j&gi=gcT&Ao=Pq|yL$VGoD1%zre>k+5Q#&z`{dz`!&J>hO&eK(9NJi4b+psIWy
zpsPEMQ+jVx8Fiz}fGxjl9Z#R{Ix&n=sa~97gA%%YQzpOaP)4XfVT%&C;|Ra<=C^mh
zk^mYmQ`D-b^yE?w^cYXI{<_V_N}KkPO62gp{cW&LORy0yZ4b<=di=~0TC-jA3h%2(
zFQoi1&WAb_?8Jn9<PfKyFJH(x@t>dsiZOm-|H(AIFn;+04g4nXx~^_!#!iyr+;?B8
zl@4~{g1)0&(J>Wsd&VtHu&VA}NlqW!4xx+8l_Zyn-xG5xDFfi4rplTuF}OqqvvJ-F
zBbcvi^!_g9mHCM{4$D|@6IrMG9X34B6C{a9{TjIZvUW;`7&GSJV$WO;8Haxr>jRJi
zin@CqMv(>MnV`#<4e&!2pDA3JDQq+p{$)#ZlT73R@#jQ6UG|>pV8Db&Mc<|m_JGS#
z>{t%oHA%VnNe>ez$~>4;K44S&o$OCr@-S-(0o$I*GNXJ<Wvw3$r4=I6S&(XQynH#d
z5h(M_f{YGQMv+*|>GCP{k8vU`;%$K&Aq@$wS+WNZrDd=@QWcZQJ~z0Po~(8G4rMls
zi$tao;@WKMdZ7(p_6vr~$a{!-j%4JU_#v{q!z-`eC~yiC8`*h9s)}$;KrthP?BT;L
z82gELKA#JELe>FTIU3G2F_Zr=<}jPzMOE>fZ$8vNvvSYy<Qr2$2Dclh=O3Au8Rgs3
zWC<;1E-lU4{i~#K!bNyy)$((twlyngBP7O)Dj!xGG_|Qs>!d=jl?^FRQDZHpcA{IE
z0S`b%Wf+>>mt~`Tl-&8^NH6u)ih3)i!)MKU>(FosHJbd&JbuHbO{}jF#SfSq*NI>x
zUj5u^*`7fCie@h?Zv_<$PngA#p<iMi0`<IVps{gT@}5bbuKqkD=tnm%uMvUcS!?jy
z`^yG`w#$C^YsPl?cp-T)Rf~5!?O}nZW2@}?6?I@Z=G@LgB>qDip->dW%pJ{U1ZyVL
z`2g70o<51XGRGnCD3HH6#MPE38TRFrs2;5rPz>}lq;HVd50FoG>PP`1a#$ev6Zmia
zoosY+_{hL1Ezt4B*;djsIBH~GWyg*k3(>Se{u<ALm?hKulk>*$WzC(FpwLoj%P?HS
ziI~{ZR0iWW(x&;4NJQ!BELMM$gj6kd(vLLi)ahBNLDpVgUU7F@sVcm5ynI3y=ABIJ
zJ7`4FxaV}}iFg0_{@Ua;A#qdr8XliHV@5>TSzs*rlv3h#8@^;3TeXru7k{W}2RJoi
ziUFAZHV@;{W2lFqXA8uAU6=x5^*|L|mA2$og@;ibkGiaOs59>IEgO(Dz5o!(5d`;t
zVCsBoc0si5_k!7q@C5j}leTuvo)J9xB}dM3!aM0XblE6;7T7#OoPe_ghMFwmII1tY
zbh-J0VQI+!?ukb061kvk@|o4_qXTwKxyKHJl$XW%>q=%${risV^~dk4riIV!f5G~+
z#ZstFa6sA$t%B>R?9^grF~f6tD$EeK(w7|Erw~@^#JCsENaoU1N?*1wzm2TS^p~B@
zaxO7`oZ0`>>~2Jsh~)B!2Rqc_(JKq=M9#qg4xBAakQ3%vy65~)Wme|V*){>dJM&l(
zR4vi<u^{{~SI)vHejWdQaYI|F3-}sgXoLaqLi7^vAMW^Yoib!$!O!xGi<~dIXH{qI
z*Cd@*=@dmYkhQl{Q{_k@`Z86c&U^~5+ZpIWC{P$*97XNWZ^)2S*GyJZ;s^r~MABM{
zSkO|U(jR84|B?k9jFzWU-w0zQrniT)fZiGpT=pWKc-Fje-oC51KykS3`xbpmK!a|y
zsHpXnsZ)!=H9=>lFAk5V_ZIcD7`M1;!=J%APuAA<=`$YnUsVhCYBSc>9X@_G^^wkl
z{Jgv@pCBMz-0|cd_wAc8=_Z?Um@rP5ay)j{%XEe>Le^Y_uxu0Y%B+1^=Z+m0JX|}q
z=F`s~s$OUJ+f&h05MfJMfo1F*C%da!g&hlHJ(J2Si~O3-q_U}$t!qbS9l(9CfU9r7
z1TpsYZ|BTcCYhJ_AVdQN6w|#-94OnbC?Ca+XXdd<(@8}Wv)3{(c`61P-Vdmn|25{f
zh#$nXrzx$q)^4NCCGfbR(ud1DHcw|2cU>L=E;r&@%lNXZt|PvZJu#BuQ0$orYW7XY
zFsKlc7@2x$N$8N>X`u^lw~}EqUX`Xsd&`{uK6!`}j0~zb{CloKf{mUC7in)mfS}z)
z%~<sw{{r1$pQ+eb1kTtz7!FK_6bnG>{!w?`5-<f3cfJ~uto@E#RHH&;qPSoH7ky?s
z5d+>Ec*@K5r6_E-Y~DO%*DlEQReo+k8NEo5I8YOJQ_mkhL%0=tXOya}iP`^G)9%to
zyYGgXUfr%Pb2)c#%5gb|`}fZ|>d8c~BtLQC4(F*c&becJ-y|+u?#8c_WTxI5_)Zdt
z^<idWyp0}*HKE)^Ka~M2{}Hhrfr~*n|DnC_H*8qJ>q8gZ&3|B|!-^N*Ybim-qYv&$
z_WlPUG#)61&ll6aylg!g?alre>06;qnB`Qx4V>|@-3}-#!dQ}78-m;OzB7if#@2St
z<4awK%wJC!tMo8x0%WWKio3y(A&ih5Z{gIh&?Bku%^p%!i7G~{LD1P4-QA`y+b(#l
z#pQ<CL!#{P#y#a{ltNYBtOgaiy|5+89S}7|y(>?fbow&3p)qt~zz|+9>;E_v=2*xK
z6VtP$H)jET(5wov=$cNzM2$~7gVw&Wx>Lo3MnnQoGHoI{8}i(I`knFSBA(AW8BvIe
z0<zt{Mq539ks>p2E(mL(auF+4Vrnr%l_-pI0#BR%9w6)@j2&cI$tV6OfDAjP1(iNF
zSg)_IA33Z0#hDg|BQnn5C|~u>a{4m;U8aBwRUvMCKT2)Erm(l_$ollLRd_}QEbY3n
z=eg3QD-OCD1@a$=3L@6!J<Rr`JbHRMJ11xO>T$u})!2TGa%)X@Au2LuE_5=u+xQo9
zytzkx*w5#2@ol%M%a%`dE6>**LDx*KmfxZpp~s;GSp|R5ew)1ITvYS4E*t^<v0kt1
zJ?3}Wxax#ipQ#g59vVM4`S=2wtH!wd9!A+ZC#qkpw(}24&FJfD5>r^a*Xi+>xvCvO
zZ>Q(Z&D}8tphs?4cB8j^sgFM%GS{-QrD5-y_NyMQes-$b!tJ*Av*QlIU$WY)B$)p0
z=$qmlw~vd~u;p5`+lZaSJ;5-nlSYOFsz1I3peQn|xMzI~K{rISUUdf7gf^XRLO|G^
zF#ACaF)?blR#?pn-O?VUaOrMjE3u?jC>aN<X;N`!<rpYi&9VBy@ML~YDr4SjXZXhg
z2*W0i|LQ0bFA-_i0TSs~I{?-3;<(cZ%w*P*hYGH*$g}`I!(DqUL7=M^PSz*>k3X1N
zFM9vL!53bMtP|9`GN;=x$3GE<-)nlII+#MbrNZ+0`hyDnIL|g($$gPu7x%0f!NA>;
zUR7p`lEYbZ%g&7hQd7%bll66_ONNVOAJ3vCGzCu@i3r~3xQ&+W+>hI~ok_iabWtxG
zJ^zveard$tXA=}h`haEh?;B!#wGf4&cTGt_dm0T=gG}zmuKPf8b;*I@Gz>yhcQY6;
z;KZ!=62EmhXDxTa%q8^b(f#Luj#)w^O#`tg%%cY~6^_8}=CN%?X!C1?1uQoMV#Ynz
z#a855$;w{2TKaGP**)peA6<qVwwO9B8Vgsq9u2H;0MN2=?(v5~xyzzg3uua9IRI@=
z4!`GOT7A&{k6UXRCo?5Zp7q<gp=A1<I-<&*^_z$<)S(hA%3PiTcS>Lv8I?;COoN>|
ze3e=K{s?&T`l4)v?WkAxvQwmS-~J!cps5P>@carfMBrAB2SZ<wYNYzRnY1#78cy~_
zXtc5hP6ZRB@}snd?J;g>@g}U%vUKbhalSmjW60a31J<3B_U2|><R9NgnK%S`?iIp4
zIVYS<j@ji-OXm*OQ5Lb!-KaBkZK6r0&;gr~ac5s>WAgP*KjJzl<35vr&^fboJ1erA
z5~V4}jvV!lK0;43@6GJHCBghvr>bpJ9F8yi?c8a@1<QWlzJ1{H$J7(1Mi#xU2A{Pq
zc?0cslaAQCR#hy@XnZ|1fv`1H_Ea4gL6o5BNRYF-P6+Z8xY6k)NY(^3AJ=ByB>wA(
zN5O@$zDtW|KW?X*zNh1%ef#zeUG$ytKmYtQ;mLcCwy6D+pQQYjzw`L>!mBoyHzi(J
zz5A`uNAiwNjKi#6pgAcso;)TW&ji;ifTJAy;W8fG?1pc{Odw7lt-Zdtcx$mc#PCMX
z_&9rH8e)AR#%-g%lr2A6c2m#Kn>ypl9oK#S5&o|aU4)3;x$^=4KUbUmBr4VOhj%nw
zIIwf)ovp`rsP1U5`Rvx6e&5fW-rg;JaP8A%h!Xp%Lxb`<-SQZ^=#ME^-(TkH$^yYH
zE4-GzJ#3p_Z*g|Caq8#hmBIhK=!_)Sojq1%PEmIg7RH=@e<^*>hDLy1ox(Rmo}Gb>
zz4qdw49)&@^AW)lzf*aQV{kwh)Gj^j$v5Amh3i23Mtm`7a;vlhar(=jaKZ$c6)iZ&
zF!Hhg4U*JSjY7Ye0d(J*zn}jpC?yw3OipA$By6B%8`oaLz|WFSY=76}pynjH%D5+^
z#aDs+!JIvyY2L)5ih*;AV-I&d71VsPcily^w^5s|PQM91{>Fv8J101QS3D2Lbn57l
zVZqf9yH;H+j_iA?J|Qo*zj4zNXWb#PF@`sbowVLzOcJll=@1~1h<?=RnD?_~SFOU$
z`|17~(Ql5{P5s&a)du^ao)_&fNwGUo9dmrQ>D8O3%&&IxT8H)iTf?v46@O_}+Pewh
z{pT2#HU@~I`0a`6x6c+nu?%{<bZU!jZi<IdpuNe0+zQ8_w~gvovB^GG`8O{Ur(*IO
z#>2Lz@!jh&V02}DW8SWQhRe<RcugIC!#kp>UHFo<ULP!4&29#|ywv{Z@C<GRiO3QZ
zSUkHBbsXoh-`^&YbQ=xl_mgv$bC#aHHsu#e9&RDRBb&<E&KL82(YHsrNYfkEUeA9b
zuD#cpRrY`IQWo1&QH0YE$cI1#@nJA3MOl=HaGsu3_MGk=In}V;IA9`wcAgE2wOmBs
zoOPib?^mmN+VT0n=GP2c?c3eR<I1Fa$3jEz17IWaer)xmNIy(AmzAt8++}0zYs6zh
zrt_WamR`StW+iFXpqpahBAl5u4=H5r`-aq>vCeOJjks{uvO9SE&Unt)BdgVw%#Pi<
zId?GLka*X$cGLRxoC>d_ll?Qz`>(1A-&tq6<V!Snt#?w2x}O|6bwE(7T*PEq>z7<z
z7*)M#@6Cs079C;1klP{~8dzvl794bdgv#HTuF6#{qkkpQgVaYC&;Q8d&2zJG0Ts&Z
zV81+U8apiI225PjJlwP*_f(Y2%9_P+kOW28|FTodmOI)begWaAjP7enY>aSC5(0XP
zk64ZOudi8s+PJS3aq~pnQm(x!j()y9dkkjmKkYz8lRvWh;1Ni_k;KQsQNgU7DH3lB
zc<zfOAdsi`kG*MpehtQH<52$vGWr&go$xC-KFvNeP8s*^sMULQ5RDI*Q&K_imY}h@
zv{}hBX6D~<m9GU2k`h|v+*F*|i*pAK9r|Q<-Q&fGUqU_jz}IY<anio-?QgkLnkIvR
zC8c@!=_eODNO4cqtDjI+qpjVk)1p3E3pBzIvZDa?-^@*aB$keXOnAS%lD~mUYIXl3
z_EVDHmlcE5k6`r5l4kAjVN}_pTb}M(_FLnx5L@u*dh^mjxa+hN^Y#=q+Wax?MvQ4=
zb?%_&kI5INYT1~<#@@qOF%Cq{$TMSnxFB&@$&hGe7H_^Kt`G8y-YqLnj^Mhc-3B+S
z5$|!?PgOgkSkND#dt;KbcE-9H1dj;J7>-`fpMWfueoMAk3pZfa#d9e7z`Y8I-MzoA
zJI^*rEy4Ol(8lGd9ni041!STKWJ%2BAjRz!suuuctjn?LC-ZVE|Dx%S8-Cf0@{qm=
za;1+%l%72g&5F4WkQe=(aVy%NJ|%y=b<t8C-<7X-LDv-7@*xi0Q6UvDbh8BC!~9Sh
zw0<*>9kggYPF8uyRy;_$;=Qn)wRJ!lXIFhY9ZZPcNz|D#<0oh6)}(IE1@@W?XppDv
zaIh!!3mu)fPRUBR%yK)J6tbBK?X@j~$6}e+0M?rt4X-GjIn~5`adJ>Svv5m_^XJz_
zeYfLyXF~;+eIGZDFJ$6`Hez@WXVSMzYRd|-M=$dJ`Z9J8*e?hYfSwzQRGfE{lQ_uI
zpX!<)9||TYZh!1_{m?YS`#9aa1zAyKFNIL@q)3C;9P?z1e#YF(4dK?-PhzT1zeyZ$
zNo-!YYWnl)5>|bl!0#MC-c^<6JSFq}2Jqp-LWUHrHg1(>*=R-L#N$sCZKJkCKa>%j
z47sIA;yTWXFXJg#aBG=tL-V*Ff_wZ-te2MUU(^tPw*w2>V#`(t-<<C)W(xS;{2aat
zwp$V9PYfLL-AOLE6Y7|*ef{|71OwmRO!mhRt_+Wx^mPMB=@*y7M5h0V;{Vg$*@xwP
zzkU1)8ylN>nmi=al_-kHh{){0RU}doMIJ)tDTOS`Z1%&*E~=#ntEH%h(I{yg9?Qe%
z!K7rVR3cF#>VBQy>)LVb_s8$P?|<(9?j4R}9=fjUdwsv3&-ppe_xXOm&%Y|>-^f_k
z=%-9fYdX8q>2bG(7eGH)+2V9MDe>ac0D5;g7fq9QCrgy+R4hJU;5e~Am78OJc{hj=
ze2#_uW!xP`!jzfTQtjX??lb8grG%)kNP~P*3U;^4UfG9mCPhQ$XH*ZAC~=l0FWRkn
z(CuKhOZhMm4gSq5^c_{mmgw+zyhz~0vQ4^=SV)PQ6m}}*;rNJ|ew(GGQuTo{WKW#&
zw_2VWFAq+TKcIHNf<qA4g@kr(M0w9+UMUOqun^ru$1iOl67oqzm`YN<<n&qxuK)1a
zR?85KczsoEBfg3rzb)Sn@OzLnUDW=HNQyFXs^MZ*CDz7?#72h8>^s{je#=_#h@kWX
zoqk7cD1#_vd*<4nk==dS<H($E8L+tiSp1j%{=ed~`Nx#`or6qqgQu@Fm7c%gC*q<8
zHBBiAWXLK6y}Lg=v@7kLZezU`bbb04xh`qLAY4M|@bNoAT^ZLgby(Ckk3*`E5bwdh
zCFM^0K77M|ltDAnrWfti={KN$i>QDFj5|H$?LGqsUTr(sV0a<e(G}E>gW3USH`?~x
zca5y>a`h_r;X5o|Ut~QzAJhD7j5Zb&WfBazx)q<sCkw6&+U!bI8auaTD+1!JEuo>b
zFkDMkrHMa|zB8bV6gRS+_r9~6S8YB>nmL>(DS%u5RjXc36P8SH!TU}hcIYtn=!fe+
z0z0RQLyqt)pr6XOH#db^MY}yW@<n}M#pR&&G2wP=I@}htjdVxw0m99*%FAd%cMn;x
zot!q?;GrB&!N+LK6b3iH@$pOLKvHJDF~zpzQsOmy%vc7%1)GT^?mih=rGk|MvCy5E
z4@|;1(zHUS_Gok*4n}?CXkbMdC`uW5^i&ps$Hk;eg!ugBP_ys(uo_2NATo38YOiF9
zHyQfZ$@y|&VjCa<aDO?hyiW3<NT#9>wNyE`c*m<%tjuNS1Qpk{^t_+F4c9{8&cPjL
z&&q2fycSDmq9&ZsT*&A(2?Xi{jEclH41kK}4qdGj=K}@?OgySY1W@lAMPD=}y>^Rx
z0nI2?`uXZr;{QkbsZ~SEwx_M~sDJ)S-gatxJB2FNTa`|48PZLpS_Vr9`H7k`s3&If
zH8vd;=1Xz8G`gip94*sR;88qG3*nRR_15>(27uCHgSE@sP#X)$TZSpVSTU7FKI`wE
zT>aU$R{bz2yLn`6E`hqV<X`{n$5leIxJkUX(r{~(zHG;ifBk#Qt{>);Sb+@_fyh$!
zWcGI*Xi^1#s&5D0z;O~>tiMKbZ<z9Bx)121zd#?(mREqxrZV9m)UN)=!)_TF=zxw6
zAIDu%sm4rUswJx-@|L)Ti@;FV%<HCW=GB-O2i0r<`su9^h)Ov2uZBxg_PVY~7ZP(D
zo;F;8B~GCnYb3v;zi&eqaisqKnR%_PFy%DYXvNZjf~`?`^%MQ8ji*6Wg0AtgMvPj9
zVvlV61y8Sqf&SBvfd0da$&KzMAlrchL56g~+jLed&oA?$kj$h{q>SIe1S97nw<g4X
ztnStX$!!-u;t`CFUTuNiT<%<m{U`c=eu(L^3nX`S1xikK^`4-f^6qN=yMH0$;T2I1
z#jBy{NT&ACUzR~?K@z9_@6j0y;@GugNAWoxs(o+$Js(}enX{NgM;pt`6q&v?dOP=|
zRUh3walS|@UWW(ukGI5~jnuxq3B~({ExO)l&Jnl_6QLJrV?kZTo?7>6bN#EoL)L&3
z(<=w9;DsZfq`QEBU7Rdm>@(dLtHizJ2hf*!PBPI#wz=*Dw(CFOhxVfz?8RMD85SW9
zJ)z14{N4SuUm{b^NJ)jNBojQKfCva77KN<fTunQ%Kx9`ga0)dV=?3x^aFl#_d8O|A
z7w-~N0a$nw>Fo4d-PG19eIF4N#9%Tf%i^T-lJ(G#yi)3?(ORB-(2==UC-Y8F%{UqB
zZhf@CxrA0b@g=fu_tVY!5nK#Shl;P;THZjTu@6^*l#4^=UT7yrix$MOrGREt1uW=z
zGu@4H>QC2I!o0LPo*kim=IQ^(XENOifL}i^PIjMY26RMdVZhTLt-R@I6?T5kQ-@LX
z#cq;;_7N+`wZ$7+)3{8(IgHQHEra-WECaqsCE)HL<1_#@k+5*cMtGcn9;hf;@m2HX
z>j4_cv~ck_5leCzP>b=*B=fz5Q*nkad(IFNpxpSKzG!y#sSl+nJT%oZJeu9JA0+4c
z(gZK@ht(0TdAt^yq$e^)vH8lO<Nxfb5VT}=%n3}8N~YnYco-_TwDGh;CK8h{8`4$P
zzES%7+ffqoW1YWN3xOOu&+@RmZK0w0(B_gLQOr=f&qT5+5wOc1J6ZVH_q>)=Lt4^c
z``TEN?ka5U!*Z{<4<Z_PMk`#tHft!dazTKWIo*RA{z{yQlZJu+ZF=8G)P}BN_^$?(
zm(?M&HJ~gV>X}9U_bb{2jS!|368@*3PPKrB6_Th6v%U9G3o4Sse*JbR<6)2>kK0zT
z80Q}hp!T!EZUlXa!KNbaf!wyz`9$qs4Y0XTbbP#s(Uw|~%n5wCt-DR)icvhmFz+2U
z{7+?24VNMhq{dVOV@G=Y13t{s9PdWvDq3fuM2Yh`R3Ta!r#UZRDunuzj?-xM@?}=9
zaGk|OgY-xYJOBI~2wb$62T_rmb%9f4KVKK@+Drq>ki#y!0)?Nswe`)#)iE3`(mX2)
zEH`>Xey@$iSVY`&g>JX}G~1?9QVi}({v3jSzxKM7XHX@+$v`CPMWpR;oFej5JD}Q<
zo;cw~d`6$s0L-fJ4^Mj}&0_>aGEK>>aH}E9z+8Vp<56j)DKqZGyjF~NJqo=@>O}^R
zN2WX3|1ClYQvU&T#|~28uz@W~c?<hJCd&EJyBH<*V(3z3&uR72%<(iUNOn)zE1`mX
zEB|smx7d7cU4Yzbm9r<RvNP0?Iz6Ff+9PWv+!y!It;<%sK)+?rmp$i1Po^4G&Z6Rx
zfB!R+E~@w(Z7dFyGo*x40#}WZZpXK{Aby~T_7(Zn-)F5VXsZ(ULPDO>Z%w2=)cXJa
zR#!2uR03;es{U@YA!3V=$yCEBCQTjPo7?H%{44~CQooaGjpW(ghb}CZb5<1;stc-T
zQsIWLgVCvtg+D{BdsdDjRYPmtt6vvEI9XbB23G-YQuVooy!3Tz-NRhJ1lMmFE36hd
zgb>+uAFL0C4FicXlxb17C|AUsw7mrD8Y|sBX-JZZ)I2lgta(Te_}Pk2dGn@kb=Q@0
z#G|>fiwS<Io4j*=OJa4=A37JEfyy2v&Tq<7E;8CzcL!>AbI-&`R{JYoMv&<XOQ>HQ
z0q(T1lr@zs&=;S}w~p_u`&#K#Ml#~#<M?}PO5T$wktz*Np`$Bjn@~*VaGnnj(`7@&
zuE0-jY!J0ZB>G~h?VI#uRhkeluT=OuXvK@2Pg-jyv~U-F7T_hAA*{39==DUyT~Sdp
zO6$nBU8F$Oii0Dw!~;k2twUVimRlORrWw9z_i)XpW68w4*ka`g#93`DH?T^ahi|S0
zez36l<qqKw8{$hIn>qK3r`jvWXGlIyoZC3pz!0^uQletOIXA9qo{vwZ%pm1#RmMDa
zA2i!O=_35F4=(9$-@z_mMN%VPtq|A(P|bYxpVH3Zz1aEMSgNK<mSFPE;oS5V`ULBE
z8GP|uw9y7?WK<DxJr0gQxQ9=EvZv(SH0B0q64_E{yZhl`RQ{Bw->UmoZ^KO~8&yv8
zAr@r3Z;;u)H+2JLq+)<pn15dtV5z{kGcf<nc>gUM#vaE@Jfj(JZCnNFBra@X)d+8M
zXI80J4Im{VWl<BQoUOAUmsSXj)T(sMUy`)h2xp1)10nFGZD*Cu2Q*Vf>Lmov{OgX?
zF^khJb%}M3x!%l;QecpABLVg~ElTMp`S^Xzb#hbtT;{+C@zurmeR-fx3-O<nBNJkx
zm#-K2i+};A`3vE%hW23<wVCjH21Hbn>xz$H*oqB>IARJEjTxJ;jijAHaZpvm54Gs1
z+g<@o(2kvY@@@Vqr>CjTrJnA%{SP|M99N;z=S}jW&bk-z+cwU+9ma3}|A+tM-RSbD
zify+i_sWxgDkm>maDg!|x<p@F^QOcfw;_5?R~I$44#@-$Q#XZfuqS}Nt{?K}oHa$>
z-1vL2PPW1y=*x29oXKU`sw&4rS$n&kVQ<9_2*rDe#M$9?OA1f9%8U@Tj9ikEKGfuu
zRkOxMmnXs4(+4~rSF3)y=)`Y7*r(f@{MJ51pPuu31AVTo{IGs2krEW(D<>r|;<o+c
zTF?5frRtywy+L?&@k)gVhEXrkHF~cc`ZULsj0HyM<C1?h=}sBmk<WKvxTW&nfBevY
zT;bR2jQH93L!7cGdcB0eMeZ#MM65WZ&yHSrve4^tF37kPqR}{(yCmdI^{Ho++Qr$~
z*=ST-m3;ttlIT0&-rk@T6v`fqJaJ`XHsB4K$7WGi-4^voZ*HktUdzFMK(N<S*@y`O
z@}@6v9e&Sv^v;2sybS}6nyOT@r3Z%n%hU>MQM@9_nQ~wAyQncp53;IPMMUNUrAyGD
zboJWYRaI;!v_b@@5$~4ch$^Ah^^;FNIr%KGxwR^Dq>%bV%@k%i1kC|jsuOnun`$E`
zNYpjHboufhIE3s+m@j_B(!9v^6=+53Z&6p%=8#HdY5LO(e)-0n&a|aO7mh~V2@)RI
z9_)vmB8n8daAE4U-)p4WXD!y^D6w;Ky<>9uEK*xYsiG~G#1X*D0#sdU!~q{VCIl!K
z`qij+ilnT<ZgIu&yt?@7^L*i>7Pq7dlBwFFjTSy996^F49}WH@I)J_!NxL5SoxF0m
znQwl%)UN{ViD=YhJ~uEfSL;iRqr-X=t{>_vkNipRsg}2pRV`(Zz^K&dyeatyOsE|6
z!e&qs#2E-+Hv5!3w&Hbei|brso1W|L+ZfTHiQoqpk^VNAqs!m`t+?8RP0y#<cxJTK
z^rNxiCU2?M>?~aBPG`eZgD`HOP3VhYGLVgD*7e;<9L#JVcKE)3(AA$=gNkS?vej@e
z7-`CY$&|P@2CCz~SNQL?TODPy*v+@slV-_mf*U8ipIkRk!g#@{fJ3o!&DsifSq6gU
zsHNd=(K958lUW<f*_%6OmrAv4a67<mBN;F;Fd{2Xj69L~=2*-3avll3@|D!-@@&A1
zWCwb}VFnkU8@aYY+pNFR>m2xzCMUOd8}f7ZYtSK0vz>RpP_zD!J=C<2?rT_Tt3^#m
zsy+!Onfp=<W~U0aGd(e0UDV9<LL(**4Sf=iJvaEknb);jq2R{~iE>y}Qci7Cz5dqz
zwJJ@F0z|aLG-1qomqgA7Ba1y6*Tca)u(o(r+RLxotJYUGY#8d|;&PmE0Qi{g&yLW@
zht%LVxF2trFYMMncLwAn6N0}abJB#L^EYwll1b<;kL&kF@6^Z&GXr}E#E1psIf(R`
z)8GGO)?2NO8yo6#g3Kpf{$}jN&%b#~m6HMBD=4{k`{`w7W@c4FE)%H}_jM5hz_^Td
zwN33+A*-d)o;?@X*_J&j1~e=SPl3C{t{cAR%5FTuvQEZ;K(h@uk)8a^jL?a)*EhBw
zJ#A@{nMfD4YGFIx_5vs23HN$l)I!zVm8L5V52e{bnj=JNlJ;eB;rHJdt3poQ`gM`Z
z5Q9hEOwOPC`>V~}`{40k=eLS_$FBC0Nm$d_9$mYdX`q-ck97RX4gn4&0oB}#mT#+a
zy0cjr6E#8mD{8SR&(;t<_AQklCbNnsSFYfy`%-;`v=5u`cvV&2M06y>@28{*rNi?A
zvsL{frUNjm6M#;N-ZNA=SB+|!Y@{ss$~;;lrbdnJEHD3kWhRiLs9!`oB_v59N>|)8
z;tHTkGCS&Ol)DecNg2w__bF78r+W-l?Q`q4$vdwU|EswPfy(M#E%XC*_0x^B8VeR!
zOAZl~x$`dk@mXSvV8^hJ+{7$@vRAl{GH+JK+DcDeL;5CQ0H+T3CUi<mGQ%J%{DUNH
zftD@ESnYW6^O7Ri>*DRhPJ**Gm0;fQ*l5{>Yc66h0MBm#SaW>N>_s5Q@_=cCSSWZP
z_3TWab-}MDoG>K+f91w8*dKi=Gjt10D{iNp3Cq2^2aOhtXt%Pb^fP=xOiPn#EVMlh
zMHNm~ol~Ytl_G66`U!3o)~u%8*7F@sK6%qo`TTLT(2(d)udpVb=K}khFb4c_Q1DCr
zBZ3yQ_YwuPaE1KJ2@T@-k}vTO$}MNfVor04sSkdNZy6E;hdvL-`~BEA$P5Bu+7Mh7
z;t%^ldZm~RS}fTpvCGK|w!L6f!0-$CVd-+yjTM5NjKd<bN1UIN@aucQ4a$}1j$?x@
z-4$4uKmBhqgMS>-|8pAg-~I5J9n`!Gl%aP6g}%Wy+LlIUb3ss(=`yK^*BR}O?PMH-
zJO8~N*7N=aA}E7mQd0M@+#gp3qfqp^hrTd=_v~5H^AE}7lp)Vub+dZrsz2vTUy%HY
z(A2e0+X+ik)9#j2Qmy-7;&e0H(Z}tKGs^vKp(4_znz^{zy>u{jcy}es&H3?Tt<~|c
z0TdJ}r`AfgTijI!V4fQkWVC4(Mie^=m%a!vjLWHidcPs2h-1{zs9AQV=&+;=?Y-Q2
zqu}&59XoecpBrSJJ~3u=;Y{{qQWwh?9D*-vrZ<=|-lpmURm@S3mx=GH>|LKGAVd!F
zm~1_E+)+u&4;n=;B8khiQk$A$q06sDJjeNEDV}M0(ZLWTG&eE%44||=M9D6d5ZU2A
znc0+`IqwUGfY5+COrKr4DzXLz#EOoNHa{Ou$pe3Jny5*;+xW&}Sw_;~rucs)&kU@c
zl(9SLh?SFnkdvG1oz=E`#N1#g0ZClLIo2Vmiv7KGNtrfg&#9T@KeA|4I214uer{@S
zER8qHCTRtdD;y!A@a`40<q?!;KK3_h>(|sQ*Go*Lw}FJ<2Z#KDPi&U+ttS3BXsq;5
zDRd<6h_Vp&t67jGY+(|eT)Hc+bf0NMRWMi+>$*CtlQWaepLAN*(}X8DQEr`cTlviK
z*6MBmvSaSo(n}p2!A@Bh`I0PhSoxEJ)awnvRoN9K%Sa*z6J;<B|DHZ<4_b<)mY9bx
zP@9>#ch5(^-Hzh&{t0zAj%XW>D&G@7HQfVZ@j2lR*`(rV#_g`kavC#Y#Ge9b$6Xw&
z{R0xqyD=L-4h7|0A9$7*!(*JEM;3ZPPLPVVWLp>5_~v(;o~CRNPV3nd5Ula*DtouU
ze87OA3dbe?3AcH&v-9kBnB?u4H<P2S^hC+Fxeo?20(G2rFn6mHgXl6VwlgY~x@<Vy
z_Y;gK{rTN;x?3;SBKVbVo1}e&zf#Kxa7y92-@XPXM#iX8vATA@&R0aAy3hLb{Hjkp
ztJ}_kj3<qI0=MtoO?(9anq&)R_k6R&bskW)3`4i@yinm^i5f4LFX42%i2i*5zMHPY
z#{jL8i)U~{v-Sj|qN<c}=gZ$f$*kmQrPLfZ2`o1RlHaX}ltMz#F^U<Rm#}{E<|~DS
zkX33|x84{9r&zSm?z;v2D^f^OcmPtwDQ*yoBGfX?5|hz^$F{+}6}ue8y5;I#yBS~#
z@M6pArMXMm?4TtYy2Z_lPR%`^QFEMb{}LvFXn9vN5g>pA_@xz=JuBoWk%@@xsfg+K
zqAvOoJ5RCtvMX>?aiELu&S?ky&R3JkS2F)uz}CpRW?1CwP@I)L2YfpON%GkmVS&2O
zT=DIQW{f#)y<dJWV^8VwVIkQhpdp-7`6|d$7wirES>2y>i?;pB$bbmOcfaqy@VJL4
zkx8(wRId^)K5@eZBPYeAlR+_EE-XKKu?^)`t68f*V!dZ3Bn)#L3Ps;!;J~E2tMDeO
zn-a0H8DqIp%T&^@m61{*WWN@uXvaLnf`-&dT`pAno%|k3Ll5ou$FA8{eRKU-ADe-H
zICpJ4J1A(z0dKm^W$B`dx!Yh~a&>MXIa1`22~oq6?i^WjrRR9-4AeV*5i8OPFOW@H
z*!4A?)e`0rV(|W{No<wKEJ>>7m&K6lmA&RH+Ews5Wbr>zm(5yvvesc}^rOMHy##eQ
zXuDQaZM};pfI<VUlwPy=v+-&4%4>hX<#m4SC$RvbhPr~D_W+ygtUD}|cB$nUrWi^d
z^r22y>7W&UL3zvZx%C+}twHBp&@ihPfqmzC<VRgR3cMM!u0B$ofb)#>V-Iz4q@HGl
zb=BMy!I-zE9k~tq*z=Dzh}7Tp&SR>-TqxUq<wBB$SE-Ff91ym70rR`GqF547uF?}7
z$oXg%q~0`iDwCDHQ?CqJ00X|uHH*kc>5$1RW?q|f<RIN<?RL)$FPUI^wfkZ4w|JTz
z)D=EOP_FsK0mm=G`0V5;kx`XTbvb_8>h%UB`xs@YH3D=Z6s<!y<xBe#4NHo@=dxZ|
z*!w(dqDJ6InUYjK+3yXO|1^irojZ56RSo5pdD1%!DJpg-kiZ=~xADM%HDbT(e&=(_
zXC!AH_h0WE@i|c60|kz>&t#6z!R347^M4Z0H1|9Be@DbxUp}VYJShH_-X{6SRGH<}
zrORdNpMaTmByEXxGo9R)JZT#|n8QN$7<7z7VmM1j%w41{sG=x;#CbqO5%cq&l^q%I
z=X&agLvMQ0oB6WnMb3q!@685Vh`zUrWqHLq&kmF3<dyol8cq^Y^O9);!aL<%OR5;m
zf-(C}0diY7xqjv3mik@X#iyGG1-lhAo{eF8tCQ$KxFqSGtRoJ-Ldc&0MknrEBJH3r
z>*;H>te4}tLD`J~N5(kzeQ4owYnsDo<7G?Y`zm{@r;@j1t`Hqp<JPWL;ogR>8Ln{)
z`Z_tWvQUZ!Fw=A)Wrd&tM7e$%0qJC}(8}pN5^cazkZ-r@-yG`?6cPU3iLjbJ<YRc%
zcM3-S0J|XS1@UD@#Zj_sSi-JL(`^6kQ{jyG7R8VZw>KH<ir||R^jh@ZF~J#ugr=jk
zrOVz3@9U1;Y{%r_17ou&o}Y8rIA|%^rj{0^=Q93x?z4;qy-#gH^fwOjmQ9h{{D1oh
zOohdUO`}D*lVIu?-Z1>Ytpbq%l7~BVN;KUWN<*t`9XAOT5rQ)}DLc#sc+B;<M+v?0
zHyOA}S~U6FtkKVJjPnWHOrdq!#~{WYFdlak3&qg;IFw7jh?QxDQntk{*^~W`hv&Q<
z0@IH#T9da(Rmw_m0W^L28@vnBXSd{;37kY@g;(;~!T5BFTFGFg^F%zurE2JqA2@;!
zk~uzvOE5Iz3FNkfy3>IJ2+o1enwSrm0CX~0*x|P-cKILPHI|m<`sO#QNqfXybKvnN
zppooG5pRgS4aC6YC}4Vd5xmC;HO(C{0dT!HJ@+}QHu43DL;_M%XO462_UfImdLEKw
zj`5!JdR{0{`N0v&qOl*rUD;1FIa2}V@fe@+?WRpsIF}G~!COIyC9SDm>l?h$05w_`
z-^%3AHk`<U!~j+Ub&!V$9Vy&umZb(~&l_lEru*8lu)ux79GceCzZFQDa6ttPNf{16
z&Y9p~0^gzRSAx8dCtX$gX^eiFnF_l{CP8Mul#PJ|XYsFR693}T^FQnAsn&2@o+=HW
z(YZqh@5ei-t`-yGfbm1bc>8R<?^>8F_E&Ydu=#TF)kAqZ6~=4IxDiQv2=Kc{to*s>
z`(MOm5h)i!j4!n;zdCmWMjyiACD*a>2*!I)HX-#8?Soe=oB$HPyoEhn;nj#CfhTf`
zE|#|tU()6K<>s3w+vb97`)*P;x$|Z;3fOUfMjf6+Rh2-21Z`nFtT3dVH^j1)g|v`h
zFQ$bl=j=>Lcq)vW8@>Z*t&=g#iHDM>aKYRt@sgqIrx=aVi9@G<=-q;+Wk=G5vB1HO
z@J|v;cY+nvf8h<&MRjHk=29XH2LJe3B$c7~pU~c(hKJLT5<_4p@gOo99W)I~q~R6o
z+`Et|3;mc=>oBYYA!AWB=E4~U#TXz*KbTP1iTg?NH50GTnY#L+J4u|%3$y-FD+m2T
zQx#S!UBHHG?CWGPd!?iK3IfGCBe4GU+mxIyB<4s#wa=2<B#tj3XXW8M@MB6K2N^%#
zV5<hVR~kROZ%uIZT>}aX)y7|%lCFOA<=_WBL;7!!8U;MFwW4s(-1V0J|2oHqL8gk_
mTs7exU2W*n;{WoW4Z?o8=KI$5uU5$=X+CrMH2IU+U;h_Tq6|6!

literal 65510
zcmd43cT|*F*F9LaiV<mBF@V_u6%<sGq5>wgKoF3iL>0-BBxkw}v;olqL_va*LxD)n
zs4WI^C`vMlfPf@Pl4sw-?%(?6&spEB8P@x*?uM!-+<U@4`|NXXD=D7hT*|YQLZNWd
zPM=VrP!>8;DDy8Z=D=@seyi5SKR7PSpE`km;%EGLjtBm+#Om|~8wzEk7y17@pY}lm
z3S|q0cH*e2<J0d=PPVEe)3bxDGy8py{7L=&(Ty7~lC~|5*_pQGrK(c0W~7{Q<@}Rf
z8jLhG4@S$QbUDGx@o(McpPct_+m<cc-f#Qc)5~qi)@56%H>g+ijNCn}`lk7>Y%Qkx
z(n5S}u4OuOlm(BSx%Hu>$GvY--&!(QaUT28D3rtZnN16S{X2!{mgm}kf4_R3@UMS$
z-1!@MBl0t#v><ox-zcrm7CZjBNa6971HaAv+S6-W>%W)dI$5~v*M~RV`G0s}r<@#0
zSX%M>_f`%|D2M-kds+GX`N!1QgoHD7b#)eri>*REJt^Zqf9@0!scwp-ShRI^M&ma$
z6<ytsJ9q9hYOnMxdfCvsbj1pFb@fLdi;J6dR(Q%TUA*`~XlN*P*Nz>h`1trZ>(4qm
zI$k%sxZwW}ul)boTK<1}VNdyk!HV2#)^xwVx$K%`aG%g3wrPZ^4cR-)qiFhEU&Phz
zuc2G{M8t?n)5)=CzTuKic)0!4$IHo?S(5C}de(Za%%x~wX2z*GPjp3>T4~T#R8)qi
zhRS<-Gg|ER;ZNj;v38!Yr|jfsug${~JxRr%txnS=9r_a9zdt=RQ7$!kMo=TsR>SGn
zXV;L=e)#yYN_u9H%f{`;*AEllQ_6?yj6%Dfd2QM+>@id-k>)Y$?%=nW{f!D!tfH~s
z$-8U1zP!Clk9c)`Q4c=s-((ShAH&kX$eh0H;>J<E*@*x-Ik`qP0ruBq;*S?D3w!mY
zt!)I4Do3B1ni92aKJ4T&+tTR0=I+U}QsWWM!{4^j<P;Q`?>2jUf229*TbkS9dz_<q
znZV+)W|yRF+b$IYUQ0{MBa`1#UY_4gkKh(m-??WGqt0{z`%cECa7)qmvFD;w%~(P-
zm5Ud{`f~d7c``O0c6?!x(<@F>xp3i$)c99l!`PMVZ-k|7Z(UfiC0-**JiueTSWn=<
zfy~<8&F()h)m;^*RTX*j+VwSN@iY46HmvYce0t#7W{)Z5!NI|62R2wRe(mDc+@#zk
z6xBxo0nwKE<FZweYnD4SJ9Tnfu3EYB*`dD7<He8moO5!@dL*P1hu6!Em1oqB_BNzl
z`S1rHmHIf~ZBdcw==VAvQR4?%rgNdhc(LrjZ!sCAK@#y15hv)P_TSInymhO#%bx4-
zfak$G4nrNKA3uIHeaREwW-Cc9gij`^WD4p8?(K_>4ynqv>j_NbXTPCm(U#VQO2UfK
z=ge4%Vl0oDp-|R%`bJ6HiU^u7l@=KpInnPi+w)_>bm-Nh&Bh<-d>ziVwy9By!EuQ?
zIe`JwlP$Bg>E;cAX+m??yM-X9B#bqXJ9Ea5YS-VabNtrwSXM7tzQAJKL`_-wp?pvL
z2fU-8Lf>e--n3kCfG|s>{BU@CaX`?kl<KaUsqrB@yt85KT#yXghmEVjI%$=((|q|q
zTu!d(8BI=3&i;IhmrBdZ%9?;R>s-)HtE#FR{!uI()I%0xlgsyg7Z?G}zxCSZ%=~;O
z9cbtO=uvkXOV_u$>*|ks_Ty4`u6wSP2|l=Ch0QfF8jrNQ>-A9)^LkMmr;%=^0#5|8
z!5-~QrSb7`dc?}@vc`BQwe+FU(b~|N;T;PXEt02?_BF}JljKIBRD6L0#H4ILb${pf
z?Z@e&#uZ!ax~d|NPOg`7&AhCy&qup``}V}(yUjZ4ZSDEfqguT!1$y+=fBf;hV5(iX
z<9$<cnq`ZwO=p=HtHo_h$3GxI;)40yV|pWBgq7KU|8z-6!ge^3dSUmP=`U*w_OgOZ
zU#7i&eJpRVh__>+G0O&FM7_s6>dcep`EHYP^s?YX8nJwHSw}{mI9l1YYu8HJ=gsF3
zHP!RwS0GM6H5uz`l9QJ=tC@*Qx%_sifuW%sor7yr!Ux|SdV}KX@oHR0h96|jeAp5C
z{<BqPuK3(DX~1$Swu}U8_g-^%cMsfea3@0E?OMUlFUvIi3_g~W)ZF0~EE$OoagHmh
zT^=y`;jH`fO-Gzh)9JpxBTWu1S*ibCXKU-iR&j?Gw~Y0xEkC_D`u6Qxp@uO!pM1%*
z_hxC`>4}l}A#;X${3gfNc{Xl{a6<8OW>M%5v&dAqawWnd^D_rc#lj*ne`dlg;QQG3
zx)jX<_dM!{h6XL0j*{3hsiONE;|B(;XjO<x{tq6s(tqv0_HEC#o?2>8ubPt!cRA!Z
z4qmpAhJ!|vaMJJg>8iLwuq0DkFU3$Hyl2k+ti%v_EX1yTJq_QKqvI#lp}$$SS9t5z
zqgv@#A5zCBCc5D*R|oR@GTT--IXThj;^y@cuV25GWzLyOt0}-jBLeuh{K-BMmPVp>
zUCeae?93$MKRv>wKTqS*rAwU68eS_l+Eyz_(CA3mspbvo|MUA6uWu|_%E{?ZC5E))
zY{=CL_Z6F^5|AQj^uPZ4%XW5pOd>nI!6K(ryZObX<Kbg-A*of4+`P;pL!<lcB+QZJ
zvKjv0pF4CvOG>|JIm{a0GE2%+-+0L0&vuuLB8_G;EHxK&wQs^-?plpCISB6Et3)5h
zMt0T2o*T)YMi8+5`hF!XKvPh4`90n@rfJ=^T9#rg35&*W^y#^k%lr*FWzE)|J5N_f
zDTuJv3hQcjHmpJBE0t&wWf_%+?9tQH3t+16Yn_Hiw_6?1PPb$m8#_!@TyDh5bnI-#
z3wC*Jc8{s<B|-Dvz2&qj&w29&7QRfZU}#BjZaRFP?)UKFGvv)|+u1o|<s=UDcJb1s
zRoi_IUrsC-tv3@uY8Vms;@VsbL#5J4s*#>fO;BeQq$xZ-P^mpbf=WJtrxc3kEn@7?
zr{Hr+<snD-XtA-e&OM9_(?)S=X}zJ&atST|Mw{{@-7N(kzBP8mdotb^7t7OA-6rk*
z{rp5fOwT<<%_8#tZ}F#s+39|R<nvS-a!aE9<&-a9E>92oc&KV>22x{B`D}9jp8PH_
zA@RY3-@SRn<mjqON?~|wu{yW8ZK&ZU7czSPU~AwJm-MWBtIx;WM(d2&){~kldStB8
zM$+}?3+e|JO9R=!x5>h?Dw|&+@QRRSb8Z%`|5scU%5Bo2P*gkrT(qg;_NpDJCe{21
zg?S^jTAF#zX_B_z!rI>X1m!s)g09(pK1?^yS<Fd6qS-NgYI<4?kAzHk6ybO2(xs+9
zCg-9|%#!Vn^HNV^HAyb9d$-ZyaR9PNSH-jAhMBw@HYAW=^X$8PBBkbQTU+=({i4%!
zBr9UvhFN-ZiyinofJE|l`<xKB(F@b#<sRw<?$ag3Zf<TPDWTJ61l1ENMp$|+31V~m
z)DdJEIFdiNFhZy9a!~d(BYBCMR8bMh;$FN^3O?lUh}j*|cJ7TAmj*zxkalJmj6>ka
z5gnMgt^A(3-MNQ2wC~vQ<IFTOjf|T)^ebtaV(wFpHo?ksz`igT@g^+f-(|d@7+vkn
zoO|N*djRp>v1dc2x|{QIsJr&<Q;t3xYHxFr{pT<SlA@<GFo{3ERzs?ZFK_(Z)0(eG
zB_UdPW67$Lc9P9=a~uD=A}I@a!sG*lc*L(hrtUg)NXuoc&+gwV?jcv?2t4fcGAq}*
zO;$pE&cW&h5Som`u&VFN%g@qNUB)chJ36wnwy~e1q81A)Gk<(?((L=EWO_tZm5O<D
zZe1xi`)NGWNr}Vr8398ifW^?-5nbI?WXk@cj22P{#e46S_rJgW=bzK`KUS@>8~&y|
zwq_1k)vqRZU2ym9b6~5$9zFN*$p7Aule}S7x%;R{U9UZ@3OU@**SD?X&adkpCD$b>
z(7LUNmMvOW4{U<Ug#A}zL9BP-!oIvQU7PW{!ud)-Xa8MyA+DP#(~|FIH^S7Ar;{y(
z31$8Jrfcw~o_AF@yq)Zd4*iI%mCQ4@1ap6AL{z<V|NbK?Al5fvK%0Npl>k1<+GI_J
z^S5VSB_lpk6Q{{G|5~Bi`&c1u<Efv6h{ts5+jsA(@6J_)qPwsJR$GOIg^x@Pyy=)Y
z)SnxU<&{vMTOmym^5^6wtN3H|3f$=t@KLv!@lI4}?59dqM?`MDj7;kzJ(HO=Q+aCh
zu04B7|GO=*g}_3%keEf|0h)@nb@Cb^?e{m#*k3MeCuVv3@JP)$)N+s5$pAR(KT}CV
ze)Z)4gNG+>N2qI+wnv+{gzb;7wErHBmu%iKV1cSV8I9X$2#d^Iry)Oo|MreJ_6=mr
z$Sx_oN4Co93^@n*N5#`Alph&;3eRuUUeErNF(>;mU|c<s0lF%uPybCNnZX%h$v-Hl
zr{g30Q=ZXSU{Ax}SFH-y&~-YQpPx^U2tVc1ol!8&0DAlPVgpyOAQfjgR&0nNADfw(
zp-24l&xz6}BBvTMEFEm->Y2iGsDV9eOS)`LyK7=OvwL3}6kkcHqtXgyMw0vk1LLrL
z@0Z`wO!4NGFs*!k()r8n9kzIV^&D@2%%=x!*w_n^SfkMhCGaNaZ(C7`E#%-Z9sKf^
z**65NOez9h_i=*@yC2C)uW7yZ<`!plkY#>s%T$L%w$<mqe*E}BsudEXeGdQld)u~c
zjXHDAHn8PhtNG>j_IA#RSc}c(H><l;f%2a!8hJOkPWEO<+H~AAEPHw(db%`JTF=11
zAWbR&K~u!2Ox4vj*S^2GsdeH{!LEjkB&nI7`kcJtCAnAM-QO5^&?X82boLvto<+h6
z<BcdhP)w<wKOX~gL-2M8SU}km^5(OWy?uJw(}UrenFlsXyKC=Ie5!Qp*k4#;eDqu2
zAM+?OV(VTt+}OS5u6mNLPWul!{g3IXPU!%u+i>M6643jq4hO#ZT+v`zz!1{O-YG6V
zdZ)p%AzgtaOcZK~YVqpKZ1Yelx7C|AC1x}`%LDgbSuGrR1_5@>9?b}Y;zwNo+akwf
znDH8BrGa83j#=c78+xzXcMO2BPiU>NGM>IHR66PF*DJvXZDaue)SR8O_wV0tJNV^f
z``52AnVFd%dr=XtM&fw!Vz1q|ilYeT>Z+=Kh!atdAFmrKSx+(U1*ToSc5U>}pN^m`
zx=Nmi0L`9LSFh#f`Ww(aR{!1o?#8T?ZQHjqfe%ieIwi=@FDoS_RXPm!lR3E{*FtHZ
z=<tsp*Yx>92(YMldv~pvS#5$K#|wZr%}lFj?C0i{=({s7>Fwpp>9Lli06q5$i&ku?
zP}9(9|5){>T!MOhUsmTP@AV?GmIc%4>I_XjuQyx0-aI?vGHMLS@UreIkECr(Vpa!_
zj@t;cyFN`2NjGXWz|U}3wWM9QA1YwGiD45&q!|1hofO!;JHh_i@pxw3w{ONMN95fm
zN7cxBZ``=Y(o#y`Ss#oqZiC|ikC|Mj)@Ly>r;%srR20C#p8~*Gz4dA5933-ZlQog@
zYfqj$xoiJ^`caD2ZCFaMq@5hhsk$=!lpS(`tilgXm=IiykDp&XyU{c)4k4twGTaNL
z_p^{AE&}4>>RY#N^>)LGwJz0UW<0u+XFBxb$MAS(Xb+Z5Nmlk&h{ueRjvZ6Yf8vWf
zf*Z51vsk#GhV3o0Gh;jluU$!-6c!Mu0^DLJChRY8bM{c#Au-!;75r}Zo;-PCm*0|o
zcr!Pi!n7&du0F#u^ySN!2qQa)hr!Cs()9L%s<InsJ=*&z`S>EPO*J={ttI9WpO`J@
zyW^}nY8f-v4e&cs0n{exC840&B_JRtJ>7Ly?9%HSf>MiQXB*A15qBDpHI&4I6o}`-
z=n^yYs6!6<r~rP`6}@|Z<Du>+Mj@5=){CUNj$fM|E0|SNRJ<P{ulEt&X`arxgp*Tk
zc4nFxeReb1og>b}TPQN|tyTy85P5FgSU{@xOMP`IMm=epgG!?S<rXej@c!MqQwZ~T
zJOLpgbyO@uy1D0jzlp+7s-w^Lz!d5;t;4AwEHdKZ4c*nzKF*`xV*$LdcPi?F8x!lZ
z$?hGrDRM1cw`f!M#D(JoKB=mzMtuK%m2bxmjT{AI<CtuR{zP~ODZiU?95ga4nhsx1
z%|Zx5@x64}GBs^&?O3%tZ!bSAH3T*cNJ>hAldT(W-L}H(IPS|t<!AThpCv#?XRNKQ
zJN~>`|Ebp=hN6tJgd`a-gt>WnrH>@wAGm$h_)w>RaIj9V+m0PO1P>m(fHG*GFxRqW
z=gyqDM?5tqW+#}E3$QvBLMQ4K<32D*Sm|CpeH45GyLRo<dbz&08E%xDn+xP+kW%-x
zw|7rYlMr$$*cROAi8MFYbB(K^8tUp%SOkB6|0l&*NdK9T^SjsJ?;qa1d*^DtXu*P=
zTecK#pD7oAg!lzl#s?#@IEW`ZL23f8ek?Cf8qOj23_aU6&gipDVVe45F)}hT$jvhv
z$j#f@wr}3N8R6qYOH0etlBG*kl$0L)Hh+F;J4n=k+=2re?#qp=UAs0YO<pBhF&HM;
z_-%5Y=PgsUc}atd9KlvNPmiz!bu<RKiu@<)BsbhQbt&YTY{{~T)<po2n>_3|5Vm5|
z;h3)K=%C)U+Kl`$Emr57F5B+BvF7;w*<!nP(eR_0@0!(79-0InRL`CXc4)2)6f+ah
zv7ec8GOtgi1&N!-4QA~W5~{WIm~4<cHQn~*3kiit7&Sa}BLf2hMXz&kW`EgradNb`
zdaT7G34ET<kC73EUBeyVcxH;x+28}#$FMFpjEu!~vTZ9Z-G_KfLyl-6)v`47+#-7$
zP%1hEJb4m}e>1~nIt*Y39RX1!F>KrRylSoJ>z%0!P(<P)BrB4bK0Itoa6d|MIT}qF
zIV$+K#k8f%m#g}0I<lIVH=YFVw3cN3QwtX^WJ-Q)`fRmRP*9$vfNR%YDuzm#k3Knb
z_H0tBIgNJv{)U6kkzO7L07H`SEh(w_@ZrNd_wI?#mZMx1d7aO_3+{v%R8?I~5`RsN
zYGg#jPAnlV$Z*gz^6>B=TWOIYpS}8zwX0XhgoT9_HC!<?6c7<P2VgBGoTAA9i0~!W
z3T7(~z?=Lhsex8(lzh)(WzclJV70{H%gPd0e<9yJbf_VB3S_5Y*tyWo_r}luKpcXf
z6B-A>S7ervxncF{ny6E(L&1u$)GqK^;9H-9n)Q47G$A5_8Z70OUB5<*8+;D&Rd5~W
z5aDaWPjLZf#~k-sVK-BvghC6j;PIlbU%q^?{qoOkQ?8FXnN~;F3Tej-ONuX8xNz5j
z1Lx<>o7XYRwf&w*zAq|HB#`&ZZN0-Pe1D<=K&AdOWdf)ixw&m*;~#%ieUZxHUb|NC
z(4m~F+3}K+GecvhQ(wP+P5tCPhQv%d00jjFu5NCnzc(+QiJW!G1929Ynp$&M#vr#3
zJ87E7g@9Ht-D?R}Qpx5pl@jw@&3!+ycQJ603ZYUuczv63uSSpQ9=*T=*3Y99f}&bm
zFXB3;FdvdprKSdsqkL1wa|`8_7d7_#hxj1p8iR%bwypxaVdl9w!12LonKq=GJNKrs
zhP|Yxj^14>oVeOg_QZ*A4h56J1R;inZMncuCj@%ytLx5PkpaWtguAnQ63?*`bp)GQ
z;*;{@kl9E80?C~K7|?DjX$P)idL+q%k>>sS*>Nu_4Ygs?NR|Nd6H1L3d?#CRyPL2>
zlN}4DrFiQ8hFH-%Ig&^?s-~u<X{?hcPSnJneGzR;+bpgnlGBukn#;K4@jgVAN`Uwn
zRO=@F&3Tf}!^Zwx<wc8nrYFZV>H?UiVg-BGg|FMaM&5sHUUapIu6mP;;oJGTOdZSQ
z^QOV;R*vW|SX&g~+h;Z2pYo-DTCTt>x_TGV!(V^>X0za|y}f<j=DX<ifc0`7>(l9q
z2DsszX|PDs!~#UUc=19&@AZBW5%PK)rCiSww1(F!sT{rqgcp3+Nd-Jn!b^YzOHjsv
zdKKOUPXcCJ;(D|4=;@;#rKV{gP16FmM=UQIQfjhMZs~mgDLpk6$koZv@q*TX=0-$g
zcwo??O>h*$TS?EJJu`gal7(;#2&=ZV`(DNFty`ac)R&Mot`*I0LbfBp!^-Lf_$Yye
z`F3s=_a2%OA1W``K=uQI2^>z-&|2o-y<outRUSrXXQxJQ=j!l|z?o1rSH<p^3i>r8
zZX-V?Q_k-e5EMLz47@&3XV*Ur81)InwJL16m{qF`=Y|7Hs1m##sR1j_uj~K7jk34<
zpWCY>-KR2<$)kV_2+mT|eaD?qURr9dx+f$oGxK~`SC`rfYAto|-ij&d4Zr`sMNh9q
zAjy*Ki+$wS>+Jkn#l^+`p`l5+xrcGZ+E3NZ)$96&6=Ja_E=bHn4rkj^Lcp#hzdGQV
z9q$<BS!!rg9Ok3Gen1yUDDaSjh3}DPC~#^K2TSd?->WO|C`h+xio>tOcdA~En`HEk
z?}W_}m<>D5oBC1gpYQq;6*`Jcv=KB@jUSTF54ODvVzt;en5&I;_V*{D`_p~aW5Rc4
zz)RW~HHq`Ut3{<fDKB59+V!Zz8KzMpB#27dSA-pP{(N(-s6$^PPw4ekI}}gTv2W^a
z;#HAKIY5kO4-cMTG%%?MsB~N0QII1SiL07`+g>|Z78AbkDU3(+dAgPJ1$q692C>Wk
zEG7WHDv~P(G?kc#V#wjxfByN0r{DmJ4-pE7=1b752wtk)O`lSX%4NNNA6bdglh-2i
zs6?aHb-<^`<tR%?&G-J{c0*%plv<{idT2o=2@ect10n3!i_dc&QAyIx1JZ2@EARzE
z0}xRKTorR1Xq96jpYK9shaszKYd=*CmW)J^6B!j%yRH_guLq$6L00?vCM|u0RU1`9
zP}~PEmo+prEEgr!lDs_NPXgQt5&22@^&TGEubrKO($X#FJ}Hr?eD=HC9_w$>BmN3l
zmEu;p9IWAsOK(K4e%d?q5|)k{V;87ikohFDnVQCtr1Rv-g?1TDb41N{>`!5#jLnIt
zm7Wt56M^CTy&pa)`g2R0s&jbWF%y&6I{v}~J(Z6a9i1pR%30J?Yus}&rLIJri@QN`
z-6C=CY2I)t@A*`aqpI+B?@*SI(NyTbYLv<yud1=Q#BUb`9y@ldq$L1248W~q&9WjJ
zv#ET670335oNF7?PqhUi5I<$R7TtiQ5#HBri)<`X|9=1BgR-t}Gxw>_x@D$m>O9kC
zHI})xSH*+u`x;}Xrd(+Oy22wzCR28C+?r=4?4Uqz7MR~`QXM7M<@Fe?yZHKI5(ODr
z8MOxjt98b-8iuS}3+JI?ZgT1j7Q4SY3yf)+apksNP+%&BTD)z>GiX8xEX+ln^ni+B
zpIk5PzHi^Y?z)s{q_ViCepDZPI^7pEePL=Q;3)$|j1;f$u>1S(zaQA{YF()Lhjiy3
z*)J%RIW2&+G|7>GjfVZgZFg6-#sa;R&OApWsN<~7tsG|Z`Q6=Bk$1;WR{gw|W!WN?
zQM+(#>5?V;^ulH1A4GNE?dk1hATbKeBPbX#QpepR@po?S$chRjn=6ch0*}b*XL}Hr
zJte^D9z1w3+i~zYO6ck^+G^_$?fv0<E?rYc-m_@RR_j*o>FJ5bF5kRygTNLO6O*+2
zt4GGhGyuNq%AQ@$G}~|@W?i=0_h_wsRMBHPZPShC#b-R!HVm2$_c7KDIJ>w6%}ZK&
zhUcCq%XrtNdw)s%R+Nyin*jT^dU@^JNWjm-!?`zYvbdEaTWFT5x@7rs(H5@n_AcRi
zzC{-}WQ$j{If=X9vtliF@g88UTj&uvwhlo(&5Tv%nzuelY<r_v{b5HJ7eVkyQ}>rq
z`f9M5z-yviGslk~kIy?y==;F$qK|GhTbh@9S>NeewAuBdqOSeHw(E*^t30VYcJv?#
zR*_tW3hK%9D(?m`CQ((ROO`C5s~X&pJ4rI;zE&{{z3PXRa0`-6)j$qeWVJuy>2H&}
z$1#pJphn|pRFFkn)vedR^YUu-pF6DeGW))yr&+MHhi;3<Y##b&@o+WPf%ROivGBfB
zqw-C?Fp#QtQs^TH)S$i^Qf);iOX5P#y6EayU_10D-(Rh?LgA`z?JS_(2?$mXTn#%m
z(KSRS+_+IEQo<J5Kk3uj`Ue34Mx#F!PM_X8d3*;ye>Ou+{`fkZ1BFLcAZP5l`eSfV
z8D*Rpx^*cE$IqWX?_2u-NEgI+&(Fa&N$W2+<aOU^bZc;q?LxJ(=yT1@x0m;EP1G1I
zNLm>Z0CIoaf&TL4euFDlYRtZzX|uO;3+7pL2pGn=l9KvckxL-eW{di9QA1Dp_~nj4
z&Q&)94L@J)4sYYkYpD<6(&uQ*9(Kx|;wrC-luxtms&Z~BI>3xcKc`{JYiIUU{am&F
z#LqS#JMbaki3*epxb$c34!)KG%<|rND7~nsXYy4T-%y0`;}9V!DP5EUgkmkJrKT-e
zwoJq*N}S)dyW-jPgY|q!alP9el_b{OeLM9-2l2M|Onb;bUA6)+cElP{ms3&E(LF$@
z9sj6xmIcR;76)2SXSBGR=^CE7Q3pf{=WAMC1*SqZX{8?l1WDD%s)5tn%uEZE9{TwB
z98A81C1;{ox4HZV?OlmFs)o6;jEtwNt81WLswqO>xU(4lg+l}apF3yQQL-LUTYcQM
z&FX+Hm=6MmKh6-g2S^4m8dPGCTirccE$AKebCi<?fK3FpZF~Ln3IitTuX`QXB5ANB
zVFH|`*en<Cku>iQkAy;vdtndWD?X!vP<N>Lv18nq6J|y2vb>8mj+8wrIdD=tdv!=u
zR1}e0beM_o@x6UI@cl=5BFfPE0uPT!sc9r3c0)<wCBaV41(rb6s3l^nsAlhcHnULd
z@(!=_t{xRsd}(;Y8Q)`kxfoPxc+S|^IFdGB*uz`ns#$B|s4)tnn>TOz%f2x*FzCYF
z{P{14JnE68#U)#|m8^5vE8D`eJx@6F3ggSp?-LwJZWF_?LoQ*c$FZ2G(#P*FO#N-m
z-AKgJmPdOzerqny{Nm?R8<F1m=I4u_!_{ZaYGRb_nvLY-?w}7Q01VnPGn&!ON(v=q
z>h>h$x9qn2hfH$|>Ye47wO+**87$|T+MfdxJ$kd*AMJ|z#e{{+6MEYN^wiIsIpc5c
zJ)INpu(Uo3NRgpepiBBJ02-!^iCP&JY${@;I}$D2osyFE8$NKQio%o*i={-RY$S|K
zU*kcrHi0RHsKN#L_#!{Hd?lD~d(FK`!jym&tWPr|0;C~7t8gPdh>*bS*v?FiOP=l4
zA5Qt+l#{W3_o)*n_I6znG(<i;=Pev6BbU9}F$#qoFlE&hW3drkq20S{po58f%2d^m
z_2~Ze={(SAJPJ5CoN0RNJHwPGug>YL^JT7Il?p%{7}|tq1KcmIZZ;PWs$=a(*gJak
zD4gEp%c-w*FcHvWtD(f0nV!_hE`vLuDTH#q3JgeT;mD8jCRv}3q=K3H+>(|SJyLZ3
z{kIa{<*)14k>`aGaRGZsF$MhK$_1`30brnKDsR`}!`hXVl_*hF;S=1r0jkItFn9=Y
zan+QKZK<x|{c-!^GG4r>nA&fis=HU$tZG)F_Iz@v#arF%=H})Oaqb;^;@ACLzb9EB
z-F9zZ_~jN|LWT6&<A11SpG1%HE}|Z8&p|wTh;%`cVXd7Q1Z;R?kvF>GEmoj_>m*-&
zd!nPNdWB=OX!6*F3l}0^$#OSbQN8MaX)6ET#Tw%g<>Z^M+kJ(3ii_-+3hEjSZGhP#
zdNxHqmHdn^W(wIgAyL@{>nRo5ng)eh8n7wm{W}V0&e$Fo>#GEHK&t{X(ak-uNz(Q-
zeYoN{S697R662big5Fc_y(4H@Fq;+W?ms_!@$^f8btOEsg@pxUyIYd&>R~w+?^c_u
zSFbV}<|Qhrs62;Y=Nlnc?fo>d2coV&tvJ{3Coq4%*54GyO~{C{?R$yvq#M0i31Q(~
zyOcp{69fqIENE6;Lb9mKR^0j6QKV@Y>f=3ICq{eYk@fcSto27LpWxDiGEaf$QG0_d
z165VGAvgrR;-^oa@?77muR_qPU3U%b7t8$o`STh89_O>4+APF_N;f?Gy8~U@k^z>Z
zu-$<TfPb-~N6ju@zU)t%z{Zl|TryZpl^6SzqH}w2)v8r(bXGEIzU$co0|T=CT#1)*
zy})-oS*4c2LA>YpLL*B`NLcB@TyXy&JVm5E^=3Zjf}E(T{9fDF9iQ}wiWc1k^mBbk
z<c)4vAML@kO(il2rmhrn7Oe_yB?c-;;7z%&udmtk#kvejJwoW9wt@L2T$@JZG0Qav
ztI|>OB;3=}Q!xF}UfIu9tJf%)gDjnBXqk!f^YfDnfb$t6rScqcP6hT@^0<N@mX<P>
z_-frV!>NrDicijhRhN~O4Oa9bJPx!8C0npK7I(b8)qs?JGn|g~aR2^$nEUiZEV^83
zi8co|bbNy-M+F!WmAuXH1qJg=hz>Ow9rNbvyWfFu1;LYaiSGFLXsD<V&04ps;dR?x
z??LW{qfqDO<~C65t5sdK;x~$CecA`heG)CMDa}yF>^c8JX10HJTAK6(5NLtM1EoFO
zgS!7*g-6hv?$(%Sz0QTgN4pdn7?emNXE&oZS4M*!^eyAh`HSK#Tk`2M!arU6^TtY~
z{>BD?z0%CO7O9@7eW7<6k=q3EwLV$j%Y7`Tr3+lJGV&syPT)YV$LyGg8H1JC0$PD{
zqhzYC4|DPR?mzt2hi5KvtW*U>l6vvgZ;bQF&#T%Yf2u6^n6bg<{LU5rX{x&6nYioE
z0Xf!sw`roqsg~Ke!pq6jNJd5`sH^Vwqqn5~ZlpL24GoFx(23uP^t^ukdXTL9xla29
z1Qg@X{m=gV^G_rQB%Ga#pyM(=T7RYR5c=Jurzm&&G`#SLYF9QK)AXa?F4D>miZgRR
z7p=$y8=7Xy6d+KdsCVW^K*8f;a8A`wPuTt(xHMfpJ9XCI+z3)VmYzpm{WO|f1(o0@
zk^dqtq!_AYCxRBs?urT;e!Esr)ujP4qSJKy-g-52A3cqIhDG_g8S7h@c9w^-R-HYp
zY*O?YBp*6I@>m0FeFIM;6!AasUXM3<Q7Y)`dbu>KO-3n{RBrSAt!WLt&S=dMs@=Tq
z<w=lRN@vd8BeESRzD)3teL~3>P|qFt*?K*2@1?)QEgEC;T*d@IHlve+b_)3aZUG&S
z=@G*M-8OL)6V<mkc~dQ#B#EDkimH-*Abl%T06g<`p)+Ai*H8_ox=%X;hX{aRCxtrs
z41{`7MVfD3zh1q5eJp+%hf*2!mym}PyqwUhSlZh9&b7vWBaR%I_!cgSz#kzWo((;r
zRFm6S(^H-m>~?2&>6hHxH&PX^!H3C%plun>X_;0eg>dm$fkE*aV#Vdsv#xm!GO9ba
zZ-28;Ep7QMv}z@rnEeH_`Czxcb=)~zj3&YetT{00HeeAEcc>{oJkem7d>%-NpGJ5Q
zg!&-C-x_yp*&>7GU@FQiN#*{<<xvr{W5EZa&2PLLAOSoaX*b5Mc_?EBcfA;L^0BY)
zxoD@Oan+A$Crkzw`-q^(2d!#`(p=4uVzC>b8-^pedw0oG&lOfTQKyi1hMWUwZQ-&t
zzxT1l2cEK9TW`KY>zwo`f$z9CZ|<*h3kBsw3<6EL9?|u@N^9{nM4}3lqIOj&X8}yb
zqt#@QxSYLUj*`FWI!2%_jEA`rgV0Vti`&nlHA!ee$UgdCPXC64oDG1vqpK>G7dw6K
z#EBb0@$847Fy@I&6mACvk~iKFB-9j*b{*oMPWIW>$^HTs^g~98(&&`mxqaJO?<o70
zVQQBkZWeurXao&LxpqNMVva4?#_Z&|Ms7ttQBAD604Am^<cJ^*U#$d{)by`P!JPEH
zm;CeYU61cssfV_)k?&#rTZ2kF{%fBT@r1&~&#!NS(4j}{Iu}LyUjOy}W!b&|9Ebi|
z)Hr-Jpp%JqVUMnVB@$*tDlYWZR&peCcIusWL4W?_>eZ{i7X4CCn8}hq6G1DX`{)G9
zz+wK=K?FjQ=y~@*up$xlbW8}l4=&?ZpU-}81)?g@z5-rY(yG;yN)St1k@x>i2{Df1
zqqH5>|J}-*xz+kWuDC2BBH|l3D}O3_THhXNSy=xopx6VU6lE<zr$AkNv{n3Ph@2d@
z$^N$aJR%2CLH&)=ZVwHafanJQ`zsN3h*p6tMxfh4gG%TIX{i6Z&_oUr_7Kt8L}>QA
zk+h9I-`o>1h`>Si@4I)a5!I<QB3=bu^WWFx$=6oSK&Tp6jH)hx`LRKO{Y7Ib>Pkw5
z4AHnkG|ER5U>S|q?*8{QZ0Sk`HU+J%Jv5&ESN@i?{|<j-e|b(4!$|_+<6^RI|2+*`
zne)D^jL2fCG&muTn5o}?7hOqSu@K5(HC<gHS~hf>?QL!UM-OuXyUBU+Od@|}$ectd
zlM7#{Ifv|xUeQh(DdsqquWSA9DTsB9s9F#AJ>I8(CGCg$Qufm|URV9i^Axb4MZxSy
z4=bogzv!N|zSytzV(ZGS_y{)~9qJwjaxDfbDex#f5*AOf`ustqwFzk(4qiKf1{6JF
z4rEL8OwZYkDxNXyEu?0e*|xFJq3rZ}b4l0z`=@=RTL@QXw#?2rjC`vMuQQE1;5-&B
zJ#CWGG&emUV~3(R>@DH)NrDt<9QpA>1>~@k8#E6@A%5e=4a3;z++1y5@v9Gr$`#ra
z^3!S$C((>RSBv072^Pw*B&hXN30b*JI5!LskcKM!rKrp3)yU{*2`2@xv&x2sQB;%G
z!atXCaasPDoKxb&U{^iY>_AtH2+>MMT>7$35z!4wCO#U;5q~XQR+D98L?b=T5eVot
zOxr3}*D|z>ghaCDG(^TrU{R?%0Xl?p2mZQq_bwlerR&0oIvc8MkOML)Z{YPZ!*7##
zP&3GXfvGQDdX_)l!Aql;8dh^kG*(<0o72iEX^^+_OUj>+!{a`rBIfo$Wb4P&oYqo9
zOv7Y~8C$9dplL$SCOj~L6s-!xDL48`)Umk@QQAx{a}QM;w8|CQp$Ba`xUbhBh6eTi
z0VpzDB`+*SGfy>QuOwv;EYNd}J^*CL6%!OBOlU8)qZ^L=c)$I}*ckL55T_6_MZSl0
z@{rv5Xh?ch1E1gMWdAFMgo#&d17@7t3@F`n_04V-gL_I;32jC{KHK(n8vZMAk{4JK
z1Y_|iC;R3V>&TPd5YGSkdP81AlAbh?LO?tE^@A^v9v=-bkOeraqY?mpPB$~YbOgM2
z&Xt0>Cr^|qTvpf%3ir|O3zrjHpcsatZlvA|Wr6!hyxx-@k{0m7JUR_1+x;Fq*g}KG
zi*QIhd1vP?p+U69_Gl*&hobI+r~y4T^9pxaZ=OiDU>X9M)XQ)0$g=Fmzco#8l+S%l
zfvAKV)HdX~bOt+!A2^YAez$Y$pNmRHFn4ekc%Cy+P+0hE4=X9a*5t+jJdr1N&fq<x
z$TKtUZ#ohOl1rWr39u*>6|%kkaq)@!C&vgefUIFMkYAf{A)LLrFpw~;x>(%5@@NRA
zTGhGde?s1PZ<r$H94ca1vNu5bXX|_eBO_>c(UgQ6h_JwiH-5{WdthyE^1x_Onj$pO
zsVIja@65hZ0E(d!IaHX2{tFAbH?4j*SEz)sXi&5L^N(?%=RBd<xxJcGSC@8%dL#nD
z)2h&JI0pmGx!OX+^$E4|me$1kh`zyY^3JmXzg_1ZYtMZ(^)$A*@i``uLfuB6WVyql
za%^Xnm6Zu_bJAP3Rw!Y}92|QR%#TKg$Ri2rzyF~JS29A}?6?M!JA8Bm2D%efIbHN(
zAq9$SaZ?Xw7P4STpoC>}0tz(@K5(DvZ?<FU2z4!+gAo<0;i5HlFxt{~1Py52L)|B0
zv%Ys%=G=mZqC?rF?J@XvEvZ|2%%Q7)h6^gviH-!>6W;QFEH13<){2dB=*`fo`kHdk
zdtio5Z-Amodgf<1J880X9D<8Ya~rV$HE3DVRKUleW3Shi4yxK2H$jkrpd0v6L7kU<
zT^W5cC?JAe;kC?Frhm0M0Of5IM{j4FkF<DzW!{LolZ#7EM1HrUlT#N2+GKu)Y66i7
zpuC-b?^DHSU8i<#oBipC%nDETM}Fv@%dS=n5iUorK`{QL7$QuAJj>9K5Q}t&7j;)j
z6E<EhUKuS(F`JGvHg3(iPMRr(-dTMB7R@gGCgVTa=~O5<-)(X!S$=aa_h`mN@ro}B
zQ{y>c754hg8-LWv^B1i^YeA6^9Yej+)5GVmAT&DCS@eSk526Ip=l}&eB(VUVLBGwO
zK!O6OQjDb?j{IN@GNdCN2CM`{Z(!uFnUNw)mw7q^Ir_^sC2D8AEXyc`c^ZYf@po12
ztb?XBJ2yWN8jtCju>#gv_aCc<UZU>ohRiT-bzigfyN%hCeK|plf3F4vqGZX=^v+$q
z`lvr3Gx?<zM;mW0e6C+#cSTJ79p(dw=o_Is5rwfFi$q|EFPRXDZ@c@pD#31UAw*a2
z;i1%MK}?mH(<e!UM{s)09D7C|y3CrEsJp}Y<R3lmfvVlCdw;wEgu@sjX+pb>L{e00
z+%POij$~1{S;Q`38$lQu5)^&U$Y>@B_%t1xaD;pVBY9#L1sK|~8SaBgM>prWNXQ3^
zEm^#{BGbC9QFAWl%jj1=pC@yDIp`4tmH;Y}9;^PY1R*KEHKAgYaB`b7VS>0kInWx(
z#&kjH2!f^tS{rKVHj=>;fJBMN8nh~SSOSb&LMwRgOAP3eJvn%RcrXUCn?{B?APEHq
z6*N!5SOoa_Q&o|-OEx)ueQ21pYVLK?NK&{6426M<eJ=eLU@bPg4*a?PhXrOjgbSu#
zW=)g1HB=uNHFF@Pf@j|c9#7fZ_wOV4r3@!|Ov}@UL3x;hmB1==P_NR7h`Y&ksI*Vu
z*Im@eU5wGlB&|PW<sCFbd5$;G8O9^)L7|!%ewh6@p6Yios1QyH#?epu-5?i<CJ`!o
z8Xa(yNQ#?aH)ua!v+@7+&ID)$yg0v`1SwgcO$?NFU=ZUVAVL1S=3w&cV;GC01Rzst
z05~`4%#Cx%1YNYlsNUR<=Pe|eU;qBUG3SX`@$16>%LAnUyBB7U5>TwX(Tg=lhYYBi
z2V+l&odS2?e(8J3eg%cUQF3^yUqrBR#s~?!LWoyqR4GU_ML?Xa0*bAa$zI36rSLe)
zfE9#fmC#b_-^$?@_Q1eJqbfJKQrJP%WR%udfPBPa0(;Iyov94>krId>sS5oG!D$b4
z0eu2o<g*q3z|klJ7cs2x3>!{mgS8NszJsNY_Z^3@Ml{0-7mmBv$KJHm706u_N7C~$
z0c*t0okk%}MH{jXVKxTyTqo(E`IeybP;PF=-k87@un5W~)Ao=4?D`ImTfN+t4_YI$
zqI>H*CXhj!8{6mJij+T;P2hA|HstlRw@0ETE%+!7U1K~_9hC+io0(<P8JH$Z9$coH
zh|vq@69N!C89f3L6C$_zC~lvlgBN~C1+NC?FZO~N8{3K=A!zw<1tePrLT7^|(~5fe
zF%%?ZS)Y3QLoK;tm_ISi!Z&E-5hnHlm1tt{mhsjNNF^h^4H{LEv^w-=10(02V2D(6
zg>R-jM*EHkecV4g`_tvt^`|H2aW6QbdS)<TXNXuzgk3>!Q+|7GVp>7HeM&u}HjUBy
z!hd5%s*RRr`UUk4XQ_)|QNL_e+IP+`{I4A=mM$(S>F0Vhxoqju#ml!Rb+&yDH#6*;
z%$B~mzr-bRTpN1Bq61)x4TmB_%kYrK_aw~gYg7%`3;5j=21HqOPAnoAp~1k1%<*^+
zA`S#a@RJwxbUV;M>eLOT5PhU#Dh6?}7{oncT(B0db^fW|TcKS?zm_ggFMPwxOK<ck
zIF*}TUV>s`YJ_MjD^rK!>RbRjMj3M?U;)B5k-4%X)5Fzrii+y)?)hZOia<Y)*~u0)
zP0jkRtd?A-lzkz4FC7am=S8`<uQ}h%CA^lsEiX2cZRs{fa)2Hr(YoQohY#!ekwb`1
z_^P;;Z|(1!H(TC|PuI<S8aK}pZ1Vs7;ub(l@=@zaje-e(abTVRU1wTZ8TUpXyf{(%
ziC*ADOvikzKYK%=_)Hqg0=Dw^^75h*sRZ~Q1(oe&fx?`K*DEa|{1~7cN*=EjOb`o3
z2M-<U0jbV{6Pa!po0;q<Itm4q<KWl#Z8~J1aEAX}aJikewKBBVRJQzOtj~OSf?ShQ
z>O<VR_&tQ65GPV;iHV8efDF?(nPd?0=yeVn9kkyFfTsj)quTstpgrc~)TKAK*0%vj
zh|1E~S2;wiy>&mSDC3y|wCw582Bb~xUh9;)gdHJ{uh0KXyDIvx7C?Ct4}A(<NN>$I
z@EFu0RNAjjZqJcZOiu^o5brs-Rp?IqKl281|Mb8cF6{dTGDjx1iG`-7LZB#(jxLnb
z9?jGk<(|_RT17iWHTjYzle4*70K6gSga=g8a5g^3HbE<G5+Jl+qfYVW{D9qpdPJiq
z8R0B6`Hn-~O9M#AOj9LM{2ShTcW-@l*ugXWzPFaI-$vMu;m;(<c*avm97qE$(@Zn<
zx1JmYYWg||O{J`5<MZd+7&<xW%H2psgnXvLTxe(|afZZS3m3aB#hyj75+jZowg$=y
zB}^_*V-Yz}FR={Dv%p^68^rYFz2-cP%V=yvM(#%?GjY1Qx||_q1Ff%$GG9jH_7qB?
zHQN~Of+65-L&J-=cdd{CN&xh@9#CO3U1T2li{u}F{Bh6vc!P*4@*t7AQfbhE?E6cf
zT*6Zy37HaxX2wZg2@5~f%AC<#ycRLlPW-;Hnu1=K=+#r+2mue;1P}_C6bjNvH>aaH
z!?K0DoIQ>>=N_gHAc!RBjQdYxFyX=F@NfFz!-uIfk^uk#d9+VqG~i`Q3LgzhYj~*A
z_0BleG==%R46-er`owy|IG{Kvazq<uM3VL9&2flL=+t<o^T{BFnoiy!;CznqbIig&
z=X>uX8$h8P5GTGVbTLgc&1uMx#=U24k<Vtm{)1#1TtD|(`$rDT^mOpF`iCtVGdW7F
zpZ!vI(uar2r7?URq1?gF8luwd>kL$%l_ui{^a%7oK0syMY;)qo3BvZ#Bc49pK!h>P
zHl%ofjL?sYK@);5Mnb~*`F%fpdt;V--j7bN+Br0rRyLDITzeY2g(A)id0Jgy+4dEo
zCmS<1j!H90-Rr{@Q>#5|dG=oP+-Dfh=5~JBW|(Q=A;Xf#9OX7|I3?BvG02b;7|Rh!
z544zyegd<aS#1=1#>HL&?Ty4eVzxmVO`>81x@9{>MN?nds5M<Jy_`%Y5A38Q1O<y1
z>mGs_R?2NcH1mGo@VCn2At5332tU8oWKg+#iY%vR2bmO-;YNWiQUu;S|I2usMS8mH
z_z%fNu{c~nv<PEH^t#5(vvr0k7udj+=s-{}Sk#QOy{i=S*c4BJbjuctd#Aj4?A{JR
zXt{3M;w)rlU}yD4H((wDSbJTDpvF9j^f)(1fbXEMY+}YUijZN29%RfLmJ4IleP&8=
z=IhHHQq<3dFsbfpy8Ncl$B$#62B#5$8HBbzvaxN))FuaBR<f^gfS+vSjNn|*q(@-%
z0+CVTV=n3m4G6%g7$<W_o2rV>XBOeodl1mMai>@f72N|NAt5^E@8X&DS6J*V6y;=J
zEZ)@R7KX<~>vI(q6$2&h;;bi$#MF-460VCr#!ebwiRsJH<@B%J-EqMC@^p9_6c3gE
z^Ezfwz!hcgo3BeFEts!pKgT|A=><P&cG%+8Ukhb_x&X&~eDfw;xhE6Czw7t*nX})i
zaFBqdEo6icSMX&I<s+aO*NkHpF&w2+dFC{#528+H-lr)js9n995OIf@hFZZPyjIB<
zi(F^SuIx7uj7*99^eI6_KdG5vOK=lB8*kl?53|lQv@|v2&^l_$oGzH1EI`mVqf5qN
zD7K)a<Br}UOnwXJ%-cXLCCq>z5gCPskZ1U<Z-`UJ`t3?OC`d^eg!34bh*B3#1U?#y
zqZ*=Z=a2n-3>HVkW7IAXMjt<XFnKgLk<Ej#T1psGP*AWMR34UdU_c3xFX%f_iIYj%
zdkyGYG2sFrrG#kol#~=*$IlCqJ;W|>u8pgY+KRfVE|EPHn|Tf4-BTT{WpeldJpz*p
zS(%xlT_ePb$cV1rhbwXR1vxo6M8$8)8!G07k#b)B2{uai)VEG2g}^5*o8cD5$1D50
zBnrm834wqNu8t&4Mrue>3=0e6FPOQB2G36458!8tmcr~P6sx`LyC)r-=&y}H*l_N6
zViefxoQANb4Bfmf#$JAnF<@Cz9WIdXcuIzyoKFc;I|QjKRBe2;&!0b|KbvZEa&#XC
zHubgZeAbcARs0~)UNa5QmA9=!JFlTimD3UTFn=dU{!AGmz6<>|M{c;R_e*`gQK&61
zJ}KocSOe|qDM0hGxhcr*zu=-3rfA@->s@^K$PuxNg_J`YSER8y=r{XO$s`wg&-Qb(
zjY@0TZ)Apij6Rhwm2<O{c3Bz2rN0s!9Zwr%(sjvyklV}Xupf_sDbDq6ATJw|>5ndb
zdU2`95nWgN1U8gn6Q-W(A0PvQ0F!=2JV<MV!^|H|enBoSE}9BRz2t26IAX;$_SGT@
zUjh6?uL0Rev~gwJb)k3f-eu=wNeoUwerDErhuohMK-x6M{$N>6@4-*Z-pOe`iW7W-
zOl9$Fh3wcXWCubs{c7Fi#!X<$r|+0tBJmT+G?{prW-vxOkveH~GJ=C_aY&=)8YKF$
z#7!d5IcuBap^cwB?ZxPIYZWdA#fTlcuVf?(X_QQDL12Q$B)^f7yB~`Ma>aWPBf{U$
z(~3bJ8ngR6sYptsg8g@uc~$07jGvK#KX$9k+nbE=m>n*vu0BVanxw#oHoRj`#Zffk
z%|E?(mFWg$dtGwiBNNDe!K8&m+|hlDt-=p`?PQjngCK)Y(>RMF80}{vr|$q-spuXv
zG}Gh|4q51P?(p!KHH@p-hj0?3>2eFojQXtSkD@h_bV;M)nvD%PAO@;)`fL)pa@b9v
z1<y64=j4pBU4Qf)duMR^3{iQ!di4tJ@G20#BcQ<qK+t0`=%hJI(ihSC18ZH`|K$51
zU#g?@)WG$9_ix`$*5}(3P<OSK5CG&^>8m$yP6FRlSEj~JC#J`+r+XmEu7;8=9?uU2
znH(K}4A&Ax>7>~p{8Y+KQ_pS0KY!|nJZu^xhv%VNuEnbif718)P+gs}zTnOy2*9<G
zbtNqH&|$Vuh%yKQp|*0M=<Qo#XRzL72E_3z>cNMRvBzqZvzKq=z!k~bVlr-AezexS
zB|nc0nF4E`hT>JM4>~Uvq7o8TZB6?jKd1@N9i`D>ozI~yC8k4aGPt@Qb>jgm8Jz3%
z?|!dBW_b~g2Q`JLgIUkfh|>rrzT6c?!}>y}6Pa2;7}g=&R(-0;b{d({0<y;##6Pl0
z4a(TTkLza9{!hjFh1Ok7tbM&rT)g%^lWY>jlsJODs8QJlw9e(}kj99DGAw<1@Em~s
zNJufF8YG=$L<3z2Dh-2dwdC{`WMxs)ru*NZy06c7%b|Y2F$ZY>y|6jCr^Q>M+4vr~
zKeL3rY|xjw0j75G6j6{Dk!HIE+LMAIBHOkd!&wR})b4yVOvyo*5km5T{_*r$V{itt
zGUIpkrj9{BWS<z*zKn}Y1H&>@wFO_HC`7m;&+fSl`ysPzC6wdFHfiP!>f~4pi|j7`
zwtPI_H)L;h!0d3HcgjxzJb<S5TC@;1N+?QhyjHxY$Ppcze}fHK<4ie3|8K}YLEs35
z5UOdw{Q-n8gbekN1}his1qQU_w5q>XZWqe3-WZpILuttH7>z6frV78$wJ*cSWPawI
zCujJZa83jnm1*zm6QZG`_!u<W$~_leo?wSq@jQr0-Eg#0Q7=$$eKOCxwvG;UhTUmq
zKOcF)!Wo2{Fz$6I$vR4dz`FiU<9^nOdL|mOI5Ln_#uQ@!cK2W#*f12cj~~mp4~7t0
z;?qxIv^)^pQhu!#ISHgf&bFskLC<yYW@5pNb6Y;j#~yMZ1p9=I=(7#U*-<!IECv)_
zRpvBQ2t2(4N)DK{kQD~hi9?<)`IQ(1$kGdb>FSDItLLgrdSjIuX=nlvnJ<}*B}x#|
zdVvl!SkQa}IW@z%e+94j8PuR@HYe|!!Bi14e5oKiOpw2S+4>e**(j*pirhK3#ZUkP
zQ*3&e1A4?XRl`g*_zGw;c3>6{STP2b5INEeqJJpj+B-XUK#b*eeqjy*(H@;_RgyGD
zI;CgpuqHUWLz?Uc#0)-E8hFb}WV&{cgq<JBQis`XZ#6$A9>l0G5*<kfnb3A}=s4Gr
z2)J$+dYl<3Pk2)PS!06taz+zn1f5PAf^Fj9IxzdvQ=hpox5U8kc)O_&cu35n8DsVd
ztriuCHJ%0slUZakE=(eUGfnqECu9R4sIHOO4cosd%QF8!DyT*^AqL?W6gzrEkBO!v
zpISn969!|!usc*b0d$%otPqc%lcUF=3x@)3ulm9UEi@iUr$B#I77pEFp$(d0qjtX^
zF2%&8ykBrI14DJ+fJG}{;94#L>{5ooGtY7rggV~)PwAJm?L>aUjQGlunk(oefe22p
zKG}>_N<)Ejw{I5B2@<WlQTH%W!!jUpRCQ@YfB=EOLSriy*?gCz<iy8<b%pb8+`83G
zj-LX#{+zIcjhV_&ZONKlUq&W6P=62-Ls2nt^gJ>?6H;7KwIYG176+*i)yQ~z05ZmQ
z8s397pX+eNcROuwm`pesL?g#FP-zIxX;2J^;%J^j8XRN~apIZfs{K#SXuiDi;pA8V
zDg)Ho0aT*>d6bdiU=z?$fss=j7vKSPcT5qt6Xwy#Fb%mq8P|RL_9W5}29ur9`6KX)
zkcOeIKR%p{w?)r(mxqUk+tlDY%(kzKcdU#<!^Q*$rIB+va9J{kg~SdHs^?R({zq_|
zNXp*Q>hUAb4}!VG5etz}Jif21Bc%{4MmZc0{@(CHla{V-B09<nIrhDDDnw~rQ0OLF
zGo^S8zg0ZT2Jz(WT}!&<U>0@B`q&s`YM#dg95**~KBxJ83EzXesbO*w(=Oi9_)h_0
zVGSH*FxX**n7-plsUgDJdnQv2Iq)G^qQ<5stS}$stN_Ia&lD33)vB`Cu{W%H*%Ex=
zkouR$IP`-|LSancMbTOV$aC_(Nf{X#K_}M*5kyUr`-mW!FN4Mu{+DR~^uv`ZQ2J)X
zoK>Oin?S?wir%6MI&O{;7EmjxMep{kxr?NK@Asbw`NT7z&a*xPmnIV_=<JyS(2)8T
zJZd<pplF}8>K<*~x3BUikFM+BNxPm}Mr9JSe)=<)2b88>si@&e{v~7Eu*$?_{deNy
zX@V^;<mKf>nq3b#{fI!NWecS;N>+LL%5Ok~iil)H%?Oj@#+(wWsp><YRtQ|L2f`sS
zlnek6fhmY$0wf{41)+igl1e`5x7CkJ3{xO(|Ju_d2tgT1H?VhqOjBDIV3xQ(-CPhO
z#OUP>%@R47=F@nARz`h14iQV4fP?s>5+!qKI3$k@lE;s{oyKp?o*IEkCLu`_gfrN^
zgraeIbn&4}BXf}GWhp@#KnMy>9?=+DK@p^{tnqBSad%bVh$t8oX|j>yB_J4$Ao^_>
zq*&tq^z*ynT(w9V{-9d0eHfQMMeI)4V_e=2p{6$J4a$A!0thE{f3r0E@F5()g_1q{
z``AS&wdoOCkKfWB{j@C2q*@kuP44Vj7BH3R;iw<|uK4?cT$w#U3y>#8pkE5XV@>8P
z;0@Lxh|reggd6BgRNmgnCDW03N^*G7#7Ivxs%&yR!OorE?x<G)pyFVa9yE(;;4XY>
zH^d3e8pLOH)f>xA>>iovJkfXjI-7&^<3O&SLv-bBmju}amm!j-!A|sl$th1m8$eiP
zv8^@SfsBPPaUKF}Fb;`le^9Bnii;U0+R4On+gWHvh}(if;?bCC@I{7!QXN*)?Rp<I
zA2z!tQRmS1b?BC5@Sk~H$UBCOeQ}p;>~}R@C<D2qeumZStLZ6eCx$YZri|o8SKFt4
zburGFm*J!fIA9Oj5ButcMbjsAGZI_KFfCa|a_GqYSl3IVK<nR_8hUD|nc|O-W7mv9
zD>C~I!4z?f95?1@dd9w>T6ad|iDwLhmU)1{8-ccm-hQeyb{;B*8mNG=O;NRNrZ@>B
z3Y7;9XR@63euw#ck|QNCeU8SY$+LMi$FXqSIFJQmO%jNKyx~{XT+&k)hlkTN8TBbf
zJ}}vEyqOH3s46<QH8~EJ$hF{!hF=PP6}Vo2;|>_wSt<n2ZL+bkVQX@L<R#hz>btXY
za?IezL=}ir(5N)Xg_)3Midug;fpb|DA>w?LV$<Z#gnO~@9dfV`<e1$EZncm`YA3ap
zB2FRA;RFpb=b=7SYfJK$_UakQg9kC3uZNJ0adx<LKx1PghNQOvrlFG!97V__OzX+=
z{)waxl|DHx5y4Qlm=e3~jJ=&5qn4)-CuG6+edi|WR$~MdAst?R@QjRc@_Ewg!54Pn
zJU{dXiuyV67tX-ytKz)z27(Pcu$J#ZGXjuab8yhW@Xgnb4t@;{jjZhKkBI7JWo5GR
z9PIiYk}AVg%b1BVoNIUiu=yKefb0twjKJWrSXpHs3;I7wwa`g<g)f^nW?FMj)T4y)
z6tq@bnCmYEwTHsvy+?@-93zIjc^UP%0NFH1^W>rUBpOg?sG6O7_7D*;G>Q+Sqm9CK
zhvX7nvijd4-KXMQ3rwpSPYibvwJc76KnpcNAxOLnXSgwe4>SnIz^fEF5z2)mZ*YC@
z(IE#_2Si4K2jmLR$exgMv=A`kb^3820g+TgJq1zaSu%(Y|C=1^X8~0>DpzPTL1*#N
zApIDbXoh2hyg*=+qC=Hwp3!9gp)~Xs*XBgHA;Ei0b?f;thXCUUM}S+f$k7}6?^Zzb
zff^%GlY*0@!V$7fv0rq7V<DcN6e9j6yB8fDT`5lnT0u{%I6FIgztV%!VFc%u)_w}*
z^)7E}(y`1PyhRRDfL&L7yPJ)34x!IJhKh!)2A;niLEOQ?0okdo|7BR1j8WNBWlSgo
zkoKfLi$SGE4&uO20UKa0S)z<P5$2!4dWp#&WLlCWkTV2fDsThl=Qmz>xUKqW4r>04
zQ*j&)8knG0MU=B!8dqwVQmOJ92Xujok*Mdui2k}~(TFa{!ro!kZ{N1xQQxG+P=_<3
z(UPiZ<%Wy}n?c)94^KtdaB_eXSxY3Sk$MVdILT>zh{)ua8xr>MFuQOF#^P*2kAN3m
zK-%PF5t0)s_dowp;O_xuF5X*M+xzL0Cnf=GJB#1nic<s&P}wwJ2{OdW%aSANyc@7U
z*=Vw*;UqE821n6N;xye{p~}ju84HIu0ag~^oDK#N(=l~7xHvj~1ONYkY05-vGl}T+
z@Sl4(6`*1i3M_}OpxanY1cGE*kcpNB5lXZTBD<0?W&~qW8O1a1V`Y*^sE3q+HF4ws
zf)$lnzc!W3O?M~1+vHnnD586Vs5NjR6_KwCIave<=}>8qq*NQsZzkx8kmQ5O%Hgsm
zfDZu-Zbm?OLT!!;amr7JP|!|8+DZgxXrhgP&DS_oYnoPj=t>VpNXc;ph=x*^TS|~}
zh+2RgRX`fZIM>Hufd`R{qLhTlRa(s0Vy^Jm2?iKFH#H3AqnFz~JY})}$=AL<BDcnK
z>J3jl#vG!nGZk${@a!MozmKt-|0-KI$+T57>rN_`5<V8tD`=LciTRUVAcNwXS(dxB
zb#%yFKN<V=@*+ZMBKx<pvbx)+(-5V=ixz^w-n}mr-@#m9A;zGAC%*cgrK;tAt~}pM
z^Nfp|DH$3b&i%S=A?1D~z{#l4Jeln=6WYDKy-imqX{vT=9Jx5-K*P81Z&#evSK~Mf
z>1^tyHynro2`6>NR9bwXgMjA}X-Zn(erZU<$vKtC=~`)?V=6$~R2pz<+?oM?K0X>9
zyo%{nx8L!|$dEnc#Imv=I(-|_tsW2V&a|>qHV}UI{{4H+0^Qilb}u6&Z~g%(4Ztd6
zFc~ArulZYl14&K}yXD~IiNNj)N=TreJ>rkHU<^*$bS{?wDk2Y~Fk!Pf04I4wbcVPn
zq97~?ai1`0TeX&RAJBgj&Ub_Cwua9`U2h4X83?AY&s$kI!3;p093%{Ac+aOFt444z
z%#z?Xk^*oZ%CqB(2$Dl244Cq$FihzJ(LwkV08Fys<RFcD<i9-i*Xbv~?-0!_nnvV+
zmh?NFM^fr`;|u}Q&^3FXoWcB*mm4bnkbckj#pS$(%y+uwObvrVUWhj%hFIy6Z$Oc#
zynQ#NS!E|YG99CRa12;*eNcc-%#UENd-v|$-Gt-4E5JCUVJem!<Aq$rEXx{byW{V+
ze{!Tp!%P98HH{<JzmpXyYysL3(#UPGUlzKc#Dsb@9_sv^7z+hYlxQ~+fgncRB_=lV
zl7%LFY}#fk1L4z@S<KAgm(268g8ckclN3^fM9pe<@6pTG2BxY;QR4X7YzHKswNL<&
z9D*uCr&8NLXqX%=PHcW+c;Ad@ty}<9f*X%Gld+&G(6=Ow<M_pH%+Sgfe9Mmeo~&iJ
zFbPyY>LLuNp=%xoiA5|($+zbFYuAD5C+D^hO)mg>(Z{AZLC1zPGY95$L+bh6Om!`d
z(V2yT$Jz@v&upHbxMC`ZBT-1rqdbr9!P59K`9p2}K&{>JfLn2UR$uL2E5w=C7n#4=
z|HwRZajjtS(0Eh$h31QkNAt=KvIjL5r3N?;>aPEdIoAIP-BlcHMMe@v{`tSid-J%Q
z^Y;BejAh0)#@Na-Gs4I&MfMrDOm=A@WsCMwh(v0}GBe0>5u@y7sZ^6nX(3~jHAM>{
znIx5v7Ae*DdA=|1`QCs0{`>v)`}Md#kNb10>$=|8`~7;Iuk$?4<2cTavtCe$NKjyz
zIF7xzvLW6n{_QQ@J1;0Hf~(5?YsWVA>+V&!(25-)1~Z^3K#^CLuQ&Dd@aQ_uW8B!W
zJ9;LvlL^B6$RmA6?hD%1wqGx-h|wvQ3Xa5YGBX!_)o{}$%ASl*TtulzE^P1Fo+^a4
zW=zjSN-;U`vNy%PeSbUzbV&#`(m{PiQ$ETQ73OchwGzq#?P&=V<-Kom>Qab_+XbH*
zT9r?a)Gn~T4)wQ0n@uCTgx<)?%6f76A=#@`J+KkqR##I3$Cj5njT||WRPp(V+bKjl
z5_ktIgW27Fv>4dAw@-L{w+CPjVi@H->0HP-k6E}}lV)i)hM(M4lt}p^hAvsJ-a+I>
z`i?A;FeKT(oXnuYADv{%;<sJ5i3$c=rM~P^^-ox?ckJ6sa>OEch*^|Bz<YcX-`kHL
zKfZ5Uz;<g}8{OJx|5frnQX5u_qh3TypU{Jx@*I(UjQFN+PY#3Uq4ou6yD*t%*pqzC
z@Ebnwrf$j(b#}ikkN?uZ280wDu%%1KioYKgrr(g9jcceW;NUrc)zgk1XQ@C%ux=4y
zey{cWsRt|!3QI~nuOH4BlDcfTl=7ZHaTIFv9FrFqco$BXtx;vJm+6b*O}rUWsqF@^
z2gqT~XwPI~97^b@R!v9%GZB&q$Cx7xR4F30#WR9XdjLen&vQk}Z?)2|67o7FTi18w
z9g_A4uFi3`W7a{-nV>h|hx;h%z#0Q5xMY2^D8!=yM|NW*lQDIVNp{Q1oGaH2!g5Av
zxb-(B2ksE=xo%xBdEH7kw_!&f>Wu6?056DN(%}`5YKp+SOV$9>#Op7fKK+)a3A^|V
z!OtNl{BXweZ;=m4HVqFPaAfbrC6?Vjm~9P?@oHMQl7j~WuwezSd&9-0ze7k2DPQkO
zIxFtOq<j%E;WutSqe2SV^1#nqXHr?n<@2GTp{ysog_eD2XZ&FOlckYo@DXYie&_I*
z(*&i3Ut5~DIB#D2`Pm!QGbyAdu1dhAARbz`Uih{d)22mD(yPu8PVHTBW9#(BtlBJ5
z^e;Vncx4o@3iOWPaA9+3(Y<x5eLSO5cO+W^(r{$z*I389fBt5`q~{??RT`EH@yrz6
z9~Ob6F~_rg??PzRLtrX(C!DIh0q>-8MLomT5<d%3%?ZE?CtK2W9EF6WI071gVNJ;F
zy#%{)<$rg~h4$n2gtM(GxmkPRj8z@EL&hxYf&A0Aj<u`QGeJRJ17Zk=*)@@;<rE8Z
zsCf5oWNu3Y|C6!Mq1_v@>Km(o8;dh+qOLC+-tBTh)~S_~-Im{}pXxVm$AFPx(`L;w
zSC^(k9>ia<iRu=*yvNV%2hQP8s3sv~9`Vop9c$JJJv71Iy-CFNWwIf~nq<$OC7<o|
zLp~*HxVo=xTY?$w(-tm_c<+^dh<tt{yLn8^)Om>>Zoco{(Q#0T43GYhRTr-yeG*#e
z3TLb0b1YZ+iecHKN0wwXIl<W>fP)zaVsX8V>1lq!K6lFy%O)R=vd}=hgiZxJC7t}A
z9#DgY`TciL29Tj-pbXK9+42ib*p21e7a=CM5XBqJhZBIG0<QtYF6pzP<DZzYVjk-2
zn&YHjiu+y?1g=MCLT4QYI%Q3W7l`<#_c0yY7(~6uHWxfXDoBbpfL|dnr8Ihzz`m`l
z+y^Mrv}sdu66D$?aMA{FFDcCX(kAMCNSoEAw3CfUpu3El@(|A{l5M)>;E^9eJ27UI
z{+z%+-+v$XtltLzZG3?a*%-W-X_<$|g<G0FA|)cabsKP1>bCvIvYIEVR&x`J1stU(
z6%2F0g55<O3o3VyepPOWXTFE0XNdr^u$@nl?l@AOX)=x=!~YoN0v90Q3I7%-?BTnO
z&5yFiOtJHmP(x<HLKkCGT!QBK=Af*+$6XL6%Q%J@NWm?t7E`C1YP_kf&gI=GJ$3S=
zJ=B0M<2F)x${7#%Y)U`C_i@h68#e@yWZc4kFt?`cxrdIyA!%lzj!V;p(KdW@ag>%`
zB~hUF?%<b~=O6GrBWSBb=J&om$K}?<=9ysLP<dlO#0y*@&6h4c!Sb&n9d!U3YU??v
zRZIZAFt$<e;P28m00oy+_1`_<DLFDNz{!S&`Zuf6N9imWiWjsQ4nQGQqp0E~NuaKy
z+60U71CF6411a3GF_k1-s$S4Sf-Ro}>)O}v9>ppdKA@s1ARr*J!x6wu0{yeX!sXGM
z&|t@v==-omCYIhM&<OipniC?=fCdqfYfbMXr}4v`#;&1b7i?OvBFU7mZIS2^DZPG+
zUF)$@+TjLoL}a1txMHZ{NiXh$h|Y%U{DW`BSVS}_H$nv>&9&xaliUz%28)6VenVb#
z{OT)*m*tRrgqMg?=T87P+OxDROZwE6x*&R0g?O2CjmrhbllC@zDwCy4|5-lr$I#!&
z+yp0sX)r)Xr(~P|-y=5@T7@+s8;0O38lwhkF$YqK85f(_I_^1i$g-@cj$%94>FMC`
zjbzn^Z*)w@+Qx$2k<Oi0Jai}C%+gs@79Ss9B5)hpYVnJeh0g8}9<)nhxuG-*^ZMuy
z*n2$)JjwoUkdX)n1fj)wU64T0%2U3JHB{iD6@><}oufUg9HhPmHIeoNlC@)Of%#-w
zWUHjsIQ{H2J3H?<F0#esS|G;REVtDhwvK3hW3o$%icV8Gkz<{qG~xwWJ>R{1m&2Vs
z2G%5qACTT0g1T&KcznLy4U8el9cbf8oWo7@Tf#%qWv`7*rmUoz@}hMhkCC#sboEZc
zCa)5dl1~8nH0{u#0}=RK-PE`nyWcU}<SrZ0k9b6H>*a}&m9xD`yFEcwd>1|^;mL-_
zxS}Y8I7^2}_klxh@Mm_H;mC>3su-{i(RiSyR|7XvB&ZU?9G#lYQT+Kghq%UwaA?Q=
zZ-8<`-3n4q8ROgu_*a5e>RGpUHL$PFkm+GtG>2`8Jog)VR3vP#R0HxCY~xN#2SsJl
zp>`v6GJh3GL#K@DH3vATkGK|OQ+Kb3{;P?9(*YhoZ_6H`khG+s?KsIuW6)k!4JY(!
zn#N{X-+dKlvCh*|Ghz>1T%6}~kF5)T*747<mCc8Af5@4JFfowa&+*x{J-+c4R6_Rb
zV{W0Y{;_7^hV12RKps1%@8{V=Dr9y>B0S&j8=KdWLhAwU!D2LHk4l1DnWDAge!{?C
zrGG$*pBY!5I-y~Yt74sP;gg+kTetguy7*%KKtQ4ziJ4Qx{XQy@{IPdq{hG!<4AQ?Q
zYF0^b*)yf*u)nNWj@W6xkZCf~3$P-!YS-(PSv>i{Gg2F)+&`CW;;4OTxu>8O8K2;8
zN|V;3qaqQqXw`J#C}C8e0Sd3>EYxzI2@Wf)UKXhs)Cb;f9sM;e_u8@*-k~(9Q{slu
zpO!A6H0oG7W6W|!0LJMpzwvcJwcPephwkREgCp0Sr1yb~=O<=~AF9V+rSC|*&?`Aa
zgGMwyfoY#de@*9Fpf-D_6E^w*?vO}J)5g4Z%D#B1+La*!b`%IiwtHkhFkFRH_<*~3
zj`%?;H!Z|tQy9FaPn3Y)J0)tk`qkkz5cq3P&ugvLQa_!hb~c5Y_0F`{tEt8Sl&jHV
zX~iPQmW!Q5X%J)j<Wg-`#LIaK!7Kkq8p5uAqRXY66_S+o+B`XWq>auoWL;F^pCiru
z%71*fZwFZ3e|W&k?gAIHW%?MASePJ^IjmOung|_l(0_=<fwvJ_1wrEd@u+vh1-B;%
zSt_v!5r$+&c)|BHQorulu{UhCYOd9`Xj}JG3i=g0P_#LHq~7QWU2F`Enh_~7j06}w
z!XpjqO7h$pXhny<qOJ-NLjyW?h1-hOny^v#Utgv&$vA(4j5DoM%mzY{^zMVUW%ndf
ze7tL}-zr3kDZRW*;70B>$tYb7S|s8G-~C@yahFenwh4Hl(gx;{V9RpR`|EZ#u*+<y
ztSW6X4HE3Mi_HyuB#|$Ge;aOE>7;ZYrzlJc+C!nj#-LPB@!~wodcE=87)s^nv<iZd
zoRvA!tzwfh{iwda3SOZj0I;uROo2Zv*kKFLJ00vX!2UhimWZDpT|{;C@p&E77aoP}
zt5>f)>9CKGm3R@nMiN*N&x;|6#hpR;3(}E4zQE<e<Z1f{4tv#$mKr07dQ=fKj2zdF
z;C4tcs4YYrm~}#~Kyj+bXM+hwVygD>k!Jq=AdfDwHk5460-1N9)JM!VB6kxr1GIga
z;*|&Sj6CZmAh<XtfsF_;hs@BrC%2J_DyC~6x&3;}ChtnZZs=H|A3Hhi0OTA8n(6RU
z6JA=GZI514W?_9x$5kS&-36gNH17t_Nc_-gCC!3Qx{s6qC>h5q_)z<&9T~&Nia`X&
ziOVU69`^tr@YF)Qc*AeN=N^vO>`4k-P|Z>EX~EOvY;tT%IU{{Uog_#8_RfQh5f{-3
zG%tvx@1XdWUnuM>8Oo}YJA;MXggc4!E%iU!{7;s!6$ulnAOVL7U!V4);;VmL=4(m6
zO8P<@Ym55{dy91q%X|*7nWS;N?+q&7ypyB9CZ+*QIp7qasIKd=--ui>XV7#mXIM{z
zl|l7iD6A$+(I7>;a&6+uB-`?}akwVXsD9HtdZHcPEL;qBlQIgQrk!ea)8JowForK|
zrm*7AHv_~Kq~P!X2LtjvcTD_@j5%$jD>lYlA=<^sM2Cr?*252u;8ZWDCiy6vH{<3x
zyd$m0;)6))IYw<G5Z7J#OB22P#~4?$Oy>(XZb#sW6EEi}Lbqdsg&!AY%z0dhp|g>q
zi2LFT;&MH#2ujZx#*<$cNoZ5gZb&Xvcy!XH20orHQ#csdv$$y6OZ-~A#AriX)}6z>
zQ@$S;rKJdPj%A_s)lVVBL0ud)sj$S#NKnKc1C%=^^HTT|`Ike75BD+R@Pa%7>7+V_
z5iQnOF%!)d&hD83%J(UJ`t-OW2o@&pp+io5{OP#3`!%&F)iK?$dGd>_N)316m)T`O
z@O0RLjK6H&?_Q9+7XaX>f-pz9=JZ^BKV3gd^lYXP?I0n*1MS6{vwvTtpYB^ut^esh
zjaTlIg>WBmMOIHL3JBl~c^T-C{186Erz>o)T%irmh9^-_E$WxDTtc}gv0cQSVt50z
zVt?MzTuwb<j;Ux>oCcLi$W(5}8KJ9B3>_)-&_s5Q&4uSDCU2MUAw3|R?=#Q_hl-Sw
z^T}y{skz{fcxo6K)BJ;j=bxm6w~1GjN0r3GyADx8MdGf{poG3jd6?N%kq%j{J`AWU
z(sJU7{oT?<(wa_*!yM?gry{XURs>1Wo+NlBApSwy)-<O#>3q%+=;$XA4Z5u-5{ih{
z{kC4tPbxD1O8QK}Fz>HB#C^#_<!97jzKw=S(u^w<>~O(hj?tt_Bwgque4K1)Wsx~6
zV<|8j(C!a_0o`fx-b~ly5|XaEvJ+oiQs}?!lrfKbvwG!mTcjza3Zl&<458rx1CDG1
z->Zb?bA42G)SdBj(J1fDoFgqRU$YL>11B353F@e>ay{BK+r+_H6~w71^n2gY-ZZQw
zdKFYdU-@R%tUC^=m9HKzpeWv>%~9hBO6hAcX-3F-+ZQzT_bzHL-!45)*2QkQ>#IAx
z#yfvC^_P(`BL$d1bxm_uf+F~k)LjTw4n4Z$CasSh7XEUQlIFMy(N5!oIjWsuF1LJg
z9AZkia$Ofvg2yrEe|$mjiw@jiMGua4M>A}vLdL;WNA(OM7wPDT2QhLP8Z#7TexuMa
zlr4k}q~VI(0l2(K<J`y?=fs$AF7!p0HY|mr$}2E)_FlvR)>mS=KU(!Z%EH^Z$$TZ3
zbAV=7#wj?^6Xj{|TX(6pWPtUCht9E9YkEgPS<tNa*{f2mMr0uDd|Lg(QYxP)?4(pN
z$e7rP`>K~B?vi7lfHlh=Hd0aHbB?TiunG`UbZ?IqURYV!DQC{!kS0Uw6?JCvFd=Ks
zo{T&QG83<rY-2~i8UCE@2Sy)W%ZaI3;5WehhE0@}+aOaGeD8dbJOS-#x^I-5Fi{c2
z9oxp?Fn|dQDsfFhg+J0GoSW}RmtLh~R%q7Nc}Ty31E-^H+NaHL#SQPWdnRdgZ@`1y
z+E%M>>$$g6`+E#(Eo->TZy#fq!<(@rQ*$ZiwCx%0WVq;Yyp&<XcKZ4{BiVQ$k9Oyd
z9Zs1Sl-qgxS5Zr}ZqsZ0qplH)E?vA_G0-CVTGt%gDwmTXO<ohVwPI#W%JMnTpqRJh
zD@!SEqJ(0Fp-4I;)P`|WFtN=j$flPro!TRTBq`Dr)t0_DV1Ou8ASbr+pm7QMob|s7
z71xz3hntmKgqaDjmbYr)!o>Ssg(b@Y=`ec`o%#_ev>V#`7yQT>GnevRQO)=4+0%Ep
zvVD)eSDwly9;8+QrA~I~qmY15AxgSlY1=m3`tRR5q{KxwU6`vYParu<S%;c!x(=O~
zmxrzfh*g@K?&H_e2v+Vdke-bEXd50#zbxfFr(Md8hXW#470tB1)k2w>lYd|K)R~`w
za58#chE1-oe?F~9d<s%(x?HNc+9_+vr|3#6Fx+l@@QDc5lY}{|@(pqk8g-J69_tLf
zb!dv;pe}=L)lH!YZRLY=Ez5rNtM}=nccz!CDP$CAbNjS`E4u|;R~}pUj`GlBe%rBQ
zL~iewpr;cWKU6HgGyl)UKMy+sb?nyj0p1~H9PN=EEI?6k7||LlaLO>7Ds_MvF9F73
ziGDz+^X_We_f~IfkhFx&k1|wca=`;~EyHxAk1mtYDC|NwW%WG55f>8BBH|Qd2_G-H
z>#m=R(HF?4wh@(tDFH1kRT3XUsA#W)44u8d?9sz{2SeQM=dVK=K$J^5;ji3->M3l+
z<UZn6txS`WF0*uYeN%z|9GMi!UU-zpP4%^yLiove_3ilad975Xz46Oc@+F#@wXfyS
zO*R!-r_aql@B)J58MLiu0(UK|XY2zf$zqquoHV3d7}MKjY6=C1_M4Fu>At>?N&+1m
zfBm^xOMMHqXM^%&@`q!zvBOjvF`k8f=F9}zk$Pazf>Wf8(^Mi^pa;`8IM5*g9*TZu
z-3Gax8X9r6a-7X>DN;UK9NMOvLz(9pN!m?yFHNz6IMM4~x9Fv8xYVmxvq~+2s91s1
zLhQcH098pDsFYEQ`$^Qz`~mM%;d>WUQ(q|6>MrjOfec3&68xPH0ypEmS?O5KD57G4
z`PnO(?;mlCX^YKPiScdkv&um`N|C9R*pbeQqvwu?^h{Gm(RjeWdviu;Z19e*!~Rkt
zKw1&pDXjuDmwFB}l*goAsyn%_($$t43~Yx+yX3H+Sh;M*84Is~TkLS7wQ%@OsOB4Z
zxo77%Dk;(ZT~SFKj>&uenBG@NC_?JKS#{;gl}=r|O2S(#<UR0dJG&Qs?ii3*s-23M
zo<JkqKhQ)jW1{Zj=B)X4f74ZAwWO2@S*iLEE%nP?+xHruc>Uo|%GLTS3Zr&5ocbvq
z@ZbPT4~*T&&Q9^*G?-!9S^g@yoorv%p?ndzpo;kcg00|e>wBQxhYk%iVx^?xIB{>F
zp7I6!MIWTmPHta{2XJXj&sgwn>kZft-Eevy+)Lgwyl>)-W*W&&1V^BAvPWBX6C98D
z7k@w`8bnZwbB^$RDu1dmL&B~`OW{4pm_*~ryX#<RE?hl5D?gUphLMcU*_z5upCQUv
zjh5~lZNU0QD$x<+&G?rGT1QK<Byp=_dYwykt@?e{e=R&jVC-p1wj*V-O_Whcs}1+=
z-qkjBy3v{a7&}`1?&|AU15X~DdWNc!5=KUp3b{}Z`hLM5tk<Kr6;^+~PN9=*wux6{
zjEmx&C?h^pBfJ|vIP99mW5xU^a*E<|;A4}#vL_;HOC0zRE3M3^7h@!XY}ssOGTX5W
zfBqt0%a$MseM7wWiSRn`w|0L!Nv6?Ss6(}>j7#$g{Tsohc=WG7pj@?`;lNVm|KU5g
z6o^3m)9&9n*xmmfus!PE0o$(s*MRL~vH+By<R2uCDL+1&M2o8a`TzbK*EGI-yw<~H
zd0u1ro#|*~f;vq7-uQ%%$^0MI^wIfi>YOHftR8pTGP&^bwr#()9uPWq;F>j&6|Lg`
zy#Hl~IVS)4rs>^1-~C=$nOAq~^{x-koxb~A``hc+H@_>5Uuj_b{7~Hn+phig-g#7!
z^zk;2lKFP(A6H`uZpu%7(7(^Y=-=nC_uuEx@81VJ^G4RorW(gi-n@Q2gVg>6l>rce
za9X5OnJ`d!$MmMe;EBBrXrJ#NvM8MR@B^7gTpp>3xFpc3?aa||aNf7dU%Re}w$t!2
z{1cIoKr@1=e)Vc<CtXho_EUPU`wvx7UsLsrJ&UG%1G|Y>Cx8%!3oRa;eX5G%L>d@^
zn21g>@-+px2!A@fRaTXcXCUIb5y!x(?FmOI)yO<5Ar4_G@FdQKTFvwA^?newa4!2y
z3<3lurutiMcwM=`<ll#FVy5CXNzX+gdH)DU9DV57eIEG*5u}TudYT|r&@Kk5kkKU!
z9}0+8-|74he_U>z+<%#^ts~j0jKZS%Qb<yMb8y1+OLdQ5e~mfnnsq{_oHitBnUm;P
zgdC1RZOG$Gz*NfQ2lgkbCOc|Qaq+>vCjOL;qL9;w*5En%>j8Solta<lkZ6QHTEpZJ
zfo_|%Xp;^NOp4v9zS0IEJIPqJw&u;#+M|u{k7{1A`gK5f+>e55V6CEUB&KeeRH-re
z<7YT3$Ck#RX1I8aSimuiH|s^o1y=p__y+D)ClRce11&lKRnj1-0_4K|+5jOA*b<$y
zaD4Z6^pd|m6LT<C5TfuT1`GI_^;7Ps-8z=sh-m=J@|H3Y*|O_`c8fo80oI~;ej!tm
zz=UK}xjjg?q4z_&xPneW+&9{reSroUymup&(6ewWyYcX0LU!tcft@8FDtI|7L?*u2
z@3T;r|HPZ%c*h<@Vo^#qArAFM(p{42q%v!~1THW=CYghW$Y!q4#IVhA>RtzsSGuYF
zcstc-U0q$7!)*=W^B|eE<KfiL5ixMqEO)wu1>nFq_5E^?>AF!B(CO_I-!AkuGEQ9-
z^=qCW??`%kV?aDcj=rVbOJ_W~FV1mXy*Zs7p$Z}ga^Qfg>iJrdj2|aW>SJ2jIehFY
z^}{~aS01(>?e8Smd@0oec7f|YPfSHPAf0SM8r>e_8xT$MZy34?0I8Oyx3&>Uwvb~S
zAFquaO`hPd@9ZC0kI$(TNvqX<LC2OOzE5|;V3~7r9)w#(H?1(nMrXshlIFQ27m?z+
z!tr<Fa_KF2O@IWxWzsx@Mb%~8D=~pV(=tuPAoV4mT~nU6hTy*GPEH6G=Pe#c#U<0=
znVl21D;9_vHw_v~-vlpxO3(9o{{aI+FFQo@km&aKX19^)Frv=>#ySBB7gwx*o;k}#
zL)5#=b)nplE%dSnEjWMc{t-RL{qWvRi`>ZkqS1;Ae?=uznlOJcWVN+5E)pSTLmc-*
z#=0E-@kQ%yKynHJj?*vY$5rYp&XOx0m<UHM<?MZF*o__?pJ)AXET$Bx+A<4Q<c~u<
z^_}3$`DM3(lj^!D+UCqHP!#Xldq;OjdNlGyjq;9)RC;kd912XYW*Scp*aDtdbHeuO
z8ZHsMbwua!b$H-ZDwCWOg_r<KSnc`%enx(lbVk!rOu_og$cJcjP-3EXW1aHmJ{3H=
z#qlB<RphzC$fk?rvfE91{(R8=`x|NC2)yd)E6B7cSyVc;&BR2tZLa_E;c#X#(YSuU
zL3${KW0iMJmXdDhVe2J&M)-S9-L%t-1W1IYDxJ<hEx<?>Du&hRVTddjEZ<s=c+mRR
zpUwl}z8s)8r)`g~92`C8%nJ44Q$)?5yxMPlf&OY&hsb%7i3>7lJUMg=k6lbE&SsSZ
z{Ds_IyB6?XW>Rg+p4S5&x5%qhWd9(^xrP(e0mOAY58h%d<&m;?iTOc|TN5#=d$6v{
zLPFAErlWaxt=qwk0HdC=H*+A4a5IMIPGk}OKxsK+jf?Uf^Scv>EWe_fs6uzC8MT5*
zK~*$g=28clbE^1x&7C_pMq0g!v~Pw=?964S`DU||z_-#_27Q$2^+=KeUOrOaIfFQV
zbipG!A;n3eqbO0AA%;T~H6gAD;S&65P>CfM1+(vBbCq~iiTPDlO!1=O$jsMRFNqT#
z9o2->-QDW#KVrqL{N}DkZfn=(Kmd~55ayCrPHIYuXiB89-{Dn!LW1Ib23^5<YTN*&
znZp6s$A~QWE?DUFVBwZ<ZL_$6gAI^Hi5EBR{J=4<;wEuTrY`NtyHP4LL9~}?@g~O>
z6i4_DYCe=q*Io@H`8@x%b6btU^Kn9xPN*nL#e5mI1(Y-yW}cP1VSJ-lrb1;rLGYJe
zFBT~LO7ByS&}PKZ&zfmWUyA=RlY`}s5J7J1Jo3}=Y`Lq)7&8^8=)tS%#Inup0apPB
z9SMzwp}&!L7*mM%HCIq>4gY531g=f#{|b3V+lXVXmVA8r$FaO4TJlZvwPoP5Wdqz#
z#sW@P7i1bAe<kC`z{q}5iSioOkN-!n+M!xh9HTzx`vamrz7OMU*Np#p<$nwrr3134
zm<kHvcdxo<rePzwQ@?+JpD6*f9^U##$!PQMip}p#8wS2B3P~~PWC@J;IPQ3{_@ufH
z7sdf0NwMdVTYg<F*w=_6r}F8jRE{^rkr7P$)A#s;6G!9FY#wpt!zT@9s(o~5X(1f^
z^gWu%_Xs9yxEc3>rwW7b{@)JDGVtlf`cF2pa=Y3|V0C3ewRHDBeyGnW<%*pcqL)%g
zGz-On`r|7n&L!q(^vgz}BBg^$XeZ!4pYENVh*I;P3G2GxEkr%02Rkm+Q$F)pf8vVI
z-ck-+8Ki1tj5Y@WuJg*t>R$&-^4NMP4B`1<orV+xWY0l)_TMYQi&JCJM0nyOY;JWm
zGDe4*x1nOjr#nC6XKsbXZxBIdDZmn-i9PH8(^}D#M;iSl>(-w+C6P%S=df@5=Y5=2
z?Zic=|MCc8vU&n|pFWetwCU4lrwDF22#@0JLQ{rx{F6$|o4^+`o+(rUjfvJA96p^-
zfr!R)4ZdY!xjksm?Z3Av-|geC<s1`gSsx?mOpAdZ0e^3(x&mr@TQY2ows$y2#6JTR
z^enbBJuvc(NPN4k=f;{Pv|OTYHq$Wajh292l`;$QhH%7~c*9@P%M^(W2`n98Tr_sW
zJF$p$2@}^FNI9bZ;qKA2Tvq*<*gUbLa><-E)1~i6wkWZdapdHH7Sz#QzQyZ!7?$H~
zV~)`Npvr)8(pm(}k|Bunv^j2Dj9COMA0GG57gfdgrT~AKGDu~Uuh%f9bMPPS>kfrC
zH**u({_GRmfsq7{5o%YBUIP%R>;>>Iuc_0gXHx4jkK^^$fsylTIS-?0{K+hC+1VkF
zQ(a<9sfx{Uo2!HjWqM%{D3s__psCBOJed?vjxW4W!yR2^;%)Y*=Mer(&`!}i-J{WF
z&|g@pNM&Fz?Kz!=<cr%II5sP~xAbXYhCo`T5BKdW#s#=0r%ZnY_{{=Ym{|juMph}H
z0!k4mE5e+7JQ98!XO@^<Ya0=OBs-UAB4GJzT`m8|F6OvdARg`_ei||sla(w_3Jna|
zEc|072ZeBz8Sm~$OoPSvqs%ASwoRb;z)W=Z>0fntqkj&Bw?((WZ&TblQ~U5}Q(o>Y
z5z|*`rHO;1PEHX@5@-iwZ5tt)N@%I&M3R9>k`U8nG_4l@NO<<j*9XOgq2_w!(YMLV
zII_)|x|!!u(AMZK8jxO$L9!#q^{s9>UL$%LA4r>kvSKbBOpz2)l~`r|vY-7Wb5uBb
zvLMZEfyvAgf_*1VJ*YW!9_ANbF<{~FO=aQ0h9W{idp#?pT9`GxfAS0BeuH(%WN8s!
ziJK1rrHXzEBTNm7yyiIf`jwoCTnD7fT4q%6a&e?_-@wKfVl8&2GOGTDvjBc14s(kH
z*c{9}XkVNp?EsyMEEi#dp<pq2v@|##^xW-df`S)g+i5J94|PW;mQ5uN7GeS&zq<GY
zVV%~2R9$qtgmxrcDIq$w_pZ5OV2$O973D{UPy!iRu&vupyJ4qh9@a53yq3V;k;e@c
zz(s%2!{$EEMKlCD93`z|%F9``GH?0OOd;wD`lTMMBVgw@BYFN%=iN6xSrR!<peY$R
zL!ZQS{Gl``Bpl#m_A{}du%aVB(gnaQgS9WZ9wQYbIq6f2YKV8?D0j#T@uQ@n_bNnv
z80gZ*V8!w@M|;&q%g{fD`&i`UhL{e3L3|DL2?jdZk=W<Cxs@B#iRqwVU&r>h@|ZFE
z{q6N$Hq$Iir1S%+#(%f|(1gvx7UX0+x@rd(AeJn%X3dfYWT<&Wc97BDki%Ekqe*!|
zIh09t10H(RFlordxhzJhLL@MGe<<h$M97rQGvdi0VmOg4tsB?Wr}0jQ7q7t9g@lS2
zi|^r?XAO;g$FUf4VpE%fty{KS$1LghkcfN;8mK!QP*<zyEGkaOJlt9`>d?=kgw4P5
z*>B9CU71lNhUG%P6as4T-K(v9u~3>@_PQHul~O_o6o6eaw9#!GYsumCWW;;xvz{~i
zl`dpPxK)DSykxD?!Zu}2Ht0dtPAT|az^X|XVi8LsB)5F~wrw)S)Pc<gGj$FhshBN+
zF8X@*(j7kB!7DF!Y)?O->Y;NIg2g^UI3v<!A&z{!dTvbA4iZwClPxWF8C!~dn#fx~
zm0@dID7t}-`0P8PyZXqL%EYwLYJFLZxY7@)>3SCao|rECw0>3RXufdW$)LuDVI98w
z(lu@TPVWlHE?>SmarGeWDNHQRoIc_?6^0FGLFT1~sv#mm6bC;hd(XZQfBwv(wnJgK
z>(=gK$e2jvP>BOs(r~yaWX}9s%pFCeN0<~P=cP-R{w3(qMrHbL&a?HK_h0MqqUTtT
z^j51w158WuDc<?oC0O^H3-=y31(40i`LRh%KD#$4_R$I9yTIJT8ezArJJK^D@>E&m
z-+uUT!2zo^2S3d=XF52M)7@e$erMM}n%G9-pr9aLnG3hq4#>+0t)?S>Pte}N^Ebv_
z8=QG3+a}8NVYBRE8v#j4RgIV5+SIbWOrRA~wQqInwqfB#VPyqNVjsphJs1v}eCoz{
zk<rk$sJh&;mmN5?AOU_4pGYacAU6DlJFT`_N-Qx-%&D0tvn)xun5TAnM``zQFEOaO
zS+VsQUIZd8p$}&oad4>L>$Njd*RNkMFq(LYa3b0t|GBd*Bwrb2BM^nyhZ<jPIhG?D
zeaUV{`{=<NVD+Ay9MeBI@tZk?IUX1;QV^lky_uEeXZpDGIJO;fRAW9@yP>bPHUqeR
zAYEviJ0%a<BH4CIPHvA0tSk{t-5Y1?Sdq9I-EVwd)*((N!ZMVonXFxzFHdKCZC<x{
zdP~yc7aHq;h(u$UwkxuiyxI~MA0NVANNM?i!vZ@9kxKkiP$2Y6G|SdMj6j?sDhmQ0
zaS6EJ9DyAH&%=dGXGF6>I{EG%AWz~7BH#=B>b4Cj6@KOmMfah%?(H%v?@)p$m7Tz3
zL0F>*p-~2arLS6eYu%7akF%83k}4u(5&8t#xDk8=S;+z-Nl8aM&ZhIuxn$qH3+Oy1
z+0)G}-L??GBt&xf%d;0E*W1FIRRK>v7;>VUrp1yY$`0jz3E!2c-H$kdvbFRzl6He1
zRxL-axK|lbQ03W%MXX~aIt&7@!=u+?93}&w0o-(=;(p%$9q0$1C)sjMq&G6$Rux3&
zJ#1l;1p(8HUen-9A6CkgzM5^$Bvkgtw~|DlOeFzepPwc^AsOx3=x({jsYsc4x`;Ls
zZk6%*IiWq^;6ak^BO}j==ZK<tUMah*PaR}44!fDjKOn^!w6Z#i9EPkLm66ziV|Hb2
z=t5>9<`62CabdIQlxH;~8BciMAeJ|R+Gi>H2F|IUe)>(!IzX|=4s&1h&)xmHHQ62c
z3N^=~vWmOqi&CB8&sTXaM8JE4R*uKQ+V}6f-I<VDaqg{mj|n6VgWj%N_bu@_r9zKf
zk~GWk$$gj-<gCvUm0du56oP`#je#L!UbOcnOUsxk>2ccW$rlS=(^R!dJM_R#_X}Eo
zr$8Zd?ysRJ#wZ|W?dyfIRlxm%&Fg$pM!$aFa~Jdw8Jt4ea1XBQ_h20b`c&pJ0hjZR
zmT4B06CpkqRD&BI(LI8oO6CwL>Q>lwRrG~q22)BI+cp=m2<o@aLl#;tU*kl=B1CX(
zm7-EolgnFrA)Iq~bQ%?9AITZC55McK%2bFzkXbfd_F;P3sT)quZirnzhk5pe>}fNm
zH3VKTz^g?lYGN`68`2&F2jqznZtEduq6|63CN(waix}7=3`fx_&oK8}#);{XJL?hj
zn4&i;9r|)9p(Ye<7D3s4e0GOq({`6b6Lzghs84*C^wwFg>Npa-n36g315)GPj@s!~
zwkCnsqBSyq{RVhXCNZ<zQ%)F@)yyHni0=;c0(^Hw2Tw>v`9)R1Vk6}1{kVL}!~-~A
z$exNj1Bd2fx$kcOIAoZ2?%b(KJ{NvxQK~hSHnHj$IY!^lV>t`4={JceI*Gg~F)1nZ
z)@K*fSN-|z2hR_+)h1yd4xJtA;$Fb@h*p#w#C`M>(tK&3Vr)7KI{92^Z^hGGhT{a9
z*S=s#A3yKYLj!x+b?bh~nq~jOpk1_vUq6$$goJ>%<IhcQA*wc)%#q{ibDpWL<vgAW
zUCa!mE=pMw(0Q&=O=K@v5%$$9vo!#DGUfxT)+v5I?>2~Un-P`?v-lsR0xEIfFnd`0
z4b56iFT%=qtxEl;7m+L`<jQ68Nb&wa*CQAX^ua@G8`INg8+B^(ddmhlPh#j5k~9b!
z-@Hek>GjUnV|=(wSS5suAjJy`cy7FVXnCXKa)Ku-tgp%Z3E(npUBl$Un)rmkPMeq}
z$&Q71#5pU1e==(8t;?+pevF$oLuN9vngCt3wojL#^*lzs2AAL=sjJ?wcvGBiq08<%
zfwZ@1+(n|&!f)*-Nnywl%{mA9f+gpcR&h7OhUfdY$HJ_4>*f;+iOz;`{n8}m6&8sY
z;o+hvtPE<q%{I4pWH~tRTk*HSnD*^=^<DI5`-uh6FZB~X(`bA4?Z_tq#@pL?Xw{W2
zr_#<^_<CMV;fC!i|8UEz7}t{tHL4(Kqs&WunbZB~_eX%e4m}j;1=tdqffwI#xJ{11
zP6Nik_|@YX+TRFpeyy(_IK!VQJ~?h1tZM(sEn6cnHLRix+Lf#CjC*JKrYm5fk$BHJ
zdi6Zkhl#$r8!eg=t0}Gl*X7&&;pPO3V3sLKm2aITNiiN9omVvmY%6eE^<h-!nTYuu
z_xa>>xYk&knD+Md^P1TN$LOg~Y!e>4Sgsj6q{q9PZoS;fbuOie$=unqXN_z|2%hG9
zk_G4c*fIQ6aJFy#uRnO*S${G0%CPO%hb?%qVPk!#O!!D*E$%+LFZMHc-#5!ef}Z6O
z5NNy5xzU%cC?svshYH+H!Rk=;`dQaY`npG0QqBX7-#W_H^fJ8dAF{Szh&jeWa;u0D
zhDS?f(|td5=$pi-5QjPokBiw&{4Rw&uhH3nQ*e?^s(#kX$>aruyZ4Bk>VloEHcl~V
z_hE3g<N7WR8M7~BO)6YPHBLo|vuNEfm+eU~DNtEQ7gxEoM$8JUPA0jsqD6AW{Mi>)
zFF8RsAKKYGRcMt|vK;XrwPG?&d-`;f4I4f*uRDEY7X?8^fC%z~m;YRFrAhEV@Ag=q
zU+Y!k7QNQaY3Kai97d)`4qUnz5##(mW1hok@51)WH$I;FpV6^1CUN9tRK)dV*)q@d
zQuXU6D=dr1@Tsgzkq?-_fkDGun7RE-%x7GD;;*%hfDk+8CvdQAxZf;VL#-HdI%hUZ
zL-Rm^);i;>PmWeuEz2GDbJTc?Wy$L!e{h_({^xN|3ZAyI@9F))jg&hm+d73>l;=b(
zDjwP)q^37WTWhQ@(CG&#UFBOe=80A{R2$6|tJba^Y4gzXx89E)b0o-Ng?uw7zLSsj
zhuq4IT59pTNkN&=^G6q*JbSW!LeW6ZO6*gDAoR!NcB_>FvXi;xaZb7ZgcfSmxu6Nt
zor9V19ZqM}`goXs=Xt9J+L^8!7D^h?dVEgz73X^l>OuVU{pQ@;rmUivkaN5J0^9y)
z(uQ<fpj5p4g8zxR_Hs>Af{6rcgr?N?9pXm+m_t>h^CZ6}2E`8*zRAP4fyZMUvZivj
z=yjgox?Q{RwtHf>9+@-H^w#1v1J6D8XsGu*u9w=>u}fWIWqJ7?#H4<cvvr1?HCxvt
zdFzIS{$z+#XeNmZ!aX-Ce6UP-O8(`RzG$zx8SW@=SB<a4LU1nWHK7+r10y-S1tNi1
zdj>${z_X?9HY*1%{%zZn);oP-qoenwHn(dcPFD0t`O2Oxf6(x6Pg)C!Z$n@t^xD#3
zf>`eSNWDRSWPK&VA5|s9`|SSt?zi41fVs(d`Bg4=Fe9jrLy}fdQ?he!5K%kJBai<H
z`Zb+kP;|I>fx#m<KM~*@b5Hlp*|{4KAT4Hm3)jA+%mUbVuo)#@XYknX$=nmv8~O>0
z5hh%1^=HYA@~nnMvBx-ySGv0cZE9`TF~xVOtM0_!-+c#8g%}@ts^C31uS+7ld-pMX
zA*n1!yl{#FGaNul#hw;#IkbU}minQ)4%3X3g6YdIXBSxym@q_QRJUNW&@7LIcTXq2
zVs(R)8#~|GNUs-EiVG<gcO7zb#rR@f?}Z!;bbTCvJ5@TK+M_!!Ddqbi=g};@(lm;`
z-r;fU&43N)Yb0XV%RIC&H3r4~xK1%e)-8%90Yo?k!E$#N(QKKh>v~-oY)Z3ZXy@f_
zpY3eNvJ!y(R{ylmI1BugVN=q1<j8PvKV<!U+Kj0AbQI>6RUrUWDf2AxQhRIQy@RP%
zh7=T`<%vk!(@C~OFiKfxny~A!OAZRC2nHrDl=(*VFRY5v$)hIU_3pC0nS}}`51~W@
zn5MXsj^FLoSTjYa_4bk8XF>7!4&^bWhIoK?VoD3To`QGVI5y46p9G`Cl|WWK_y^gA
zB+RI(J&(n!?bC>qDK9hSsi^%|NOYauwzn=%bn*oC6NH8-+?{^}Y!n1c#Ir&MmZ59I
zZ^lA@6(3WE<F+*F^xB)u;S?$`OR+Z>hyiBU>2(|TGi8fBg4*GLj7k?<oKQbs__Hzz
zR}=><6~X+#W&bFH@|*tm;OlD_$xI?qc*rO+X_KPMUmp8V`V4rFzEIU}+YGYX0F_xW
z{}GQt3ZJUiPoCJ#kU6~IX3}<%5{DZ3+u}=?i*OLiL?0l<5g7*eOvaK|fsmN5zX-8h
zU<V%mJ?mZLPRS$>N*;QB8v-VHH>@jlKlBdaErY3_7{r>lTe}_u10awrN5do5okunq
zV&N|aUqbTso&O7gU64j(2fuaJdij?nA4YL(1}j!Xzfm`vgdosYzBn#B!i_~AMTKcw
zSuzf&!SDK$qnkc)0V1v*zzmMmUiw>Zhf!Q`g`JRYlJJDGty2b8U_Y_F_J-A@QDC7&
znP~J4V$%LY=jMa^_cthVbKQKA{K1@?yCb#ts8Mmb^Y4)NlP3#!%H#CN&;tjz+&1{S
z&subU;=Ll)JTd~*)TNZ<2Ut-oi~LQj?j;oNUm%`~dXCvr{Oi4Xhp|N01LPsW^4LlL
z??M*vNV_P8Zh|zFjgm=Qbi<GHFCq*!b#Y|ZE7NEpafeplm)EJH*+~xTKXCq(0)Zbe
z$!M{ZPODztQ6QVi5~r&5Q**KrWH^lN>DrgrE}lKnU=juKq-6r}g1q)MYrz-?+7ptk
zfeT8%Ow_etrk_+R6h;Q8C!A+z5Cw=#p*`*q#i{O6UOrX((`bF`D!%;ZuL2b);G!CE
zOv+tY+K%c?lEI7><M4<Qj{!8<WV1lAQcL*xEw3Cg9#{ipaa~&O=UH1W`t@o?G0H;g
zHg3(Tl`BK67W~}42RGZ{wW^H=dKsk|%?(Q@Av{eT40}WeEjixRcPZ^|4URMdmSsBD
z<dJzEfoN<8h%f_LgxDR@w3qTg2quP?7N!ns-Faa8fkhjGsTWGw49LtfK@ftOQz0p<
zUW-J4De2hbdg&=HNF*y#s7ndS-=Kt=p_i?Ro|Jf0EG%ywoscyrJzFL`-kA5`zFt`8
zh14>j)*LxiiD*)7&!PL*dHRY!bq9#A(6i7u(;R0C^jLF+G;UJf^RJB5k(i616cQ6W
zNR5^Sb=x`D#RNu%A68SfzSHqKDJ)EKETR^nNi3h{rDs!D^e0y&dyrYcyd1|k2^?Z)
zsF&A^y6jV*g+`J%;dnL=VOj9^Aoch7-=Z)8H_wh+>_T})@jGNxxKeqnJ}k_=IFtD=
z^^#gQ_YW5CH5E)2yWXZU&G9F(ik5s7o^3>~{+I4Btg<ysW1M6~D_(xLj<^`L&Kc2?
z`t0sMSGW6?@vo&j$Goyz)B4QuU)Z6Pk3riaes1*K|BY*Z2>$Z4!0K-eK50gB&rCm&
z|574z(i-u9hV4uIJ0A~gO<_(??F`5J%mE{&z$0SnqW}SNgoVA1V$G7A%1)xRE1|TJ
z9u)t=I0;JNV`$e{E1+LmN#`d`nv{h{3acrEPF&XD-D`uvPgI0UZEkAB)PyfJ(yo`+
z9yoY#ol(qhpIPXzNCUI%kyjti3pCB9o#UC4oH=OJVc1<_PA8_Vlo^^Mf*5IK5xH~=
zHEr7b?_0SV!e|zT)(+(!(0&;a`5iJh@)ttXy0q^uwxeP)ydMAsg)<vD6vN8rv+zij
z67vll>RL{cabz^jtWSj(Yzj<a8Zg{%EfHjM#yF`~0eV=LwDNem8q7h6zH%0fhMh}`
z;_hN=TK_r#`nx7HjEd5FS}6=?SOL;j5+~!;;O>7|cG9qnr^*vYB<L{5b06YENzZQC
z=1U-z!uX|fh2<}mI!ap|-NjyBURlzO+Ps-LML%-v&!z)Xh&RjM@7hu8*f1b?u3!}n
z^{>54K{+S9{>#<<>AcN~eFaTBt!2}b2LjppBR(y6&tI72Dc)ThT$^az{v#owZe6p*
zyvwoCpC!6Z<#@%>&2ZCB`^b69gv%Fy_{^7ft|*68q(YKYkF>a??hG$J5<lbsO08_^
zAHikm_7EdUV~BUZt6StgDI(+=^b>Wmi0T7hzd|A?Y_?CWRbT0oTkNJ7Ln~mW`Wg_B
zd5nL2Y?Iqfm+tS6K_^d|7S9;RQHreke>=lJeh?c}Mc2WdjJDx>;IR6Yx+6`HC57l`
z1bBOogL1|v`hkq|vuh%jP6>~h70da-=Q@09cK4UA{@#c10)ct@8VN=P_u{|Yg-${V
zSShi7$4%y?i2=s%3KLkv=ae*~XF4Gv5^qQCU>g3Kx;8sL^{UmsehGJK#Na11v}ptn
zjZ`K{XryN<O@EP;splU8%Mx*if8-&2{B`}4#`>df>o|p|hDJsx(+4ybe<2t2qsT!K
z_YP667Z`N?r}n#zOtk68@Xap@-eM6#Ww!id*FSln5CWEA5miN_<R=yRtTgcd`$C$~
z8vQE3)G4c{t5$=GfEFh|Q}@k%=F1iT4p*`m&fAFd2)S)Kh7!g2u~B#k=LqTmqd%#n
z5F<4R_{4n>mBy&gK>+|Vx;>Q9z@Kj5bI~gt`WvN$`-d7gwRn=|<A1M6dEdyDg;NXO
z(tCuE`H60!rj2|ikOj4qkqQ7=XhEM|cw6R!jrv)fHVMco(O$|*`n1a0S>8tfamyz5
zp;*UKJ5!kJu+eDI+8kK^{rBIelUeG>oFPrr#h)JMbLKT9_mS?M5MSjaq=f65Q}_No
z&de5yaUTJ%<AAe&qK$<CFiiz}@hnyQf3K=JEKbXn6rf^uE;bH??+}Q##)x@@|0B4`
z)oq;&9Y52hRsfLBy)Pr9{$>1hgD}*evO_4FasdUSVaAZW3Uo-*Nb)n%I8%D<oWk>^
z-zi=epoL-AOCK`>iL9W%(Y75sOgOLt64hME^bK%9;+Bbmnn#~Wy}hGUjPp5-P3_n3
z2Ap1-w{D6!LTk$)H-YmR1RgN<nH7Bzw{^6Nuo;@+xlvgO>4}U}Ozv|W-35@>h`+4F
zC+badGf(Um(u(c15ITrNT%_k`v$q|%&H*QCy&1XTPZ3Ou+onlI7Z0HN|6zwkdx&ua
zCaqv1<CQLJLv`#Wd)g00hdYen;R+qh@LC;>1JRDnFpe^DuYM()Vx5P_hzDwom(MCT
z#^U%J-#S`YSZvzqOg{<O*t+joiaznPM3t0PQo}GU36&|;kSdYGRksL>98JMPkMyv%
zXfvfH7<g1YL~)c1Bs&GFI(+aYT6Aqki+Mo=Y3WZYjtZMQN5W{(oSSLALsM_6XcJ93
zbQN1g0{)8h(cWd`T+sa^wTojf)rSG;-#b&JCWfWSl-y!vbW-^__$Sjv_BH$!p{<H0
zoz_F|_cs@<CX32(RK(*HE|dsMC=oUWRuv3liU8~j-!c}{$p0lL!alLydX^aum}R+U
zJS~P&R!O3*SzvNx%Yt5@HJ<<c-j9y=AZ0W=bifUB^s<)$i2yI`)|Lqh!f4!Ap*(4g
zB4kU)=Ew2HIhf$pF`$zr8S1%9s#=T_ej<@l1qoPmvF_EF*S|Q#bRgJ+4&1rF5rQ^4
zcJDdX4Hldf_?Ikk!VfL-hg8TG{D_#($L2D<67gw`tjBg@fUa&-S^<aV3u)zXB;jX#
zZq~S}eHc4x1*Ktp9C&*Y$AT;jbRK%w?tdD2@so;z(5z_VW#z5&?{YvLdk_Z`8Zy-o
zqfTBfjWP{2*mM&&22P)}FN{<G<nm74bT?^{O=XbF`VNC{5C*~Y&mU3-!;n39?GY<>
zh-daN<$diQ5FSTpQDw>6=CJQ$BuShc&^aaZh<%L&^DK5{e+etH9q37<wusOr@Ds)_
zMb^`c(*G%RD@BfGf!JA@Te)XV-la&tH7rE8bbBggS7_x0F<}u$;)<R$kivyl+smtR
zkK5x(Sn4;_*NwC|EZ(3(iZqRO_=Cc6a*V_GSQntlU+&acJj;8ji~-;$?i+d6G#QjE
z4nD?{A8z5sif>HTs{qP%R`Vj-p?cu~5Ve?TP-_^L0iT2IVHmoV{MrtfuBzt>21<|y
z?$IXAk|E)3jn({6^&OeUjM{XcCe5@>D^GceHz2mP4jh;?<1?g#OTK0Y8Bfacpl#_@
zP>XUfweSy0f&3wXBc&QDHF<R8=v^IA$f~*j`p%kBiV#&}V%WTw?(;9Vns^xS>!8&y
z_jI6R5T0=KU3J+Djk!W{?Kw8?bDdD}g_gE3qmUqtN*l7EU~Z^2(C@2DjT_4zU6`wq
z?&tg{8$F9Ky9OFwq?p>K8{+m*Rjr)tvko#Ujwr`LF`}k5nO?xbniz2nWNprzJ)9A7
z^Kg%ld8Q<D1mW9*TP9DYJGzIZq%EfWB->S*r|B$Jh5t+(_a8*AJr443?af6EOB3gr
z*<$`8li>S54c8VbmH<=Q#fXK3B^_2*9*%ubtF`qUbYKUDq3RG6Jr+KyW;b0AEkn+b
zi}lqDC?#g$Do0spLeA(;OSk0Q&B3|}W2NGeI#`i&!^D;G0W0&&S_^rLPw}<xiMb-y
zd5K0PlPj*aS(twb*~HLIUF?46zoQ4<UeY4MF=^LlK7B<3&)Rk1q1YTS4&(4v+Uv}J
z=oh@yCLBnmp~$=8bnfnVEcXmPh50QwV{BHWvl_?JFmcgn`%Ho!ny95tLvYU$`JMPJ
zzU+b%GktnZUELuGFNaG{&&W|BZX$uu#bBa|_{VY>SX|t2LA#hEbPlklbnJk1$p<Hi
zU3sLpjC{XybO^3|QB&22+{_&$8k+anv03EyNyI5Eu)&p%Gd9d4Z6j)3L3fDq&OiDo
zd*fIiP75DHl(QpW5Ok#<4RcSQ0=rwRu=2!Tkkl5RYMZ6%S0q2+9G%$v_uqa?r!W$S
z9{wg2YzkC>BX?ic)Yhw&w{|TjR2<twSN9BTD;u?#OT*%uQ@hpKl!nj$gYM^FVb`Jm
zOW5`Qf}rUC$8VhOBt6^2#78fO^KcGr4Sk?2w`9&Op_>R+EuIQF5l3a(7znh#6&d`*
zsFXH}{Na=^e!#B-p+Xs{*jjL++KGfJVeM=E3#0L`d(H%T-r2`qP)9+Mz`ay5bB-UQ
zPCvZH4fbm-Ek~Q*`J>yKq^<nSm1K-$X5?Rwnbz<EbSe{^Ih^1=Eu_`DqliC*F*{#J
zE<~*hOaA)ifg$D~|JsbHnd)6%x~LEbb((QWHUDgft+PpS%_Q9%9)*n5$A+QKe9pwZ
z(~OKVI3BfSN)X_?ADh=Lfu5vTd~Uz`osoc(^x2k84U@}bnl=CB(>E4(b#{V!UOBY!
zmxsIn2=H_qWz;E569?<Y5%#|PWtuW&mXCqWmW}G8kMKc6$lzNCMb4CrLlqQpsUg>{
zSl;#vsuO<vGsR<vhM{`Ec<0V$1sCb_N+uq(ZGPg;^HKqyk%EsB)`<Si^|-MptXsco
z#0x7U&eEW}Z26B@$~XclBw`^z@ZdOafN!B=8g(k`fxF7xGga)kQ8J#0pHs9FV)E94
z*QFFfZ$CHe2ETL_4mCgiN$}dlsz}j&5Xu!|_r!{im+@egyL%S8n`KTb9>rxS^lX1|
z7vX2-@s-IiL7&`ki472Ms86p+?vGpTd_d6y+OS<qsMHjK^kPMl&daQUK~oAjE)I(Q
z%YR>GZ;2xS3AZW;xebH><vM&W5K;K7QNE;jlBcPJ#0Hf{sd)bTTR$zAxXW=YIGc{*
z<}y?njH#h<dN`F6Ja6wCPB7f47u6Ti^W*jR;aigu>=g%wkpcQ`E1PkDZCv%#;rvV}
zwiz?&P^wHK@8<&c*-z~u%?Z{GIZ9vpQ#a{r$aoBV>QBMj^Ec+?ii_r8;evhcbsN7S
zj*~WMn|@oHTSxH1y0@s|Z5KSh*5e+0z@$Ts`1w&vWQv6586dIGG)39WWp|{QP^@Jb
z0fSTB@xJ4{0RY789QJKNJcSoX6oT@cH&6^jhUjA`N_%9_{qYdiIuOeWSXMLxuS*Or
z;U0Z&1jl}_%*hhcT`n#>ELq!^GE`P8Oq<tf3W=Ye)2lgEH4g-Y%#fTB!xxD6eT>B2
zNz`Oj*mJ5#Edpt(OKPLiCi_EuMsqhI!;uXCfHE4BN8F#P4Km$+lVPFNJb&MfX#fa{
zSBf@fb+U6wCnC=6A_0|Cbl`{mbb7IO(ihK{{JR1^@MY{pX`NC?*XH7!I{<~J6j%m(
zH9_e^ALn5Hi)e8#_@8#MaPN>(X~m3ZY1^ySrTNi1@1Ek><rsM0-UEr!j*AHpaGm{J
zFN`Mj7aYa6Ow0|8a#5E1OjCR|xCe-<9ir`j!hlnJMheF=do-J@rJwA!OzEGyOMdJ-
z|02Z;b)EgbTCVk^I_~()O%NsGP3RylZ^}u)WZZ&v9kf#!tr51<o!+n5zBxtAi~MCi
z8}Qgiak8reiG-|vo!U41nU?yZhFP@2O6_Ah@MZO{6FwZDnYs&35iCSi5FBZHsAz31
z(51&kb2xLiTpb7{63w+yxe0H5#h;PR`T<B}U7TiNfd)S9u6YIfEa4heQ%BM5(K>4=
zcT2DcZhNIzITpMHpVV}E1oj|;Q<doL9MRQRX%?X|oDHC;UCa|r*|C^gw*6y>E@su;
z6nVLHN5uWR@(<1e(ThfoLtP5)n&~)@digQ(&Xj7JKPxjNKK-PQKo%NAb1NKARIv7b
z<b%sjq&_;gd^>@(kCgP8GB?d35hFesHKD=>E|6+zdpNR+k5PELE84Yg2k?jq-L8pi
zb}=@ABhSMJXW08c4I5yi=vj;v++oqBh7&0))ZOZ+68)w0%EbZIh#O{(7eP|0P_Skm
z9Y4rO1|>*o&?4B4wVqOlpbTf?=pRzJSqBt9A<DU5LPH(>CZ}IyMSJ?n6csUed)Ri^
z`kZg+xr4G|@3MS(e~4NC^SZ`eNp<aZh1_Z!-*9lBQSZ;Sn>78b)wm8zHBL_3ed7CV
z`AyeEjt@PteR-$J+0~P*yu*)$#_hLqdU4{rGp&A`_obHc$&TM_-Mx8)@i(9M-+Sv&
zQTrnXvzC4_==v?~*E#NmOJ}ATynCN^=yTiWy6+yXjT#2Uv7(x9dJ>rWo=ZV##3lXp
zjU3twINF{6tkt$BIy^4}xYW0M$cii!w2U^PlZ0`a0td7?lS@!3jSH)%ZyZ{cmjr)R
z>+HL5@qS0sI@n02P($p{iuU4l=Rh5Ky`7m1i>r{w4LG^+Mja|B6BrjNDdmJ<*>>Sk
zFmsVk4L({tTPA#dr9RtiY$zB+GOM`Dh!0W!vb88MylS5Q`ur^wD{z_S$rKs`0nVHm
zIn&P*x6`_QeDPJ}q>UerHd+t9gu)W)cfZH48Mmk#oA69EZ8b2rzJ)KzPo|WShr3Sw
zhy24O+M4kDF`t<jwF=B#!EmJ)$4tP?gnB^mGbjNjtG*^QGe%F}(^u%u>jIXLGbJBX
z$D8z*^9FVnbGhel>0C%hru00GKoM=<DqR2S!W5#8W2Gi8uaI7#fHD0Ku=&x8)X&I3
z)Un^kraxx<(pA+CK6lQT#3hR!e5mtW0Cb}63)zx0cRU$Gq0~3{PHJlE*;y(ex5wJ4
zJhI#B2<Br`3fu)9Zl894UER3?W1|!4-tJctgA-i4|2SKTpHLER+zc(w>pSDf*awM8
zHn%nVE?{e%R%dk+qr|!cZ_fdQa+S2WyXo49J%-R!Y4d$&L_0qNhvE<{bNhx3Avfbx
z-mlG_z16)Xcy7LlIxg8|OTUn~JA=^m{PoK<>tm|KDw2PFwe77Kid9sjtFbiZv)@>b
z4_6oPnhTiQSox%}!CVNf+%~Zkc4s{zdcMb5+@FaJ2eR~$<%=|gRu&+)B#CGgUY_$L
zFxY-v#NTb+y!i#~qx2qYA3ZHG4xwR`fpDObgVRVQ_Oc>eyJ|`_nsxb0k{B+0sTfrC
zTPxaZBmpJ=fD-$h7U-FuE{<5$4o-aYqmeNjqKogJYvpw2>5XlnK0o7rh6{Kjv7QX~
z=2O?WKUp04{nfE1x$Pm)Mo|wb?qbY)a&w!^+Ucm?Ek0T#X);FTb|L3r{QtaihULoZ
zedvMi)h0O3An>nJUinykgM9+9!4?VAtwK=Y<Sl~W&r)Y^8ZOJ&+Vm>bUfwd$X}gFB
zUrEi#&m+#H--!2pjfS_G1o?Zl^1g>Zz3&B%G6ex6u0|kqeyQAwgpbw@BRkSnWC2**
z{e58~7R4dzxBgghJxZtY#w1(YNB~A;sjrk5kNNasX^%W~Nw{sfQQUW8A6kh@&UK3V
zM94oS+C3o>s6+rZ<YLWVS3B{D){0psKQ7onmaxW09C!kQf$S~e&+oX?E+6l-uUP#m
zmT9E&GI?&_m8Y(5rrd0+k2gCY05Lb=KKk?Qh)-MqZ6kKYJ-|^b0n=S&C^0|z&*=nL
z0`2(U6tNyUo#MY_K_q-!5bcLj8(87juhRbY%5l;9=<426){}XKn>*Lc-_ET_pVVF%
z3PLZ1=cotwP+Dy$2wdImZC|pffmH--fl$T0AW`jut{?uDnDj#!@jK9^elMj#TlFgQ
zXvJ?u*9=d_j4>W$iIsdpF@tVo!kLxT!X%J*p^^wKh)yU(7l0FJDdUTg&|VZ@AF3Kr
zsE8N}M$E9(J}oqk^-(7R8tv~Qxe_FymU`zbEoG!{iUYY&H^@ul<Hzn@(p6@ph^?lY
z5U~K1$~Mz%^5iWtXADu1@Q3uhdv)q`lD4+c{fTRoH1=oSHqm&~RRmN++tdTujSlpp
zPh?G>VycRX^v7{*^<ugri`MDdSC8-FI43@^#WHS`X1L?yl%Z{eJHcJaB0Rj)EEQ{R
z4q7|0BY;cIAi)#Rzh1yQiKvL%x`e5ZJ&l(yKPe8tI1jc+TErunIO+GDJC#{g<c(tD
zF7hDpVM2Cf#T)IVP@Gl*yWt*<vuGq&T)C0ZzSq{lPXEYLb$~uDQ=x#BuM8CrEwUGC
zQysJ`a1!*Ykt#_~S*8XX8G~5K6l@4Ii{P%fZ1-p+Y&2QXJL+Ls+q;I7q6_-Bxw0E3
z?vk$VE)}$)zK7x^#jfMoha@BnYFSY5uHCPJc?HbekTi<>vR~in{xOQTBOA-*D|IZ+
z4hs-QkARC1C0X?8=p6#ZAVsWH;bQ~@;SASCDC{_gXOh{O0uPn0LCi85I9h!AR;3wC
zVLeEKJ?nz?WMbUL#`?8_yixXN^9n|=SU79h4tXNzJ=`X-7v-JB23KASZNt{f8|S5w
zA0=`?sCpY$L_Gl7Gcmd==G^)D`KJco4m9ZAuAU*Q;$aPKhv%4-lq58Hp$Wm_l_{SY
zH)99O#Q=H<+U$tkU8RVfeHMV-9+q)GYyx|+)C+=+BuY+^kC0FV(d#()q@7bM>NP>{
zh2#!Ji02uaT}DlYH?Rm?<fF*pZB3`p=7kOjUSfOk=#4vfa%5do<H=Kp^YFa7-LI3R
z)em%1h2)COI*n|d++Q#{8TtiMIP!&cGiI@U`6hm1R(iVWqQh^OiT|~5C727!6uEW?
znOUbJ4wQOcp07oBOOR<o+X<8q8OxG-N~F-GxprFGNL?0s9B+8%-Dpnnv0qzwDcpaV
zSSvG+$i8NAi=?I#Izw*wlCE5|ISzfoydz+8KeOv<M<WMeBW5-4VtEX(<@<pH&t86I
z>kN@@b;jf_#V588R%x@d9XL{-f7yY|?D=`VC&V0?<HnJ!eb{3{XZA078zB2sPIeQ7
zHJrt#5t041NW*7}omrC=S&<HKf~uHx<v=k;LoQ3qG;3;V918w{7j!!2d<DC*gjR7T
zrQQj~LuTe~q%|`B{j-ijltGs&sU8@V@t}E5ZMcln8y?MFH#S5>S%kooJh%|BMzRMi
zw>KkTd26x0+rzRo0&7bjP^6QnhlNo<h@p=yh)79dFWRG`66D0?6gbHfk}@Wq9B^iB
zS3D4LW)%PM)em)_jxEB_S2>t`QoB;Mi>&~Ma0!K)Zt5kOodS>~w!n&m<afP$cVY+#
zmnP&5wm`_)vnG5YAt?ck;;8_5akw%Ldg5L2Ic7}@R*;(NzCxEdjK}bb5QltDT-Yhk
zwO6p!=E4n$DWh66I6KLiV@X&V@_k6X5bQ&QHb%N5nhC%)_}2V~Dq2K1u&aQ#AjHwy
z*UhP`s}m6v|8v9KFE!GS`I{z}b`O=-T@E?_LT9V2a`RdGN5ZZ=J>%YHo^2hsa=SPU
zg~5GPn4Kb9fA4;L>zeq~aapNXhPFwf2Ug37yb$NTWPpGV;#48dsVw^W&h@k-VwPUr
z=5T2Y>PJb2ym04tAK#e+OeW-pCjAB@2Zp3dX`fAci%L|u7wkE%H*{4(R0ncQ+OiH1
zu3<ziZ6yhqM;lHzA-55S@6kYtprm?1nQlrDzkq;r3$*%>wzjULC9jlHLWn#NqM6pB
z74o44wSgk^E6u*WVG;Ww5E`EhtQ7Y6)TtpIYRyE9#U^MG#xt3Mgn*;Qggi(F6T$8T
zXA%@#7=;W+;qPuOr##*D`4%3J3=SX<m2!xtJ?E(zXSRrzRjRL8WGq8V>RuKe`-pxS
zgf?lDK&zmMFMJbGHo=L8S3SH}1SpOSx&8k8YdBtyIJ|A!7Z`qLZfN*8qnBQ|P;mfW
zZx=1*s~>EnxIp}Luvgls3H#hO{yDemeHXkf($muoZ}zeHswl{xP^#Pb<cwz;UsTQ`
zNJ~gw0_K|&KtUiz1M_ERt-U;GM5s?rG29v%VWsC)yPEGCf4T9BH{CaPY87E6&*g^j
zM356t;9nNrMFw>mZJY_qv90H=D1%qmd+w6wA|AI;f)!C5Bf?28=ssjgwIsIkBH#f;
z0ux^GDhK5#LhNZGwn2JE;F3#}g@>V&Cs=`46_fG>=a#c3j}f?n`9G;PNL7|~-751%
z=?I=OWMU*>o3;^ZhS34Pz6}zJ9g5tt2T`XjhPbeRk~<Q8?63!v*&5smnfc2)ULl$%
zNs9RBz2`fSyuxi1inp+7VGTfmE6wQfjl?&g2D@8-414}s9ypUCPGUd8d5nBU81FM@
z?u3Q|t)sj}*zt4c?zY)E%YBqVXqH2ZOKEw3(-X0I(~WiQ<D((tp^+@TG<c~V|BFgd
zFhgR-^Echo-fh{s<l0l$KPIqXc<3Q#&g{Nz!`MQZSHkB=BE!0p<`!$g#<jFf#-T=i
zaedLX3)S4;vSrJF<A*9DgL4aVQ4xH|NliVSTTsy5FoPO>!;XQSTmP`K*|0WOHl&$)
zy@EI4-)LrLmR<W{)%cdXdQEU?fbMee=VQ9v2M!wKe6Fs5bB8W`K(D)_8r{b?9=h{G
zQN^A2?}|6H8$BV@wW0KbmiyM*D#xSVX%&TI%~iMBFF*7{XoV;_D-sRvn8VRrC#&vO
z{`nG_etdS-%Cx;wYSHd{ynF2fYrUI3BZ<zyolemcd?4t_=%DB4vK+E#HC2Kw*n`u_
z%)G!hy<AVFhP^IFVRH}WNiT#UrINyDIq2mu%g{k6L!-Kjm6$|i-(C~)EGH08HOu6e
zWC+sfA3=f*anoFkux?k81V|0SW3T+H^$?TwZe2-MNWx}vHD@b^-A4hU9gI-V-R0hw
zP^NW^P&_>`YL+KPn4BQt?vL3gGALBv!%M@W1#CVB+s4JWK|eLdZsW}5AIwJ_VrRp@
zM+1BL!(p}jk)?w(L|?Q$)dSlyZ657~#jTnsmWP0JXk5)pJ(>7g-QsNM6hMN5-JC{U
znPb~NE9Bv*N$qSWT2ux2RYg8<qMIBX8|yZWvOsu<L{ReszjO{E(!kxI`LM>l{~%+!
zeGijnw)y;Xx9e^J%ahl0bM;D2YY{+0tKNw*Bj*>md9j(?Sia1!ZLXy6>pwSSOM<CP
z6pKiz_9vH|Jt#kHmSP-&5;m;!n^*9=Rjb+l*|pcbSFetiL<g0J#Y3R|h@g%~6GxsR
zA)Ge4!<!>JD)xrv9&#G>rEI@WP()<dJcU49{MWIzF&466TdZ&Y&xg?~xl|E`0_H+l
zWOmC7BK*jH*BxDmv@$aX6@^svOkd(osi%y^K3E^69*jz`Ob3${SrB<FnND2o;N&X?
z{pJAghY|lfef8DS>hPF(136qdgPyxKNsGJ|#TfNHd)#jHx^`>{PG(Y7lFyoR?A|U<
zhc+y(SPAiCWA6=YmM1z4hdTG$pFPC2z^LE&%WkI&vfp`h7k~6hd{zAS9;n~5YT!S)
zDK7}j_c)cpD9Hg_Z@x5p{ixr#aY?z=oa(9|TE?+y{v>NAw#lWBE=6AI{QSe;eGClV
z2Vog=?^MdruV=!4-<+s1GoT{$XVXhxG>`sITEzD&2Q}B+&-!7jhK=)RhG2sTPqqbh
zB0O#8H-mm&(p3%xIX$cL-K+m{-`S^>o0`=#F)d=hdD-RC4<Rc_Shpu${3zdrI_~D)
z-%gnlw`b4OlHsn~Fsqd)2wIbU!FNelm-5k(CA$ZwJ)M%axxy>|pxyo^E3EJ2TI4j#
zyKjBsP_bL-`mvN)I>ng{-#65Jw+yNQ;?r{bGUN&agByJGnbMx-o;H7jR$cco4=4CV
zuJ7rQ-$ZH<_i%JF&VR;hjPUNf{pz8!DfEdNXEht72tYr4s3~#Q&HU=hAal8s?1vZ%
zN$D5C#y|bxc@+2cdDlcq|55E_aU9#|-rCBoA<LD^M^-GdFlsH?2iLFLN2iTB9S9AN
z!cecD<*J<qjL=-RQ?InUZT17L41<fO2_UKO#+s%biT3V3iv^Y0CK$4+!;Z_PhthAn
ze|>Og@VXErGPka0AKGP<RX*#mb;p*?n;*1()8Mw@pQz90rfr4b%&|y1_UOAqD@U3h
z-4x^XlX2Diq7CiLyc#&XWp=NxXBg%adVMWJN4Hy@RUX+mszn4NDA2$gJa<_&+x_LM
zkSljh#@VK|zt%9y<y=M4^pr8AfME*~XRdCT(EISUq(cv~ibn@L?d#7*KL3N`s`Z!W
zqa!=CuS`7hfWl{9xD6=l$a4_74QZBex5-9LOYK*IrDU%gYn#3xpw=ds!Q7yqSvE~v
z+N4z<_CChm^y$|hlACn=VIEd2Sa16eHMLk*FnryqVK4;)L0j>ef!~KHW>>5J;%(fq
zBMWda4@azPSl=_M8#vZxQ8(JWc+uwD!%e~~QoB7H-OtpqByx9BV&W~=n(5uPayAav
zUA=0j)ytRlmS$!)JLkMS^yhVh;C-OJ-F!G3|7dbvH_^Qzdf%`U3r=kHXg9ihJL<@Q
zy&eXe@`Yi*GkWQT-}sW5l~9+VZ#mDLS(AFn+}5M?!_dne8=g4or2YOyeSQ6%8m%1D
ze#chN{M)Qbt7BUJeS-@N*x&9CmLJf2Gp^k7_VE01rfU+CvS+;{C_Ved!7@;&WkH}T
zhs}}`_-;mbrHMx|ojq?+&aGgoyXOrtW&MJ4x2NV-KjDNRoR^dpuRnk2+f_(~waSFC
zI&sn%U<2uld12BKQ$3!@BcrS^UkF0(Ak$5_OJso@?Xx)|;m%8ifQu54Rn7}bH<+RI
zP%s3x&^}z#A`ce0a6N^pl6Jh%--7b{W9jOEx`pHMb^mud@*$Yuyf0|88au+}Eq12m
zd60Kyd2SiqY^v=~&mNtMyi{I1lQLH=_*$H{qUne}^yDk9Xn4r!Ma6aDw%*_OqT_rR
zcYB7D4=4wu)4$uO(ZtjjoW|}m+y)N?t}wbw<Lk6c_o4hc9es4wi(lfmPOafAbg1`2
zuB0uFr|}G%0T!R<?8SSarCe0xW{pq=7eP3)RE{P49L9A_dtiTf?^VIWUfmknM)*=<
zl<Bbg{0pjBnedH$i%#FZ&!aD%6!Iz+8znv1l|2U%=kcPLH#HnwGFD8c!x>4TKR7PG
zsg>*H=0Fg$_ExRLJ}jMcMycDWD{0)0F3x#^WKO70{33*|nIK<sG=W(g_UhbD!)NQ=
z?;_?2rVNGbBxrXjlcVRdbe4C8dNVo*4C~@#-=8}o+z4=qHxhO%-T=XX=p~62X}XIl
zNr3)Ob!Q(|bH1+e#tbvFF=oawnna9b+e6+mNQN19V=*Bl^yW1wB`OlhFk*&emZ6kh
zBt@xahF%pBg_NWhibR{#@|F}zbw1BpIWzmebDit_aZZ<O@Aj^>TEE}#`+dI8bKm!K
zKRin@NRd-rx)591D(fM|)x$n99HDD~+)hA>1xYLC$K`a!)Kd&%X-C6QV!jtfczFjl
zbkiLZjy+0Dg;z)pO-g&{VZ71T{~PCn87UtM|6TcnsPls1vw`Csk95x{MT_r#p@${I
zWgB;G@b|*pLoSXxiD%rPn)9U&@`jvpPO@z1G*7fWE%*&xI}Oa)0txuOT}X#!ApC@m
z-7|jxLk{z3sCPAx#ue0tOzKYoT7A%MH`k@^l}G(^ZQv@9!a*l5Ze9l~)<l6IJ{%yq
zab;;#a<xBy8DBtrv7wu(`FdDuBotE(MZ$Hi*RFake8#9%^cBR=0bItU0*=(vI^}>?
zAu`n4$+=qQ975yDpkJ`76J=?f$hBl`qMR^Sfj!a@*a9%0xV!F41<aZnF<(|4xIj{t
zq^s|CGg)|gN=yIhomHSw(IpKM^Tth?I<*kUFd8{@{W^K1Pk4By&Dp^OBA_6NpR&Y3
zFlCgYw9wK4JvVsi7#z?TAymX@paBXjqs-}Q${$;;E1)Uq>bN5bGnoX*F{C1ip}uK8
ziB%FDPi)1+(Iha0QkZ^Awh64ntq5O*mext#AtNe@%i>vo66+6ddV2)=CA-R|bCGFf
z4tSQuYdVu|GZ#(pijz<?&S(WV*g5!l&>=D@32HdFs=@f>dXXL$TiS`M11IThI`*5G
zv<f*E%iZ?qlundZd(xt85hD<a7-HEhT1UnEjksPi-w9TiBI_jjP&OW@#+j1GwYA%9
z8eV$j=~`2UP+rL(DD#}xLW1OH6O`Pk*n&p-zyIEQ53y{$Miw7rbg>KOjskg5$r1j^
zFa%o?xdA6gtl8rIY7C*sv9~c8ha`KHz)~VnhJUm@>tLj`PlB8d4C#^wvO=MFU2MS#
zL(t3JA9m1M@ZL^y%gY^8jv?8$5cfJ!M07Sx2|rs7)Aw|gPwPgRYT0!}m&wI_y5M$5
z4K7`V(>Hb$6@p>>2d*JBA_6fDnlRx`@saQ!WPSvYl?Cw1rZ;upSq4=70vdOSRayS%
zY#Eb4MaJl7JmbOvk(sg<)o3?)G0&oF$y#0}>xXgzr!gMfG~#gAlCKNqkpXjwb7syg
zi16L<UJzQjzHUdas~cjIMmkon^d0W_%XrVbH$_Gd;6TUal>Vl>913RK@0wLW#SE$H
zGJT>*hEw)Z1=p}GMrGX=#W6VB+T5G>wr<@jV5zu_6#g{&v<BQLWaoHj#*GAL>~X51
zqn%N;V&*Q3mt#wG!Q$$3$-osOdw8C6Ri5kD*H9aa=O={aaTNDP^UB}H5Q{wf#j~t#
z^&UdVrG#K}BGOGT0j2vFw_3zjkc;YRxz>>jHS&D6kVz6LWlux0h9eodbk`MoG)mtY
z;00I-^hn5Kmd5*f>(2F;DhXy>-I$ZOLff~sA;ISu3svXqjh<T*h21fFFKd{Hw}Y~D
zPhvR1&l&<<lfPyeOGBNwry+@!x}775DP9J*I%-wwKRnyn|6lj-9}&ONZ)6_0kghr>
z&3M1?LxUln350vy7k9Zzl>J%p10i+1T33{Ln_Vvykx@V!qv04eTjtS`pC*6P&NOjY
ziD@^^^#S@qE3u&6{*6lN)yxT=xGeWG#Xy5Wrzk761DHS%*iJ>6QNE{_nS25O|9;gI
zggTCIe;^z>sBM5g;Gh#=1FMt!EEB=_a-9Sx(Nwv$o4&NtiYi_^h)-4YmezB)du|)K
zoib9KP0=fBh`ZiziMvZ!zY^yR6~Ir|g!ue-;;x>s8w0y6^E>ou+d8Wu*V=v9UG&+8
zXzl{^fh$z|tz0xb<y%&;*L<vJ#x_2Iw4w{GM#Jj06Gx-X2^JuyQ<Y>Nf;jaQ7u8$6
z+sV<NKn?}VP)3qy$!vGcHIHk*=66BO#2Gu)1X8%C%nmc@_?A}1t5l(e!ey6n3V8fG
zclru~?n=1~^t-Tl^`cP}p|jhWzMTR%GX8ImzJP7i=jv;np?SRF7clIOH#`)CXY~1W
zZ<dT0GH4jodsVffV&JS>x^(G7sHSARi1A5SEVj=*ZQOrxp-l*oHsyFcd9DmHRmuV+
zgr^*50aUJMNy^J_mZPJTd*My@GLYeZ!Q2H}7X6WEDg%le!Q#kXrYL0}5s~wkCGyIv
z8nQEdKIurX=(tDTXsD|b3#vTD6%GOqXocG3JDc1T-s0rCz8NvJn*szi)2*F(PvW=e
z*%&M>doq^m1!1IyU0e(rEFHYllqC1^&g_9vE4(pFvVVG48dNzRoeHaX09Q4#vhbVY
z?@jJHR3QX5Y*Z{osfN^ARbg`B{P`(LlG}!h{rv+Mxz3))G{W^80BW8;zeit?)|m0>
zQ1U242CMxEsblmN5JiIRk)p;pcLy33b2y3Bkvv@ajLym?FtC#xTI?&!UK2KL+^Y*`
zZ8TIv%LesEdPcq$oGVRNK=MwvDMZ!XH>q#n_QgQk&$f!&<Fvc8%x(j+F=`tA4syfx
z+j?_;qT@aq#?XNLe0S;I0BLQ1^o^Cd`5{Q9QC@x)+n1+s`0BKH@(|az3p|dG(fx?I
zyam>$YodZFNW2SMq?v_Gf0|N^I3=6+o4W+L1&p_$0)iKpwXHgj=gE6$Hc7#8an__$
zCO}{2Q<H~{JJLc7btfeV$`un<-H*OIT=XiZ;#Bmf6L|(j>Um@6VV}cz2PUJlH`sY$
z&E@9l4CKJ8K&daScHmSl&zeXjq_C4ab_zvOuQyaF;rBI6b-_x;2%RY_cprkoB(8-9
z0vV+noJ%D3g&6Dg$7C@JlK8IQRmW(GNL30NxnH@`_9*LEvi>XOI38j7dUZweaE>3f
zAW#Ne$=dEZH%uRT;C%Y{MdyC(<lKvF$E5vXZ3Br0^qyLHv#WO{n}l5YlC^D$>mmM}
zPGq(;TyQ$6wPqtzWx+VTeIc?z429<v{&c7tJbZpi96hsQjHrynAE%S*LNm9?);4_r
z_i6cdOLgZ!&^CbSz~rH64<k-E<{}0n<24HCX=wfT;JmeSogY+J2k%XSZmv~MGaMV+
z$hGOK1JK~Puhwa4>5WiK{$`F=-b5VC7@%vZm<3o()7Y94=TP(;n1t>Vxu|8YJ1@T<
zj4ixw?ILI;K!@=vndhR3?w;vvJD}YuN>e|b0d{tGPPkm>gCL+A7H|haAgFwpxr_j0
zXq^9*`{Zh29w?{&OaGKP_QenSIFm=6E#u*a1C9(-2(u=QOQw$p(^decX&(#@x9T_k
z>cm_zQfA{HCKq*jdndm{bau8Z2~`Pe^~!y=f)$ltbTJ^xb#m(^hNuGL4l&V2OS3>K
zZ_4PI`n6<+>FGL+&eKU;$xj9=+WX>WIJHUq25wPLo@BqA`9{U7hluMjU;cAPRrVg`
zIOni#6&rBW)+#!RW&##OpS(lERs%h2<<_UMjTG_L;mJgUZ#7}+**sc>#fC>c8LJk#
z(_!6#v>l55qW&dpv}hk6HP9pWDr}QF1d&vi(s!eC1gU4inzwIqXJftJ`f>gG9WvjL
zgW0Sf@}yt%#q|iQ_}F#L=yX-60t4I^559vL`PK?VzTs!zzS!d33i6$=YJZnx25)G$
z^~^D*<6>f5`TqU;YghR+2j9#U0X5Fe8xEI{m@R0Ry7n0~WJsQFC~Chy6dj7p51StN
z<_~T#<;L>q;SWG<8>rg;PG7cCtTV-z*w2=&;c%dRr9@Iw=BJNmPM^W&t(hzmTFPca
z<fz@7g^vT37fV3!PQlddok=mJXk_kl!NWvUYx^(d6sx64nFn5x$mNy$jW?NOLEh!{
z;J|WsLXC>lm%wnHG02wiM;5`JcvwCZr(+~mLC3%1CO$9FXqIy%aiEIR57F*$hz<@j
zuQyb|55wBtL}T)?s5Ra(r5CYU#-<>zSy7k<29i6en>uCb6Yco)Hbu$kOV7Ia$hBr*
zMky}AiOgu_fEfh1B(;~&^PDlG){~pPl^RyA7V=QN^m1#TF=&;9>^N?EH<odfZ{F|M
zkC*^4_$vh*acP>e2z?5NPD%=@AFWc2(yP)urm-TUa!Y6H5hkq4w*s9?bE@nm6@74v
z?<~@A*YTQcqfdVb9njHV+!rfPeC^%T{$83mUpZVX!z-pwnX=FmX<@;=_I2`s?;;1v
ztOw62cdSM7zIj9Y-uXm&AmB$8vnb_2+aE|?q-Cw1)&x;LkEWjW8)z#sJWfKC>8(jl
zTsH_v{fYo4lP0vZ+S>Ks$uB9tV+eanBP9I-*QhV_r6&8!9DtOsOV*ep&H;nxFRnN(
zhg7USFZXXh*%V|tQ;ml|I!8f2BdtaHJ#g7j%mNW1!KgHi>&8detN92Hvm}*SzX3FH
z;O$L#jy!lw27}J_l)0oo&nCAA;v#t>o5hTtb3tqztCPfv6akXHg3LVa(I2@%5o~HL
z|MtvE0Uaa8t86&k3dzdzoqN)Gu|2DwLVo!_baij(%jhjn>G|f%%mw&D@P74FEve>z
z&x6@}K`Nk7k|juMm_Z!Etha(Whc)<$h{<p28$bxW32#{EqujtBcafnTx^XZvv%Mwc
z)RiA<wpz-03rm|LaDIw`*<~TfXCUmQT9Td?s!nC3cnM%C6+gHKf5rKA3agv1A$#F(
zF~R;swjVjST0M^wv~w_+>k3#&)Wsg=`}oZ#etp!EG3zM6MM-Xs&vo+iK$R_nm&K>C
z=NuMO<>#8jF!PACB1<zf;)6o)DIthC-D*HLEU8v(Z2GXitlG8NYd|Eu9{rwUZNpn^
z&RMjB|5ATF`^(6yLz)Za|IeJHS+RV(d3(F$j_<quIcllTpPf-Af3{qC{y9tQ|IHig
zS@zU2T4!s!ad(HxB47u!AZ`cRCKVKZ?tAvmI}1*dor7v9C`mf|bYsZe@AH;wck;PD
zaikP05Ug%Dd-#-omNjgb!35flj!NqseW3jn;&cKO`dD^yyjqIN*p2{wal=6%sxFGc
zMTNJ5y`zkj2MH{{>bcIVqbxf0*Z=s&{-N!=C?$=zaovJ*l>}r3It7l#?V_<vk*0b`
zv;hjzKb+hH)xt{)TK4~r5LVskiN?wxL53B){UDQ!+Q0p8EQeqHW637MQw7|Xc_M-d
zg2F<+%|sD_PReN08&Aw%LTB~jY`spKlh}WYGmXR&$$0SCH%RFTy1JNFbTLi6m10gi
z!%V#5=rk=l{9QL-00-KDU@#hz&LWfCoL!I4nj@!fgG)W``;y~%#IL&Slz3rSm&4y!
z5jCTLPXw=Na2ZKH&to{1zjGr~0%UKF)!nV55`I#wy{USf_ISOc<@c1JRYbf}S?dAd
zij4sXf=!$%6-c8XMNG9sg~=b3E^M-;5)U))ID<;Yc76h8Tf>1F(C)f<lsr^CUO3rt
zb8P^S6n{@tu_cslcB)no=m4G+6`e=1g(XAlXOJd&HQrn<;HWg-6DJH))*<k!<KPu5
zd%!_ygWc%uO&`X-(9rmfoj-Fx1|uZ{i4I9p%;Q}3jj8V9`7}*sXak}VotW@teDN*{
z%ZdGgK1$FD9tZOgzc*M0x<|P5>w+%Rg37Evh|OG=<kA~Hi14tMk-;BWpc^7*0hPwS
zrWV1dC>@y|11WJZ>a-Id(4D?eY3>m;85oN~o<}ASy0l|+Qb``lG4J$9{pq5r?t3&{
zJI1GygCb=nS1>@~yXYuiWyat*<ZcQjcDFG)&7C@R3JNo5qkYQHdJtvxD_}TeWlDDn
zMMqT%&!n=#+CzQ+Gq=U;zVsqzc%W!IsT~v}?mTJ}D^7g{rMyGr-`>$OI~Eaf?onzv
zirc@uxN*gA`-y%lrXhcHzjsd$yE|taGI@PW>m%|%yaD##SAo3E<@`TbA9=&Xv`vLH
z?N=Tigt(f;=oQcEH;O~iJXO5^X=@O}8tFX)`<YD{?YL|2(9|aj%*eSn@1(25&AtRe
z0OE*7tD3emT1Wjx{zuZymE3WU(>u5+VKCjdecD<^bjf;SDsxUly=n)!eGnES*IaNC
zUz{uDz&+A8&SDibQIrRBV-w37<J0DLr<tO%1JH;7p0EH$tM5%F_bIv&6%q06th__L
zVq%&WXWn&vX4*%~<ci+F8&@!cXInHiKDMed>t9N47-WiYQ2c#Dd|F?NYL$m7x)>CW
zGV+W8-bJ*|A)0vP*qmCczOaeiM|PPsH2v9@B$wMU(y2)zm*6o@FRXjOifZlYiN|!h
z9K6&8?R2RKX#xBPguH4qfsGZBH1@^E1+byla82T9Z)?JkgATdhCcit>Ej8?(IWn7?
zJe!6bTf`4E`4}ayE=~;kl9b*IVX#>Ii}AJKqZlq6E%`lIe-QP(41%#(zJYX7IX!((
zI7pkPA%Ylb4DcV{80hu#NgTwmpbW)HcR9Z|tDs~Lyz`se^d18(ZDNH>arh$Nbsgxv
zF#~lGVn6bl@OxL85XuOY$ZSi`5DcFND5SD_wS`rdPpT|gj6R~-)G}jP8Y$ZWNv9~;
zYOyU97Nw4ozXVMzc3qdbSs;yk#e_J@j>NI8i%m^U<H}bO`;<ngh<t}=YgLZO|FE=m
zhl<4K$z6=#mD8<*W7lxugzC+v_S(KgmuO;*G=U(ZdNs(qM-ILO6duY*X+)W4hh~aK
zovBL5uPwfSE0v%6E()xA-l!@~X#aS&xDtL|*t+_Um2a<ja=z<Jnha-B(1mV%e(Ac-
za>;WD7fl4CM9;d#YJiACMcREYU!E$<q_|V#`&uap1ojNDb%=$LZ8HL<Ur7rgpizce
zBNG)z0O}>b7EL?_9gkuF<1|XXVuNXz-3qDi0ZlMfDVCi4xa7v29;J>u7ipVrO(cJb
zQP7~sL>Q<~qvX@W*Rk7{@g!5TpF)=$quZq@Z3gkWfnjx%$T7Rl9;Ov;j=;K>@-l8+
zF!1wuo?)VGcYXDajbf0yBNliKF5T%6lkS$QSBcO671Axbdd{%Nk%ex?X>lr|xftBZ
z0)!8!MLcGB`qg3@gZ9rad?zzYFhA5CwA2T?Gav!IicClr@JnXD0AF54ZdLJ>dIx)=
zzdwmXkeqTdMFBAnfK+jU-2|DUAp9BQ=rWv*?yDd3+Y@<-x=D)v_1`#=|JM?<<UK(_
zA|FLziJUhj?Vv$2GjQr#EnDW}&k#Fpulw`*>o-()y`bN5k}if<x?SmvmZY@ljUE`>
zRy|_ggKSO3nK=?~U-+AY9^=S<QqfEspo2FzzoxZsy}|Y@b5+Z_F}j9FqoVePgk{@e
zN(l>&d{e-txR%+kdy{O*^jDCZ<NeMBWdCia+>HI2mOht5C|(Gd#miE*k-+hYQqy)h
z>NLM`ZM90mlF_lcjcr*9GpBAG3Eete;Cxq8yla<fn}$q~Hs@4lc*~ITA$F&4ZF_lc
zJZGL|fL6+{0X<h)yrVV7!g>_1DENC!C!-&!GPNdFRqmq}*1GjA3KI@K{y_zqFKHp7
z*~d@7e@LH|tq@JV*i0s(<7CuH6BLQl_uuoI!=<ZLbf{QsNQ)u^FxybqUmH|BmpoJa
zn#dZ>v2c_;U;IWXJK!3Po5%SCS=Y@w_~^~SxY;Yt5Yz?Djn8OYAn+hQ-O}j*-0=dj
z{CvBOb=%OJ+CE*a^V}L_!jZw4LV@<^-Fpa<V4->fY`u;lpj;RloTh8yIa9)-6dLRn
zx|U7F+~EwvH+cHTE4-aHe5CvJ#(C0FD*{M97(CYa`YR)8hj^%X%1)I5kJ1w?t6x5d
z7z)cQP@La?)!XYk<~vSpBN-q!vHv`5ZTnr?vlN)IEwVg1GBU6nmNz52iTp88#c1MP
zu%vgaTdS@@9rzJu&=)TO&GtQW#wV^E(N`~go*GKTz)PEK*Q3ug<#5ir!#MX}>2tKL
zn*8vqa$3C-)>&Dad^T|01iy^0`}facMEdcUCR0YRFU_mH9WJ_IMIIJ*a)`ndnX#WY
zs@*zzV~S6OQO5ng^peupzb&Ua6OWYIm0~!319m*b89=TV{d~(pn!9Na%M^2O@y4Dq
zLi|sQ%cf5{*gsmP)G2p?PKCK!0sk0#$^Ol(N`L(zIPI=)uUhLiHcJhShO===5lt#l
zVL>ul%<Fh6s#aDr4ycgtn9($A)p&W15WZH2lUL4mh{-v(4?!O0Fr024+0I7)O5F&f
zI|^0f-Xx0vxOKC(S2+`EZ_9}<qG)_$rk7PfQpVz+ke=UwOd^d$)`F8%)r7jN2Gwqy
zTN_hD%;+TiXUNsw#}?&NI*%EtuI8VOfqJn(6D;j-5E^u1)T^H7*KLA)StP^S#A$|v
zz@3&$cDTiyIS&C3XX9|Hub^#><YKYk5?MFNx7P-%t>L^Fk0i4m&@k6~@+;SXoCYX3
zk|rD(7Jq2eg(Y_#`D|(@xj{_*5U+fM)4&+(eP+*^C9^a`XpbD#xay+YdZO)Z)N>D>
zi@hci2ARx3_CpCLAXd8^B`Uj>&fHk(c5!<RB{~;P?)tCMo+O~IKcHHba@Tk7M`3l6
z22+0v%If~&p*a`I#W=vHXxi|wuA#Zq#?RXa?$<415*nEdTHYvh_GH~fi!04PLdnRS
zZ!RX`gzt&*vfu}&q;($7nh%Q3){W@LB4W5vTw;jYcf0hg7-lU<CIdA}Ue9B6I(~*F
z(p6!~X5U3@Fjgg%Kh9>6ubeW1)5@)|d49mi-Vuie_351D`uO9naIeEkTjXFau?+Jy
z3BW9A-~3$Y*%_xks~+~gau?%CO4IzZ?hbR|FL_;66jyIX+Qw>gbpCSOxbj;rRIj5-
z{B_3`TA#ssbU`^ba7HXhdl$>{kGn44u)B4K;Avw#UQp?PC9Eeigm1Q0rNw0E!Lh}o
zelAM0k$~{&r(Y<LX{<~1pdJ#9w!T7EBICqV#!j$1h3rY1f5q2sM&IHa{@cbqq-mNM
zJU~H%kx86)FQQaHfQnz+uP?1ntdlz{irE4Y&|FiMs8!}y@6m?h9GdY$dGp<7CLA42
zzAf5Zo_vA_(*4IGCQ#6jR_<6NCE`>igRo-2Ns7ySDP6>oqsS`j=E43%|FvrTqjI(E
ztm?FwMb+#`o9eZ^y|_HM-&3h9;NF6k{8oYZxV@|5HOl9KgigkB@sF}Ju=&A*2cg%z
zjc5C{%u9Kf9|=3JlpyE`nbSlVK1M4zrFKGvT%Y~YrGd#<6l|io?0TA>#mh4gyQ<q?
zq3iFmc^v`yX<#PemDL%}UA2czEc_`aBHoJ`)%(55cfchO`$AM#ALvp6pU}W}{@rzT
zt*ecfCwwRL(b44Id;&$xC!rQy&%u_6D4g@K9usi9z>Y=SYfD-;w4Zr~k8aL{<pv?<
z*R-_<{6qWFXJ36>c+POny*-ol6&?R+9R98O$vv0N49m(ce7t6E(lx`#Exj)M{Mi@2
zCm-66o-%pL&mUxU`Nh!ox7R~Ac@}Ik&W~&J+>~M0*W|mf>VwrLc9wd5LPstMtukKj
z<z*bP?x*4~+nxI&*q6k2>B?woZuXFV5C(_2BwO(>F)^`;6?-Q8!ioL$Ps6Y<uN${(
z_wJy_sV7!XI?&mEKuz`4uxX9K<=z>;rY>B|*QM7CvEH7(eUWx<Xh<(-KMs_Y>ZiGI
z?A<2^ce!uBfAG%M-}XJ4k#J$^AJi?QMvr##@>+nmEnOg?F)=YlW5*t&gVli+gET;y
zEyL@MFB_(9RNm&_a8R#c`=^hsADo$y9@^dP&$r%uXX&zK2OhKJ|DoN*x21c2{`cq6
zsIrNy#0{YD#|7ul{}^wZ@ec(ZK1I)-{dRM7XyQSi{Kr`7p05u*v$KgZA-6Hi$KHi%
z>iku!=|ujUo7!5vFP=WTaOSFt^beaWeklURIj^DcUgePy=dE>ea_X>riz}6*g{5V1
zY46^>?Z<^CS|d@v>Eht9!}I+|Z|ZWrPp(|EW>!c@h%yo<!R;BrqdyhHvT`@Q2&>Bl
z1$#EVdG$%diEoSl^6q}uYh>W-0EM5QU#&_t?b)tY7sQdZj~{pa{AhkRJxp(}tTl0}
zaUCO(T~SDPhC3T$M|N~nRFK|k)srW40dADuSHmW(>N9jTCBj34^j*|bV)U&nDk+JA
z5*SkUv^2qv5loXWZn!XuUH7p@vas+I4-XIfQ#5MP4syM+GCV3O>dKv~`T33;H(CSc
z6jQa#v9k-?ajdo4?ECL8R@bA!vBP+-IAEyVmA<PT9j7MPosnbV=7ga`SDk(3S@N7j
z-M*PN;^)mRI@;O=SFY@e^w~WkFfg!yR3F%2wKw<r>?=CNgPg;zMo;V;URSr;?GX?3
zUoV?`eDIz=jJ5*hMmL||znEiX<&;zpYpyC80>x}rIKgSeuwmh-&X;Yio96_+6rtg|
z2cN5|dx$;Cg9bv>uff4zMqeM4HH-!n;NAfNx$1Vbs3=polc~3UZ0o?yDn+`8x_(e_
z9{KOT+>J>lqVX`DCbJlsu6DL#)xG(-;42;Pu9`M&XS!_yT}kK*;oik8lNbk*USn4!
zMGxLU^-UNYtVy+gFwZ~d@c#Xu)j!S4%fqH{hhxS$b2^v!@Cv29y*-KcAv~?!%y`45
zrY5vT;pAvTHk@(z>%|-*WxxrW_xmaIvYj*BpySGwa~vHVFVcfI-wWt6qJ>XaTkDvZ
zl*G9|D5j3)hq)FO78k#!ErrkY%blnGwV&TbQO`m}zhL`^f2pb%ZP&N)J|>QG-!yUS
z$KM!w?zd~#Lt3)9=cU|2LAIzjxDK+6;^N{Oh|aW)uh&FGH;2KDf#?rnsUC}qGb1fl
z=H3^0je|-CYUbxp_kQ!r99U!;eHe)*%q`2#&Mqn{iUhGuz3oC?!hdoq9Xoy6mMB1h
zYW2dq!w;K%eSPQ5o_$zy&iZGsH##~v6g+%*6e+=C+f%Nv_pH)m(b3<rM&^P>9-JO6
z#7$I`LB__{v$*_zPd`b3QQds;^RrVXwET@jzVDN=^71_YAI!|mm`*zO?8pwElG{$U
zv9WpndSh)(jp7B~XY;PzP^&$T4@1ai(km$=WBj%2@3f|{q=VY(!NX!>zq~fPgI4B<
zKOy>vzkEAJ&A;*2v<Cm&TTLSH?Lu=6`SO4K+qQRe8g2B(m6vUi&#(CB^vUtxF5L2;
Dh(!KW

diff --git a/articles/rxode2-speed_files/figure-html/unnamed-chunk-13-1.png b/articles/rxode2-speed_files/figure-html/unnamed-chunk-13-1.png
index 7b0c4132daeed2f01834b804d495cbe178421d97..8d741af84ace0e2ca6ed2c099df68e25c5dbaa35 100644
GIT binary patch
literal 65303
zcmeFZWmuM5*EM|Et=j}aLARJFs7RLyVvCfdih_iMbYrrWQf#FZ6eLAJIurqm?naSr
z>3+w&z~gzpzu$3u&#%wpdG34PTez<CTx+g5#~fqKb?4Lx=|%I|<}(<KMKXsE$}$+U
zEE$X$=jYDGzfqfBppO5Vec`C|LHsxV3>`>tz<<p%I(*iI!C31=|3A&WzWWk`v5p~g
zaG#u|XIqt(tz2KyMEAh4{`K?925#+L)o-_S=Kd<9;P$@$7Qdnok3W1!U!)q#BN-lb
zw_o9<Ja^c-y9(RPwni7muXtpqF}>_7$NPJBUqrtC=v>m_x$MD~Z5MZcZ5^`o%)Hqt
z%r~^g>dTMBMm`_St3qoQPh&n7gHh>LRy1qspBc#<)9JbCkI|W#iIe}tsC_fna_YxZ
zk1y?-KKZX5U;JD9_ZI~Zr7oEIVS~GlEK~njxndgc)IY{<{eStv_028Q84Jbu`T4_m
zezNa+ku`tbywhjSJP;8PQB_-<%R7Dg^xt)L8)S4ZUp}R+9pKD;U9_jCr{LZBw94@z
zMp$Ob@f$wj;o+lWV?zTM?)@1Uc!K}M^)@2nw7`q3^pzjCZr^?^D=SO)$`v!St%H*L
z_U-%AE4DVWsJQssPovn_S06WT*>c#+%S*(p<zetqVct)lJ|*|tHUBTX`Tu82`u~6M
z|5Dg8#LL>v5lKNoK}R;9^nO^RD}VH;v$BdxQFS$s%<gB${`B`hMmTO?<ve}HjFqca
zz3J*Qx^?T;*MbKQznBM(was#134^D&ygaC~QnkLJVVk6+#-T%pM2vsk)EB5Iu|F1B
z7Shyo*_CzI{teP@a#ydu&v04omztVtq9n{bdPe*>cum@BYHHuUeala_8!@d-omSu2
z_}w7Jr=aG1OohC<df1;oSFhA0S1Fx5d^q>_*9ZKUzO6X!Eq2!4-oCMSfO*^<$0=&Y
zd?O>z%vro9H^9d}NZh9Xi*B_2&+hZ*&);=-S3Gvi<=+h(RxVriuY-fb+3l-_1_!?&
z{TRdq`}rNXa^(u0AYDB@c{R1rZ{LnQdh(?8^PT1595Md>2RClqSRSSGKuk<*k4`>U
zmY1laNn`fwgvmT2co5&*czWNyYt;@Dc6NWhZr0V+ZRvaZ?%mGg&-chI-*hauG(z){
zr>E_SPsNWPuUffc#THRfm1obMiI_IsJ96ZR%NX8X#HiMB@fsmt$?<OY=OSi7JW0K>
zS<m>S)z#H&1_x%)S~>Q(BXwmX$Gpjr7&-Pr;EL>t6DQoqyuH2YHyX0Nf`v^Qw(Qz<
zdeehu&yw>IN_*@_EnPV!tCfPo!iv4cZK6yXPqbIY70mT&H&5tquL!Uk>EF6;-QFka
z)^Ee&TehWVWO(<)GxsGm?f|{ke?_`Q4OuVxN>aw<&CJZI`kx)!l=XDi0Yz2gTdTG!
z`8ql>o)oe!UHUROJXu*$!!BAxR5aRQ!rrdy<DByMmz<aOD~Bp1jrCVpkF>||Yi2A|
zvuTfhE@t(|YTip`1f>en_tf0O_ox^qy-peL3rDVN|G+CEqFgi9VbRxa=%BK6=~AAc
zcNX-b{CbbIvYx-T=}b~oi7_Zj=u8gw616}sJnYfmQ5BqIGq7#r#zQ3<4oM9^o^6OM
zYy0($`8|%yk?U=bhlYk)4^&x18`nuiM@5-d2%5FXRT-r%dM;?_tupf!Ry)4wm9uHy
zA(m$B{_uB=@mCf#b|p<MS>L}GY;YHQ_4)I`D$6?8_^wY2s(wutAl5dNpIYws745@~
zk`r$g)vXRKTC}LDU(L2VeYo4*AtXXGGd?ihkZ02-shAr31cT%utuee-!Ej%z*{{#@
zWMqwvBNRfEUjF`laffo$;WGgz%#5R%yRh&Iid@&qix)3m?rSO9wtais2P{IE#`vEn
zM`UD-PhV>E#`zp?O&LF0qHYr^F<3hzR4t)csB?7Gw#s2*?D_6Xcj@a8z;Ut4%#Xai
z2?tJSL0?~=_0L-@QJ20yHfY5+7Ol0Yd}&;l5f}KrCnqg!a)WtP0|#mxG;+iTJoEBw
zx@z}TUH!fze)P|Gvx4J-TX*ivO)p}ezMWD3bWKlXc~{YHJux;At(C>bFDS^~8W9np
z9B=wIzo6jK?q{k>=O+-l@qfN?n>D|F(&A-No#)J(95_5|jVyGfI`J&+=W)R*f9|l@
z>qC=E+q=B_+Kj!N+qk*SYU~D@-v)T^{-nmq$+<&4QHfp9pz!*v1-FmI+HB(G<yEt&
zGP7><vgq~W96Rjo9U#dar5tKK+#C>n@#``ql!VwLlkYzp??1REp(^b73xQ^ql)+nz
zIVJQ%a&V~W6JH&2{@vc*{%QW4*+>&r_9N}qJz0YM+SzL~#RUZwY5$5t&ic8zEoM=e
zOCOk8oB7{th7dAHs6wdn>#>mmcD*#GB_;1KeK#$;^kmUjw<_KT_wM;__OX@qvHf$0
z!(r4|UQrSGO|LG)wLC)8=`oL5P<LHsLi{o21J2$kf^eE>D~~>!GF%iMe=FgM>eV{}
z29<%g7O*Vqu%GtetG}ULs|qtXI~^uQG8}@DAgk;Ks;tL{8k-xK)uUEyUG1!$zsvLm
zLe-O$zdzws1Ss(@sGCE9Rx{oo1Ds-Pe5)hvLPS$oLPvgiwBDnK4=;wK^kj?1*FRQy
z_~J$DrxxY|vDzZO7wW8BxsrBM;7a+$kmJvGA7U5K&;4|3ftu1Yn-@FJUsrb+GaLK8
zS@M3F?eN<-QBiV7pYRmjUc`Perb111J@bQZ%$UaOxL&U$Of9~zSmlwA53kJb7g@vU
zy!I~}eH_$JpFYhKmOW-=W7GOr#USq}Z<7DuprwqgqT-X{P?dm*@y-baGqY%C?wRyy
zsbfHTtXJF26V$EhSMJbARs~G*@XyUXB|g#?IXsl(U|{O`tMaLk@vHIP0LkWZy>R~G
z-Me?2e%}#uxiQ;|CuAY>p$4SyPCIuvqD<eq>7`R|2D>4vUcsB=1?+~2FLesu#Jzj}
zF)JtMB#uFfLYV5?)YQE)9>`VU4r6-a1C=K4%k+7|{+awB)PZw{%RXFrfkdP?IwUPQ
zd|x44J?mECu#B~Jynudb?v7&)NIqsIY8G#Ug1+5)#LP^!-n6`1ckI}akwe+v44LwK
zT3Xtzm5aoY_^kWOw9OEL-!JZ#@wjSYV%8cOW4@qfGI>7PfK9%5>-O#V_?zp$>kG6t
zyf7?|%#qL=6;}#X5Ff0)p`2o$gf;rRXyQ1i3{|b+-sa%~pK<@UX9C#t!c)fM#~U5S
zr4@pU{yq~2A!?qLY-~YA!HTU|l|98&FDeXE)KK2TxF^?6a4&MVk?_frCwH8Ee=r3}
z4F_$rOi*B8uE)lsQFbHdEQ)iOo0UrCnRO>m-0x+fDlsRFjg3A0KYS1nAMHxBZjUZ`
z`ke{S_wv%$&iEF1+OM}II9#{Ctt<okmh@&Nw?bfOXx7xz?d+nrnC*$p5Fh&eVEE?}
zhsz-V2x=D9Rskox#hO2TW9|s667%Cc%JhpVyJ<`em;Cwj$Ab$MxamWM!I={$+?<1%
zrS;r7oEFBy%Zh4dkGvhmt!~}CdDw%Ejm;F>)r$2rF}}q7#8?hgk#ngp5L6U(ul@ZG
zd*}<i;IQjGeWmo>zQs;y=|z*9Dni{!%s*?_hVttbU#?EH$|#RFYmL$^+J+;^qCia;
z!?Bg#qw;q8g}iT%xchE!j;1P&bfqontG%(LrSR>Uzad&e4;Q-<P)f8j@yz$fJ09SS
z>9pm(JP_X+mLMlDU;k_Jq&8_VztyzLtVE|MNd8Yx&jYTNEG&9oJvM&({{8Pu^l~Bt
zI#xUvwFpzUZr*sy$?2#^OKYnc-c>KhTbw71xw($x`)G6ZbaZr-V+_2cT-i3t=o%W9
z{HgyQZPla2qR4!O)FEm5DVzKE?<+@YxoTy5Zk0hHv~GDD5M@v$BD`kt-v}6@I?uWS
zjZWY12Pa#;afk0I#+mSa#H|<<8Cfv3wMOI2tzCqKJ~mvEQjB!$674QAe0Yg??>!(>
zHKo<ejcA-h8*ysavt!DMmJuB(6ZX%t0(=_!J8QyiI^v{J-FU(lGuPO}8K7zB7Q!9Y
zzwT^iHe=s%a!L@Oa_Z_Gza~RP$bv%UM)BxSPhVSPj_zRBPY?f~plv(U;^oT9%Kom2
z$qo9jz~i1mr;*0w0V%oOzI`i$EhzxJQdOONxSBQeaJfg19V?`?`uzDenU<!eT&!H*
z$2pv_vAdbCkh~h(kb3VQz-kr^+tYm7*$=r;2}6~mj~FI)-(IuWX(~8tTh~)?u1>qX
zsNlg#iMJ>NIJAj>?s{7_tj2FlMN?`CbA@FS6BBz8M6I{k4R#h++jNSH4>$Qyo|;T}
zj^S?f4!@sfY0=9aeY|~F@rNrFZ$>&(H2x<6bQRG)+zvq+Jnb&_=RYxV@r^R8SFesA
z=}3@2bEfX^vWX(#9M`x1`OaH{N89S0{;$k79-d-aTvb(IFXo#LUS<B2Aqpp9jun95
zGC^<O=AmT2_LFvx|GOZpST=@-lsiZ06Ak;b-QC?D{y-+p$s-*Fc*4J{Yloc2sGazn
zk)hjO9?d3d9>VoiD^e%@?n=Vw&fLrc?Q20N!E?QkkW!B1L?TaO*M1p~w4b>xMZwYk
zkGR~+{6SkcA0Hpe?|NZz8A=uJ&foC1{59`?PU_@O^?oWSII~1z(AT~{(r2|Er^9Fj
zGjAew|6N=MdN|)Dn;yudjMnp-0S$fEml%6!Mc4^>A8&1JY{py^THmhMVMzeU*zx4#
zWDoyOpZ3$1P=ejTJcB!b6PsbQil9AJ$d4aCdiX!se6l55w6pN#0jH@YIk$zmB->rt
z1yAs%jJ{vCY#Emf9%|TOqI06T&Rx=cGJQF|B<{m8d>s$Z350JE5bj19a0k|--I;yE
zt>NE^Q4pc0OFc!gl&ZL5#R`9IA3FsWHPibMAMD4h=~+QcOr6%-YzSdS@@ycovHqBp
zq9qQ42Q=H$mPi)PSR%GAxTs|+5g(XGJFD?vDiITS*Xn5wC`3zmxjFwe>V?tdsyM!M
zq$f0NvZ)#I9`ACJ{Bj}Uum?Ir)6BINO=zR04&u&I=E5xgfewi`sn_G*fB)r@Iez?j
z{6Mu$U!F8)++;|lo|B^25cYWfJP6H+yk@4`BQET!c=zY!l+>s&P8NWLQp|WcfL@V8
z)U-RDg<X#j&fS&Vh2`acU*yJ8dXdJb0NormZM%ThzbI8C+a+`)_MDKHmydhLJld(}
zn5$=XPT)nfW%mUta@0w4p@VLrx+5iZ`|sDTVt%bOwNvVrKWCwm9~mA<FBu;nNJ;qn
z4pd}Qtpw1^g=(Z!#||n35P91V>xlOio?+q-W}>UjrVQ3Jhadx<-Yk8kBCFmDLBnDI
zQ8^Wr)~T)NaG?dZ|5aD#S2NlmsQmuoogWPaCqJG&dzJ~xo;_=xDkdHOF@N&n?q{ZT
zu7WM@4x@fx0XE7Yx5hVm+eGOUY+|N&CJHv%OfM3fiVf*Q#oo%#FQ4BWml&)VUX1t)
zL#dyt14b;&A3TiISsbPo92^$^O=$$(<!jN7*aNR$zn(fswLJ7TM(Z^+Gzi*ZTf><t
zC_Ozr9^GVLbAV6$RH*Ky3gqaF#?-FUrltTL;)<0kQLC<8DbjQoe-q$sePr$0wNof<
z=P3%DoqJtyR0#K=Nz@EM^YQHdB8q!Q3FcorF65L@br?+JGzHlDh6B7uO-U)R!XV)|
zr~Tkd%FL4(bRieL%)*^ok!J;~OdqJ*b{{%)@Ze#O^78VitMy0FzNVZpF^T;9(<_)i
zed@uc<M{+Hhk6>>^;ACS=lx1|euQ<O+MyY<D8^=#cKu9?#)s@s`Eabdx|&&ym>9qQ
zkz!$4S@W;?OrQ})>_aL;8@q%A(_EyeQ3UkFC1akTMpc9Y6SzJ$c0cp=f+_3NZt{5W
zAR33+ZepZHu$V9<VPcdQRBoP-$!MC-+@__5hK7{4hU+*d0zv-Bcw}i66r7X{ePNh<
z_R^(GQ>&Wm!`!U{EUHG=#zz|^Tkmp*?JQorR>Z$GJVo6wsc%&k)``K`d64=2ZxBa)
zRH7_WM!c@yxUo?N9TyYl5;BgBElpp^Y%8+`DZNxy9%51R7p$>8^F+=3S;XT<k6x=7
z#2pGx>VFRCwP5o6dY)vi?Q~>F>)t%+zQH<oZZM}FSgE(@jE_6B@G8);o7^|gCCs1x
zw`RCFCJI!*Sf<0MQ-p%Dau9)slVaUBf&EmK*3vWAs#0=2cW>U@xw*({zAWmN`|jMi
zlQAsKY3J`V)}~EWCq6!2RcZa?k5jQecaIpEYRkGsV=am)9x*X7T+IM<%1PGI21$Ji
z83Od9sq2_0dtF0&d#De3l=4_(KV*8n5EkSKZ<~%h<Q@WdpC)&@W+uhOfxY_rA@<m~
z)`66Xq+h>&x%2^0w`fa_JX5zHx`e(v;T?0YQ%eZHrT%k%_p?2Q)nZ9Q4KL_`GXXlX
z#&D0%M2OGm@6Gi1SVw1WX4~kwgO>a@EA6{!!Et6o?bT^3je_P;SXk&!-2h0AxDWbe
z|0Yy*{5rx6LXF#$D%OLaI7a1>TEQ%D@d~BzkdQ)Z+jP5srBkPild~Y0*}d5=Wqwa=
zVWFJ(SZ{%K!!zxNj~{D=u&i9!ESNInYLGH!j_U60%jDQ*ucMz{CqwK{akxe}mR@e&
zym=Yc>gwK)c{Rf5WLCr&YMU}^|H5^&yQ#>&W@x~5&}CC=w_^En&9?6Veke~N8rYU$
z#7I$5QT?aMcUVTN+XOJ|l#D*6wLPY$_{I{6(j84ZfM+O&9dcgPTxC%sucsHhfVo(X
zFJ1B4jNY$61vNw27QLv-P4w^{9v&`sC<oRZagCHab+exHi&*{f|NCoq=+|twE?Kff
z&8q$pb0;=?TkfyRbQkQ$1`;p8<c*{`tPTQmsorDerr|pQ!PF&nr0EOjeY`&F$jOr>
zzyMJ=$XBZ3<rh00z?rZ3aD}kao(rGns+kqZ#}EE;z7(OPs>Tesh03%rwcLEqjuA2R
z=Ge1W9v;i6ap_DN@K#n<=8^&SpuLC!g_GqYsX=uKcx=amufK^$b0mNo9Uc8PH1s%*
zY({Zj-plOl6Zn@_R#smNdQA0m^YalGJdtJF@phU%2z*;xTf*Ful9EG1Ltph9zJFp&
zri)ZT56~jupL7r%!y`2lVRXik;F+Bv5s1U+gC0J88hDf^?jRs?#Qy#JRaN;Jj2q5X
z21I9oXGLWTj*k~9Dk%wm`Enm<H~c$UO-;Y&FJ73n6mE_F@WDIg{t*cY2`DRqAJ8z)
znmKa|Kp_|!#Z#y5A^Ct1eg)EC5xaoPl55jy*{RoJG~P@}NeN3Cw<cDpJjEfU+PZbe
zm9qDTfR;S`fBm`;dNxvupcbtmSS+PT?VQJekj~t57^&h>QBi@3iK?4Vc%C|R=+?Gv
z+vFxW<)XsEFItgt@dssg$6wRa)5~eG1abnQzV+XK_y7F)^RA1F+{KIjuU@@E1$%w(
z-aqlzr`0Z?l<8>l?c#17bnVY2?Bda@Q3p2G7w#Yr*4e*ct<=asXUX{3NIk0h0-o5t
zJ&idfsKoX7^Q#4AH0jNoO`vby1_mAm6S!^b*3&vVZ}BlH-I_Mm*4w}UkB^VD3Ftos
z({F2MSKr#ojR(V?uC8MN>Fsi^$ROsZ2a+76Q`&HSfek-EnKR`AW#z)k%I(2LLH+&x
zkwI^hlGO0E+jx09%~GB`d2->xg%xOmV`GItE~7O#0a(u_VJlRX!niF)r`o7`=A1dy
ztmH1S@0xK*ORFVmpt?YEV(26^H5rky+a0JkMIiJcElA=Z9qnnzDggNS-P+npFDH}&
z6o)l_=;IT9{rdE^lJ=^zX3Yu;4jwt;t|{<>RExmCKm)ILFJHbSZO+K(-7b!gB_%O6
z)APXMoefps7Z3<S@Br7~x6Y`kd65cm=8RuUi!O*dho8a~`U2>p**ExRulVO5UvN?z
zH*I25k(Kqhe*JoO<J)Pej{A{n8QG#>=0Kt9RK^;|gGP%6|B*B?J~E7kvM?`iJrk0E
z27Gl=V(1{4X4*@w9PiL)eA<6DEkciq1iNMPW@+@B-?Flxdw3gP%jd1DufH_&8CLDC
zvol}(*$o>uxH`XF#05w&f6*fO%a@y;a8%^wop*U*<mS7Dmsj!Z*=NAZ!wVfl6@qv3
z@c~$rU9m$u%eL#nO(dp0x`kU@+4zs(18r62ruy#b@<s|z^=+uHe?4o#s>?_$x;i>}
z0R3-}%J(Lix9|R>4#fc^yDnHWh(bq>A1^|1@tZUp!w20Jwq-a@3y1EK>CUOBrS-;b
zt>~M!HhsjpZSxo~8~@echc|BCtoQrN{ft%1m-7YsOe7{Hx#q?TqC)|(%O_&x?Ck8i
zYZa(@aW_MK{lcGVPG*B$=QXoHLnF%{IdMYZ!~G4?SH&3&15!X3u`Xat>Yr$YedkRf
z)~_6#8w%2^_KFyG!(tSft5>f+eDJ`pueC%EZ~k21%5yrWThF}NvVHrHbnR)3)aJ#7
zAtBo@UA#z5#kU_nM)OxkK~6ey^k@NzHIE;I!Q2DY>cU?0y7=em7f$Qy2Fac|lMq_b
zZjQ|U*4H<c)#T?ItDmE_rEi0SkKz!c`V~5H+HYy@jd4L;eu$pq;lmZL-@MVqtGnEt
zUaMePjPyO@6SUtbqo4bCsK%ZGFfC8AiTPkq6?CQI!`t}ya-lFtzaXkodO<NB^>~OB
z`SrnOy?lR}=O9@DoYc(bqtt%ZirgwHTG6yDD?9sfmR4(zCD<Y82gjf8+=n_+5^G$C
z8t~t~sz()^KeQrW`T4EHDO|vtn5z_1(Y|fR4tWCugP3-RX;mGaMnZCcKKc2wc4!1M
zG&gZ^vA!H>%JX~l^y!u*OWf_n1sOsr6TOdF|9SzajGRmD?(Z*7wCW*rfkps_UI*2u
zJkBJ52zAx3DjvpW-Qd_|JfvJ$UxMJ)8y)PXb{)-{su>%@ao=ZXZ%9cb*nPY)=h6N9
zd430akF9b(L<pI5KI2~psi$4#{F;-Sqa5#HoEHsRxKJrPIGC?towl|%H#hfD<k=>3
zjm8?Y)2Hu`gZDimGFHwdEPNK!g20Eat}gq8b88usfMchxuy)rQY)Rg$L!|c+U!;~`
zt~1b4g{;_Yp4hXf4BhEwnfVJAP$2MY{a{5?vUTIeTMI&P!jIV5CDHH9n6tPk_1X-G
z`b|Bhs4B<3M2c89V9SqKbTGhgyef;-)+6>(8D;AI`}Z!+(a3DjS%WIZ;0&-<{jr0K
z>+@WjOo*Wf5-CxJ(L73qwF=<J9zo&h?KPE=1=&rcD<lQIiSaR<l~!}x<HubkM%wu7
zzzTbe+jX;PU-tCycy;sU%w+R<j8o_g@oiBSo$BpX35v*rb{u+;0k4$rA<?cN;H2W!
ztGlm-y_mmnVPVMG$US<++ciT!d?<eU+P1gpG(95dtctOgmX;{Z%*E}c5eM>{!+!ev
zbGfb+EuL%V=jYe_XO>YlHA#$^RgnF-!xLp8c)jLo#(s%dcKKrOLvGAjf@Gaiao7p9
zLUXat1JnU2w>581`$=uZzR1YPK=pjJU=HK9<L%qC=}B$++w!i>n1e!6*=me#ibPX9
zpXtY0ESK9#BUBe&`DekRMMYre_#MV=WhT6fZ1B9A-@kr6jaN42gqnk>ZyoHe8?LAs
z-z0+y6H7vKo}aWHejh2TzHgSJiJSHYG$DaEot#?Vf9EN<^h9lw3}rhAbPw;}KcBch
zI5@b~oD=EJ({oMz@87I?z`eV6?ZV^c|F?9*!-o(3q5mC{mM&;tz%cmn(<so_cW;74
zXK_bW!lM^2xSi7SS|jlD86`sjCw+JvAcTgknsMvvMcMuPuY;9wS@Ni&qT-R4S16bd
z)^NV;#GN6g!@u4k%jSovC2-}{FQ^@50+1WXIe9Lt`Sw6El?kXRRg{<C%EwpmH0B9@
zIU^&(V8BFZ4T`Lyl~o+@o>qvBmDN^MSL#gIlTf14KY!j?!NQmwh#I$6#Pl$I1t?0n
zGiM4~jZr+9it&$Uen<(YkYr~XHa&26ci(jUxxdVVO{?{g<Gy|QvU7h8QaP%<rleXj
zBYO^o=uZEq$2I{bZr!%+2+>0R4o=B&adF?WvlURWz^_~X=y`O)Vr12-RYC9Gy{m{X
z(^oJxjY39T!NL+yTB@{RaM8SZ>(_yVD9+9Gg-U-Jf)HV>@`?%@aUMV+i0eg_aV8`e
z12guSyk<Q{%KhrqZwU$JH!op%9~tSIJMiWH2EgizF(bXj^`gjj1}_bd9a|W_k<rUU
z*bA++ZzpG;waIVpUQETw-aa|+iV2bs#|937CgE(inyLf~i;Bk9*5Id4SqYyA2q@6@
zty;6@2fF}6NE(%tshQ&-peiT1sqcJtj7`wM6Owp<RZj!u1qUbNYD*g%y<ea2Jcle`
z5c_ksjCM)A%BfSnfFomLoeod92z2h)aU2P$7-^NnL@B9hI|<k^dOu!>xcuqnJY@I%
zF?|^sQUEMLh$^fwK#lDA^DlAkZ5AyDrQjl{>(dB?*VNQxpaTAcaN^RXA4}&j2B!o1
z3kwTVk7eFg`VzaK?%vbaS2T-7)fY5_UPomduupU2k!ubkuXr@HyGDl2CMyKXkMy?}
zz;ode4Sa??xB8=@tn9LIcE|NhlT<A?ouf@By+crfEL6}FDnhBoVF0(QKdLG#`wFR4
zoijKoscP;W$sBzFwA??Mc+3s*^78K9y(@d_)IVz4jN6_Rf?}4ZM9tfBaZrkKb8}mb
zfyStTp%YEpJUG~E@;bFlIP!Z5s-Uv8J4X>pzUfp_Lf)wu!(PZKD7<l(99PP3Zk_3}
zMpzcw1dg`6vhwo_nwpmcUX1kFGI}K`N*zr;XJ$$xnc?r-Bf`Pa6YxdZ1MtNMxl{}f
z536(Lpa9|9&;^PBrMQpbT-<eYQ#g3=CKb%pt953kLf8N>E)|xPW*FT9jBc+^QlYbi
zRU7sU5XG<UP<pqwxw)B*Pix`FRh9J1Jn>Ec)>%(OuK?AB*giK7@gQt|kiWlwg&msh
zNqnLi0v8G-B-ZegfEFvUm2*C)t|RJ@k-G$yo`2H4k(`;_(4YlB3cHsOD3=Ven!9)B
z!;8^3{1os(D!|AP9y6ElYu>A8ii%zvV%zL^E#%d!{h}5fh1p)BW^JXr%G&!mGQszg
zX9iD-byrkX^<1!RR<<OhJJ;2GQkoeqd7ybVH}}rr%ye5@fV}Bikz)}R6BX429!kV^
z@Lj|zU|-ZK&7&tyC|tbA^^&(XgIFJis($Ia^LFOr=4NJ4HID!Xko#xpsK`WR#jG07
z5QSj;AEYRoD$ZrgF5FDXt_Td=0t8J3*VBuW;kgLe8}1>r`$zH6Z9Qq!e1Jbt!LyPt
z11NhqPWIfnxhPs~Ws&?g{rbe^Q*T9J^T&^4YemhEfxsl=h`A&>k*V^Wp}cpmBZRBh
z07e8s3j<HPSTHK*Qm3EFX!!Z_B^~?EpL^p|qNC-}AXQDTUB3AQtE8Zy;J{ER)|eUs
z1qB6nJGd_9FIm!A+-=j4Gi}<m#<n%n7OJ5IrjiGu;x#}-pYTaRF0O-UPtHK@g-In{
zOwE^N(}S;;Sw=8C0F``{0YV`(kSe%&b3mq;1|Ic*)8b|bc^S`<eW^ww>UHs)IApSD
zFwl&R^d&NEg(;>Pf(?z0+KEHMVoqu6gNw>H^qN&G$jS9u-e{!03{fGE{}o?@KXYv~
zFI;e82*E^yzVc63mry|CymMWF5M59NqHG2X)(RQz6?hR^QnzgRa@V{CD7d~4{uyKg
z#_AUN85A7mVLDu~4*&v(Mb_N>0}}3|2M-D~KbYCrB;Xa|?fz+7c;Cytd7<62)&;Sv
zm`l12HjfpotbqVT2&y-k_SjhDEp=(hhfki+L0jzPSCT0K+(qHAcB1y@&+|V<#7&&<
z+&OJ(y2nd0^&icc3PgGU^0FQ<nFvlmc;+)iE(A9;wTf6+Jnsmi;WB{T&W28PlPrFZ
z@v7$5g*IL<igeq}HJsAUC#sjmn*~Q0L@2o|J-I>3WhWZ{RG`>%flNmz>nptFr?j+{
z92{XNc73HapfO<i!g{KYW?5B~3_68{h4HqHFPcApck+H~twHwX%l+VV5^;3gmnvLm
z(pjBUo?sEq#?Eddo(=!R^#weq@MI)1H8fA>xvCY=`AAWHYqRN7m9uBx5|n^34(1Uw
z)YKOnPdtA7n4>B=zqxUA=?1AS>(}qYL!P>D;T0ePYu`*yO~TiVrbl2;@Ndup<ZAku
zj2=K|7vZ&fTtQyGuWYOWwG-e8kgdN|5GT;!Ct9bFrHnaq<}4qdKI37PK?xvtv|)`z
zds(D3Y6H04Bo2C!IFm-*@zEgy0BFQU;*=TMX%}~|=KRc~^e$D&(7-?eg%YH&t6|UY
z#>Ro+kA?>h9B?ng<JOi8v2DQ9QZoue-Q~+$a{5wBh{S}b^h0?3JfIc;zP|fdc1I20
zj9R{!WUUDSK|yX_-jjHD@&2LVVYc(k=bV}O`STVouA4)idnJRz!%HaHH|4Ge;`Gbm
zQA007O%JLTiPY#?=Nd5tk%h@n_qHO&!LuqEzyQ$^8-Y{XH|*It`i{^*Rq@#9=n78G
z@YvW`0~LM!?JSBjlNl77jO;kxmhtF{b^zR6pQ_R8JbLu#Ev$XWsCI7lm-Y0{pwt5O
z+l<+1&ra4hfDm*uB~>`LxcCh+cMHHij{3Rq30@I+11@KK3P0!5cH@byo!($H{YK@p
zs$`sdzT$-Vc!<h`MT5~HmZD|1S!A!GqT*`&j}Hq6hr!JxZ39%4KfPSlpK>!Mh&*jl
z1m1;cbN9dhazPO%v;)}b>g(_Mq~6#tYJ#N?cP~SHzIyd9y=rM=eSOiD>~IrjS66u@
zrKezLvNLt28|0uL6cj8Ov^G;vcnlAojd-!uwLRK_FP=aDHiw(>8z;@Sb|lo$?mM&p
z18_tt!t1&=cEK^s!D4Pg1SPx^Zvo!TRRmvTfsESYU6GUcP7mk}nsRol@bnqem}@zE
zA&3rN-*re2I)FXm{VJgfes!o7a=;K$gDQ%OC9Rs}`T`Ih0#WiZB~WGRj{kh-$4O-s
zb?(Lfg<Az)97jx`?*8YtM)(z!IWR{_Idz(vnr4Lh(L6ohrFk0$qGI===|<XGk!4R4
zY_u}Nk@(Ss0bT%vgn)^oFF_NDvV*RrrCZ57p(-}E6=YaKJ9tjK;u$<Ypi=y-6c~(D
zD_7o_&3uVUsl9vm`j|MVC@aU$O3r4);_<;aB}u5mrj5F7lk{&apsS^&WsKmB>;)+R
zWM6Y~Iz4?jbA4b$AP`T10Xzr#;32i(gUEZ-W<_dc<)e97!OmWRM`xXXY|{fUm*@)f
z>r^BF3aC7GzYI0u1{^{7g|k?g39XT8>}+ftJK_PVK~C=BB>Vth7ibUzX}Gan^_mG2
z(=u*c$4>z8<mcyGx51oo8jBcXm6F{R1A49`GZGOV6eJ7y?jC;QNI*Dad?@@{2F@DL
zI=thiGtcie7Ft6;f;d4QHw8QZ;{0Y1GhoH(oK76cWi(E-^r%0q35?WB;BZDfJ1Iv3
zRvJW(dBsq}9o?fSGB%tl%n~aIE>-YT^_MS)p?%?dzOnMBJ0-DhVS<C5O_LaB%f^kb
zGn2s(a{LVMaXpXM_L=syXMQcJHqyf2AnV$FKAj7(tFJc9ZS;Z>*}wp<-#-tr%580J
z652|lLn9-4e|~?ZwDsori#I?he12?>Dqa2k3MiELW8K!;(t<aCw5z0M^-Lcj<Nb&U
zvf}m_v$L}om6is2dU~o4%$Yr#?WG~`Pxs4kFt$_&SXsZ!Np4~alML2;5Y~GD>s;3e
zzt5CFD~N|OV6hc7X>Ffho8O#UDlGa=Fx$<SFX6hUr1ZgqUn(j_JiRx}jsZA?xWcZ2
zE(pjyxz`?t8cx^u5r7^AFBhE4?MB7YYR~Z$j?A<VF4Dc3MC%v<I=yeG6xwx0%Eu&R
z*3*`jv0zACD*pR#E=3b|jnD=Y0QfEeZD?n|wh*`#(7VLV_x=0b;JycjVl|4FLae`-
zIL~x;s&?|e->-Hacz~cZFB(D@f*tM`)*L<t6hjObbbtQypJq}~#h<ZL{EUVA05x`Q
zmRHq#L*Kub#lmHzS7uj25*W5^>+d(OGyC)mNAV4$G}om|gtVD^>lpq~Ux0mmTO;qx
z4PM@@jSFpZ+7@1|wm&U>57i3%Quv4$wzUQcW(8PSknGsQ<<k@1LJii`KMW*}{OG#$
zU%RaPVh<*J<P2B#0Af(7XFqz>k8cel`4u1UaB;;f^|gzgLKO63Qe)n~&yNia2?3|B
z<~{<sBp6+0N#@Mt#mptD6?><&9h9L0Ft)p0w%fYK5=k$^`CgI~5CCj;66P9f_i00q
zSGasRaOvRf+qbuYiit%}O4RW7RSWhRNg@RkJA`8*clNB#Xp(5!L}c?9GvKIPR2dk0
zf)IW<zSPk@PQM`RuFYWFJ25Bq+O=z7R@$OxD2doj6KYHufb94<?PpWtwp}Zj-5z7C
z?T4E$3iCiIShqrSLZQn8VFiqB58lvpjNrV{=pj_)m21|7)YhIuf;CH6D`9IHvjZGG
z_P(a6(&>6}u>v|4tfB%;`*I+Jh_9BHe+|15ldo2^9$5xz25ie}&Naxtmp}iv)U;tb
z5HEyAq6RWETj8!Nh6_|s@yX47r1LWvtmlxYmd5w@^<|-7!w0dy#4&wMB>^AoC1xdy
zFC|(%|HOK2?bGt|MQ4k&B9y?0pPQMS4H4;lmHh{$?Lf(jhY$Zp!2<FJ&@HBQ)rT^D
zVp(7|G=(Mu#+^uae*V>w4{<Q%H8jEjWay1>67(jhSU?2Y3L87fjK8ck4kQ$swl&^x
zSg~%OYgy$vWf%wWe*5N)6oMhX&P_Dx+{f$op_*myFCl{$fF8D0P`s|Mr&oYZC|Wy*
z<C3=acim{xDM2yH0<$64)tU2-rvy21SRa;lTayRF78c*OcmzuVl3*X4iipzIo~L*l
zdPbmFJi3%%GXRRbii)1;<Km(s3=#MgbzVU>K6>=~**TxIVQc-<*Y_y>!hOJcQdA}t
zFj!l^S5&Bw90flJSTWSq-_6Zi;dZmKwEO}ck%NN+6njRf$%7Mf>Lw*U{3qTHBT+4F
zZE~nh$nzH?%Lwt~V1k=5Gc^r_=mI$g{SUu^S0ms;_oJMi?=xr5-i8JMF6_Gi*vNP=
zeL%+uK>LMn^X#dahY2v=>FR=bgV*WKEW!B6ig@mZxYG9i$oSeDTqJlwtGMVcxHNmU
z>BL~BB$3*1fSpA5-so33?c>m<y5J6ou0kqEEg83JlmCM<2l;K6LTcF^W$Y}pIsQF+
zSl0$+yA?$)x+ytf<N2}q-0!wT^K6O52o-e8MC~A4F(q<HN-F;j2`L!H5`6sPMKI_P
zBD6_UMpm9}Tha;-Wp96l;ZcuO@TSA8C4~g`r?(eyR}ltmngAf5&q>L#DCoi+<=26*
zZZjgw^j&VU^6M5p;pkkb>vi4`r0nzO&t1A_WrxPaoeenQC5yQiNQSOPF|S`A!b*-A
zX=%8Bz371g0YWz63Jx|=A!y3s9rcGm!HrH2nVBRQupCt}42Eyb(ysVjjkow|?f|wG
z7R8===r68D1NWsl@{n>9+6TJ2PT_@V(n9U(*NrqE!TqgGkXs?d$P9bI&~V0f?$0SL
zEe-Eu1*3xpsrrCT*1UK<OY008uU+aNwIWGw0Ez&fugcijtKehOgSwI-DwHWMl(|nx
zdjB9mw%-O%$ro7-hQmJiTwjfUy5e7A)Pp>78705S;F{E%f7IiI!y-dy%^oM*jlO*O
zk_)aG9Q@1ja^+dGXX`?77>1`S6bVGa!3pgI{<^O_lam{b46qy4opcqR<iQKbCr3R{
ztfTbHcGunNgN45^)OqmQ9FE-H=7M@7Au_$n<ioBo#__YG5yid<cEQew4*U(;lQCFJ
zPNBnuiU&+XmU$hZYdiCuyzuw~@>-Aem5`Ksr84$3=&Cc=sW&IQg2-5cuIZksbyzp>
zK$Kal0w$Ssdz<p&f8J)_W8K@tYuAR7JXX)^kXM~#gTH(;Z~3Ms*xVG+T)U6hJrGVn
zo+(eXiln(4j3EHwHGf*bTL=$LsH3A}c&`%nBn&nJDCieSM4yj7M`c5;*?}apZN0^D
z?;w&xtF);qrX)!wRZhF!`}P<>Xtt>tQFkAKE)K(h9a`v?59*U>l{pH7H*#Iq{i9}V
z=&XLWRVs3KP*~VwE;aPB+BQK|rsn1?FP%7KrKF@{`T)mUYCqmUMJtAzX%i=>563_|
za1Lm1QjOrpuy}DM5lSQZpMw=b!ISvHVp43Nv04nQ4bkzG+F)sK?TLtiNCqEaaz!%H
zuJHOQM_s<3JW|!!FXR(0qFwUSs0p%KF@|a?t|U#(J;9V>KafxZ;(mx*oChm{Kb6a;
zg|ibz%sWpm3a0BEt5#JP3lvC(B{V>a;9k&Y6I4b}!R6aR-X!ohHnyQ1U_6Leu(eI>
zNE%QF1@lE;AQ&1VnzCUSgIrIPfoZx@9wkGLZjc3iunYH=yFP#NWU~y?@klw#BTJYo
zK-#}VnH29zox#Sh`w#^Tf7#dM=T2j2WaHtmti2||S@%raJ1@FKJ$Pa?XW}80eUK8p
z_{7WT*{s_lGAWlrgl!3s9D570sdu6*p{FbG+sMR3__K@eJ8iz~U$%eM;Zl}q)1KVD
zJZq*(bnzc~l~%V`axY`g>#?Xr%P+Xl%9F3l>wHxH-h6{LXD+U5jP)bl_v~^o=W1^v
zad5BsrUzo2C1zoF20Tp~?5^MYq1VxqEGjBWe&8lo!RlPT95KHbm<(bgZ$5#c1sJFq
zQyIJ6#(p{T_Kx=Cw1QB`gK!8kz(~2oU;rLRa(E-L=Rtx)%gYP!4q#x{SENW1$QyF2
zWC_-o-u=LIF(F0s%}hM;qnf47@>(6b6F0V=KY#KenDmuuy6Zp=V|tU~>K}8Al%|!!
zXI^kQeVq)nIPl@p?GC%60pX71?^0Rm>4Kb2_6LvN-hz|JgP$#VBqY>x#ClxF7OR6b
z1!H~Y;>F*)a!hE@=gkX)-QyLgZFs+6iJ{-i5$_<*4=oYvax`8Gm^hs0ZtstjEr;EQ
z6%;A*^RAq=?Cim)NdrRtP-@OtSj5nj3~a$eF2fxkqNAHRCkCBRSjnsnu{_H%`}_B!
z`0F@ujgNPzyx~nA+)qDkoD&Tpi=2{~cH>xRqn2coe#sd)Mt)k4OxU+jG^gg_gI50Y
zPi@m~&u{nb`tp|7+_wiiDU<)xw=>~dsNHM>Fe0}XueDZC4+l}`aUwyojpP~`SaCa+
zc9etG&;^iS&OTkK5CCegd)kPA%08{h!@~n0dj`5VoWA9nw+q2976!=&!`Ee2>1Y9l
zBU7wD|0r)!{_CTV>fgG#xp91Kdy8~@4^BJ&5>=B*lViZ?<S>WmkRn)n#lWB#-L3mb
zBE+gQa}v$_)8&&zC+7(^ES2+?NH&zN6V1;=>TJg3LIEb0XgtWw-2BKe+@nZ&(rfvE
z{^^)vET{ylMfxC`6mB7*!B{@ilb^9yWcD&u&TMRKJet4DrnY<$43V8F6Uj)GMwNDM
zF;Ac282P7N+2NMXajYW6A@gct{{ln{;1ZeZ;Y)HGVcUGc=j<2~ZYu&&tU;n*vIKp2
zfi=_3gn(q3UM-<t|Lu`-+|~Mlj-INDlGkbH8!!U1B&j=to&34x9hF;~jlN&QTI#|)
zF1613J&68|V;YC8g229Ql!0WQS(VB0Nnp>OH5NW~S=CW<N-`I`JGrU!gHvhdGSuqP
z(NPZP7_T4%txTr+#m)q`zD8@(xOCum%(P4X@S#5{ZL~V|;_j^&q=tb=`qm%B)8Zf`
zd!+j`xMaxhj?lP^HwN|^ZW|e{t*tGPc<dPCjwqqg4e)z!V4m0l*DJ6cz#lMNOGigB
zvYoq)E!v8P9G~!nj`*Wga3ZzcX@U%cTQ%13UG?IGbHvDhj40KPSp|>I>(})UM|_6|
z2M-UY_zozwIea`Bv*`0PU_0@S=~w%WimR%|qH+u+2M#O|r}yycxAIThH*iP!P^tC|
zQNImo&7JB_d>RDl8NP<^%JdT@#gE(8xJmJSlh6gH>f!G*G1?ekdws$0&Q4wy7M2Zz
z!^ytAhj9>(V7wN2<}F#*<J&(}T#Q!kgQv|?R(!_ng*&Yq?O_iHFb&~KpSd*^h-dO%
zfb?{k;S}}6Opp8b*EbJ_F3aY(VZYM}Dn=HKuzy|5e?qH)<30T0QuOYhocU$s#_YTi
z<Nlsk*E}(!07XkAwC7Dgz)=yg1A~YA%U);d?K=n8B9va5^CKH84qp%=vM%|Aw6s(5
z_^=+~Cy&JgC6D}J?^lKSw%|1P@efcd#s+GhaDjbp*b~<F19eb^P29Tobm*ew(LYCJ
zWWffk&Z=E2eAVASt>y+W#P~z4Q-oh<pwS1y2bN|9ef0zk8o>VRurMC9DrOcIXd2im
zEXr7vqYXg~Uo|$qpMGJNfQ5wxl+UkKRbGdbV2(Spp|;%*L<E;ic=B-2>_u!ez6M(|
z{6$Pp6O08Gm6|nHjcdvBh3*lkl0FL94oW}Szhy{ewqCOw3?}?GEeH`GGNHxA#;!4|
zh9!V$7evAB%RR6}2RJjnB`A0W8=Gz-`X4VE;vnz~d0i@%9W=Z%PyFH5lJF98t&)`x
zs>nkwSX)}YZuXHwMkQqg3{&JEaWQ$YNK7B11}00TG@hc@YT|Vma^|7|$@JA;HRxI}
zOE4yxG(Ox)nhXh`=z-aj97IJ#<W8S{ne{@D8se>6w`$uSN9b+75UmUkkUv^GGyo7D
zWL#Zc$(%>n8%R47#i}j>XdAX4^sjs$&~!Rgn(IU;Q9}gtpp~`tw~C5{%m%bhfYW3i
z#2g-cYH;!qwg47KS5;ToN|D|um0BT9R-2|=F!|r|F$-x?*h9mb@MNcJa=@(hidkM8
z-CSJ6f*d+uKT!_?bUXRiY4g3QFTt|=7YCoNVM4nAj;RiMNb(={qC&6f;Rg?Iis90J
z)&v@Wfm;TG6<)bBP!kC$+UVz}x?;O9nvS)GJ-Zi3XKi2QBEd>u_*#6r^=M21_A=#Y
z-ACRSjjHPycmeahiw1g~UP*vyIH7_<L*>!IK&!u~58)|80nHOIDXm>ZL|(JaPb6^6
zN%h#Uw2nz22ItS84}FzsU<6kHuOP(~`dY|6#Ch)B%h<OQaosxCa%2nGTAKKz3NJA}
zsENuzxen+#m!R&`r%&r<C&K6Wgp^+F{1!gGGpNyMtK^_b(8wNy!=b?9C4d4n*wu@n
zQx(YJ*i8>p(>gtTd~-p7-nWNa%P}}3U|1aj$pb;ZZO<M#B-w&=7L5>CBK_=#%uGx$
zb=bR^ke!eH@LO;+$T1j=55cIV4%ofLNl6AUBommcwi{3*0}A4UoQXg~@ts$sjkf{h
z8N}7EI_@K(rdJ%2hpq&w7Z;8q;JXgFy&%Hpg(E2TiX?8@izOu%M@{XDKHByS%}KVT
zC@;@gT`y8;I~bjv%9~0BZMbJCDk^H4;54GD?s``Hg@2P99_><Onairh{r!ups;ZFE
zb}LIp#W3#+0F#8MJp(T=QpL$3HWUU>MI<eER^?=69VbHH#ih`3+KfHtQW<!(t?0xd
ziv!Dw)|a4K?=15it20miXn=47?rBC+4gedlMlJj7X*IR>re)dxfqljf!-?)@{b0Ce
zVqB3XnYfzAM><OoTDAWzgZWVp6dfHu$X9g{VD7MQ8Fi?>*~8ma9(?73kV*$9Gc&W4
zScvxa9iSR!&6&duItgMoeBf6tES|vgjIVY%6%*5!60IC6wP4<emLXpr6ebAIers>p
zt%|g}-P3$CO&X7X2F`=oi`QJ%CuYRe6}Fd(v+?^Gj(*!Q0!rf%T+IZx@2x+uM)kf4
zT=Qf~+V%EbI9w*e?V8U%BIp^tvRKd{ka);D?xC+<eX`Wo*B9dsumAPy6@)gw@Q|d;
zkg@JB<B_QF;7TOU3&HNGw|zH~a{!`0QnoejqQDE+>V$w#=!m1Ei!K|U_gM*^<jN1v
zFtFQDf{k<N4antp>C#$xQUNB#G+$VF7IBCvf$hjt9U%(_FH>)Gyn)6m2+BD`3_`G=
zrrpEgFS``2oYTZ#Syknl{0`#<QsJc}eSk%FN3BmX5D4YN6e8MR5AZqQ?cEbJK+^20
zDCa1xnD}w@=+UC`a_hofFk4J{1TppRY7B*WBNx{efXuI7zoNMVhb`+97rc-$?13vH
zAkL=dUGjCGyuc=VDRg3(QAz25NHOK$g&Igs4B$wqX<=ZQNMgXaE?8CwRFKv_EBywb
zQ^GE9Z!2<E+P;c{aND+)z(yDl5X)2Y36jCvckiI39|H6i(w#sWCczMftX&y6YS`&O
zc=Q*o-hB{CwLI?c*!6&oVbovdP>t6O4had#^s{RXzA`IUbduoI_60(w73ASjkde95
zeR+x5NsGMQz!~&d@Q0M{KX3rbtHI28P)6EAy0wSlv6>`K<*co(-5b%I$zj@tyu2`t
z%vr(#Ma>&9%%)8{vW#55xZ8kk+3;9AkwT6v)J1&XR&jAPoUfS^3?q*=XB{Z4LC86o
zRV>(jYAP%&24sB5qOyDUpEU|D<Cs_wIsp7@VG4$%b`4HKP<S|rV!I;$d<lK?rXec3
z6~Ty0JTMZ#EheUl6P-~4RdD{|#cwW6UkIXZQlpo3?*?N6##{#;y$pDusx^H&It<_*
zBQI}fVvidaZb~>sR|3FE!4G3TEFTza^NeYo=2-n%H@bhWQM*M|X+@)_&UZSnbvMfC
z>JWG}uqxW!^K(HgAxeE~Kx2U?5ZGVYDdF95<q`fCDTWN=Bp2+!Bd%Y+erBwYX21P6
z_&IiE03nn9J0(72M}>8bW3Y=lW5az?CpC~VUk3z`DT2~1E@{wdW%!=d0Yox>avX=-
zecga5%XI$98;$TSvlp`ZqA&5AuGSIzF^&f7VBz{GAqqY*FjLFtPDOFC`M1U>rR`Fd
zdP}alFEp9=rFJgFE2&f5YsVRWu!i6fPGja8(w-mKsJ^*;aC}9hVS@Gbc>d3cn1X@=
zGF*}&1<oEw&?aB{7^z=@VNo61yL-mI`1Amw2bifNocs=XjPbCKI7Ro18mOwNp?_Pa
zSdz9&1sNVU&y*#K)!Iz1Th2S>P{{6*+dl7rp-M*Pt4?29Uw^-V*6)R^J5vWmcJG!2
zllSA(p>PaQxkzN6d2{T(spYi~WJ!V92Wf!3Z5ZP+K8Ax$_A1y-ngAd?5)u;rT3ccH
zMmYSMuwyiOp>ab7MhsupmmUu>1aPl6d46mYrQO}T-%Eez>WIC@M^i$8jLjV#3OLE6
z#Fhz5-;ND1aej?HV`;V2p+l>u89)j|-p=_b&v2t21D<4rx{S!^&%@|QahU-TAn?DS
z*s_hlEue?yn+PrcGzV>xQfL-{X0jl1t-ZRIG5iLgOY{6!a2y;OoI~dezfz3{BZ;j&
zS9wWuLS*}P7mnXX)|37NGdNXY47QeYRFRYO2RT=u+{CqI%a^&v?fb&)1j3#?DG`pH
zpK1kX5#=%%1Ly*nrGp(X^3>RmgyQ+Ef4GpkPi}I$vc`k7d9ZMop&BFTR<y5Sc%sie
zgFQoK!G#dvpu5;K6y@YR{|upXb6KM8o>-xk>Nt%q#`=0uWRBvr-o?zPGy1Q%xY%#F
zt}KNIdmCD9)_>?kn_-7rB`M-qFS~;K*E-RE6;)M*;!+r0@)*J$2gG6h<FO<eRXM~0
z0T5L>Iyzw8a+z{IB13+=rZ_z{4@nY5zzmlWQ5Hg(%1Bu)UA@(4H!39!G$80XRA7U$
zsPxsK0WnOOo{ub(Q&iNx8B<dQXmn!8vtRv{beNpo#i!px_SIGz85>ho+;j2EKXgyV
zX<!HN!+jVu(uHptn#7*|E10jx5R9K;@{p7V*(K56F)2y31+gpHn*7?vj0a#ph9kI@
z90^9Xjwn2Z`T1A#8*yS!)9?rsc1$|Wj7Kv3@p+aT4XiU50ihs`k)zPBo`!|5`*O{R
z`bAz|llr!MSxo;r!GHg~ZZx-P7CH6}Duw8R50p2;(m+_zU{~#MErocwaH(OOdc=iW
z(5RRaBC-@14LOjGGs35*HWMa7w#Kh$&x|LwR!yx+#wQJg0<a^{0!_?!n<3&<)$Rm4
zp7~e<OF3iB6NJEl12=Gh+{)WRSDXmo3`@VzoX43XvB3g+kHb;ty0-QrqSy@RnywBY
z9fXX|lsqu|1f49y9$mX~f_W&WO1XhZQ5EPNaF|cq+KTb@II3p=C?E-;V-x=X9^!$)
z>))7pfbkK=Pu<$oYxH=q(4dYEiZ1coINE>_?A)Aa;Jry>gZOmSk$<%V_!cOVEz5nr
zQM<;RS+l<C3;5J?Ps?RTg=OAy1ys#z-@O2!^fE{rI*sHHLvJbLbi>cZWg*QR++tBb
zg`-V8HVN*S4?rJi-1Uet<_zf@MTk-)qL!`VYxcKjrR{7&M!q!acLk3*GcMJNX}QVG
zL}MC^H<A9&=C>yNfI40!sBe4r6wbZUer4sPtn%@~W(L<<?<!H4EJ;m<q(Bb+rGq#<
zY$HGvO*H6+Ydv;gx`wW+qN{*z&Rb4*;G}XhhP_eB=m7j*sT&?`j&}LvlhCOFsBoeU
z^+6BRrm3XvW6Hvwp4d*{?xm&nf}e3tsPyW1Rltrp>2eX;_614&^`q&IXMVwgN9h3m
z*Z0by<N?O*GX9tufFhPt8jhhHx-o}1L%0aYg8sCMosF#p59aBPg+dOT^riWs;UF71
z9(uD_@r8&jAS7vW1bl|?z8g45nNB==dD9GkBic^*A_AmaZ*EB8*|9_Xt)Nck1*n)1
zpfKLLArZ*S7cyD>&!6kT5k}f>WEefgDVcN_?6K%jl~+)Jg$bTcAbu_yo@e$vxW@tj
zeSmHE#sA3ssfG)rdfR6ury=AJJd-6vV1LzvXU|%p!NRRi0QAP}MY3qHaAI|yOc!D*
z{Eq6Wy3vF|iKg3e5#Rd?M%4n5@Fds$SZ|yFn;#jApqrgLS9>}JK11q}N1{{&G0`#?
zBMVb=IW$B_vl!s+NhAP@*Oaum9BtYx3-7uft`;E|#qA}WC;qu*4>)~-9uOy;T<nh?
znaQT`K||UG0T$U<foy5Zm%|MEb<QTo2TVjNbP`PuHNh|?X>tSTkt_J<QP(ov0L+r5
z&NXO%VJBs9`0(5-D;Pp+VH@5MOOgdhP`bqk6CG&DzLmnHr3C4VChWmPg_Ge6PsSvs
zVEn`$XD0)UqU;fCm7$=jYWt}U$(JA>x@vZJbh<a=_HuqQB&k;WKH>phc$5*?qDnPk
zIeSv1i^7W~V!PIUKhh92%gHHIICdGMX)W*qjU~b2A!|`->_plXrhuVgzm%@2M%QhJ
z5n^1L;r<&nm+T0%VsLUGHLS6mmHHMy41j`$oy<F{Ls5STF(VXeY&<~=X*FyDmntm$
zEIr53o~|=s)b5Fjj@}AMg1{_-mP~7J{ru`4=CHsKyZu6cRr(K9yQi0CFgyrzgE`uF
zV|Fuc-N9WDx9CV<&9-gWkT&>FDDJA$a2ShD1_vf13m*^-Heu<wckR0Ga3-&nz6h7g
zVJ3mbol##vUYdNW!xY{_jEP{71K@KD<^)zb*Ne$5^c7^nkU<VO517Ln5W-Mfru=%S
zf3T2<Husd;`~(~!!#}y9cb;=x+7KTX=am-Ba7^9`+&N`<Dul%bET1kUMNDL=xF34m
z78A*f3oY<lw0uM_fHUgp<jocx*?$=|2Sfu7Bbd_h>N+gI)W{T#{J;qF<HxLZOwZHa
z{GWm<pyR;v7eUFR>*X-%brMNGpip5wur*3uhSn0(4gdc8^z;ly<2~HYK&%CEgru2c
zAt>&UR{E6oyblREMxeYJG#n@<U5o?JU2goi*AJI3h;+@*RV>qP)Y8&oduf0sa)6@_
zT>y526bq1BIN^ugljDH2bqx$Uy7$d?9H;wb7`>vPaj>lf6{jFB^0#+a)?+HNsLM15
z$<D4PCM7Km3&6-4#U%`(rO<4!AbKQP;vyfq`H+|qy0n2z`D4(|XiOisKB#*<1W7`y
zHINhh+fOhucO#6E{S65QFN7gssi!u(BTa5xzfSmqHW^$rj<1%vu;T+JRv%7;8j5CU
zb<933ZS7VlaGTBa;UJtc1z`GyrZZ8de`L&JY{t|$P2Nu3b_dgB9&~~zOr7HTz%xj*
z8Ob&S9kd&y%zY^-QH0`Gg1WH>$^!aa;jV4DDj2@qws~{oL#;;S%58v$u;WVwvaebt
zWq6qpFgZI-Hb6AA!_f4j^-6fi1CB!?c0NTFYlGShV9B~*@N;_lt_>`V;inM$)Qr<t
z#{as@O`;vlAFHn!;*3OSz;r<ON#H|{c%Upz-F&F-C*?|gxgXE8#ULc1)t~D%&HnMD
z5DLo4(r?XQ%Hj^(UU>S#g?}SWrg_Q}vP^vo5gsc4Wx$v6WV-|a{ye{uUlFM95W2b1
zvD5t0v14+;3aV;DMROU$%kiwcG&3&$c)E*BDAZY_rO|0E4uv(4ph`rz-Tf>`px=s$
z<Y`m@LvRGo(Is5fmgWRI1(TcphFr|@DPTTlfFrF<mEo3}0B}@mic?6_zU0NgT@j~`
zAI}=-;DmD-zf3nB5`A#^aILEl!*>o~&!nN9+>->wkfDeb#%T%0tqP?v0D)6u0e#S%
zzGA+TynDmr@RIF5Q1Erno;`o(Pw^x$2;#d>TB^d%g9{`05m$Z|JSmzCp&F|5d&5g$
z2}CPI7e~+~0?49n3E_x#R1EO-y838P8zYn%do|$meu)t)UAVeD{9#kyP*?>F0%FU>
zt^zk@(A(kqKM%^bG<AurOx-fnP7IFM-D^$FqlstCv}wL9!a4!8cy<Y)dqv!==41^(
ztaEYwMmA&AFsDO%k04hJ!{F50SE4~35mrL9!5%JKY+MF=)A|Fq;k4-yQ(VAkYDGom
zw;MJkHW3UBxH3XV1lW+(5x;CB<&sY;(Kq;~Mc%-}@WX=)hV&TPQz*?3%Jj_(>w6a<
zhS8SM%r0Du<Uv8}2p79gmlBN3up)C@GoWBmRn0z7&oI)RT)oraIA}n%D&7oHBGzFK
zM+LNsC}QOSURUj(!SEz&@Dz@L6zE4-47PoYpU7f!P<yKU6oa$0_^b!Ra4nAq!C6c?
zfCk@&!;iqYnXFw)N;|AjINnjno0@Qp3jd4%<AR*YCT96=_7V;n8(qY<=P(+f#+O$a
zLMF^p@Dc=bZ<)4_4vr*%*~G2^V+)HCxZn8A+mxt7NBM1xJHlYN<p8TF#r9-Mjz>Uz
z#jm;CiUTYIjLoZi3Np*D>V_9OWvXhZ+bC403HV?j4#W3n{g2`{J$dSY=g$YXW2lc#
z3nnwsw<hQ+;S0cRrV*JLRgw(P)#Tb0>cq>?{gQUL*bxgZ-XMWgOp~!0?wsN9<wTyF
z$4J$o-^E!43gpxJanIE9b~C0PFcGX7wQ#VbEgzPe4gW}xxrgX2%6ypnBAXOue@snH
zeOlWq^7CK8S~<nI;RUI20xn03g1lpu--uu*_rlVLo6-65@VL$DV)P2r!qp~ThSd!&
zvOZ#LaVdplg9Lt)Sz|EHikRJ{GLMm~4Re<6<k;7B|A`l_kVJx^&#+e9caA9o=?@OR
z=V?0)<I~vNv>CHOoiVw6=FE2>-CM9?1s5>8Qw-C*J}DjeKYz|Pb)PSpO2DroH)6$L
zezpA82zMSC!SLA7ELNfur#DB(UxR_U1<Ysn#5IU?0O%rH+#{-hL@(XG91j5No-X)(
zxc(bmSz##J34b^SywJmt0Z7Kl!2taP`BWhOD&kH8?53QYT%qRDm<=>t9}&?VYw@{M
z<9khw1}QkWmkx|&1ooQl&OqRS>v@0}0WO1!dmHf8_H8hH69Equ2_5{TG#3c)gGYiC
z`ac&BgrKUCNHRIVpE<g=Rt~^mHQy7=jG+<sLqeo_B$hgxA<YD@+EvJ$y+55eb>W}D
z>x;DvfpMn0TIsS3ewh6Ct1%d+mJ#7(z??Hj4_Q4Uc9NvP_PZ3%!dWNyXJs`sk7Ze;
zawuKnOEZ>Xygywq35X6%j#QZcakpUsVdjPBPAL8$#yaTyr=A5{@DPGO{6UJii%UF6
z+8Yy}sGv*-6iB%YmvszA*BJ01F}oNkMs#!b*+M=;R+dM}03%F$4fCVZ;<d660O`IR
zjD{|s{jm`y@%x~WD15++;UZJ;)l6s}d4qZ5#lO6~;XUe{jNVzyeRo0BU&h@E7&{+-
z4VSc^H@2r_rK0Lfy4wOA-r~T2fc3Jf4RG3i+6=S;Rgq)v`EyL{hTx3TogkP3ES=I#
zk*J_q!qbCGE9pWAq*fRqCS}w6fG&CEA5nWip^|kkARqwG2`w+}x)#HTgL&OWfh^=N
zS{j0C$lo?%jVKNzXwtATs*%n|2t2&!e2%wvb~?z8jKD#DWhu&pd%Wlz#v~W*cn0`5
z2&&&v7#GG*FL-m)$w?2QDQs(Ok!Cb{!Mr0CQsqX>r;uce+d6k(uBt8$Q#noKki`_e
zQaT%cl99cE+{Il<2-H4zhtUH*@X)y%gNCG5k48a|-hTVG(2x2u?|@@ulDUoUu7U*H
zxvK+@PqWe{4L^|hpK9J^-c)m#8Id)oF)LdC{fQoSKs+NUF>w#BR>1t*h(8OiD!?C4
z722h}9;7hgs%9BZ+pc}?90_!V1iqOj%|7YS&*grn#mmh>``3!vS3-A}RaYCh*jQRJ
zUk5ehz&ny(7)i{V@5e|5rb}z-{y_xM?u;F)oSE)KsDbn6&NVuBi+<Q~IduuXoa9%e
z+nvxJ<gjA4BL26>(LEMF`eH8r^#XLL-g0sheK&c9+*f)zUrq?!MdWTmts7wFgz`=o
zAOY41y@5UnxTQ1s;TUSorJFwKvL<vR8J5T`_Cwz_biu5`O|dwHdC97g2pyiLI>PgG
zS1KT@Kn9w^UdW1abc1601><wf`}G|8==!6*oODufd5E%)UH@74?(aMa$8fbs&e^?_
zSL!fmSBF`lugC{0G{-f1pTYUs#ev*N)4lxsMb{>;^US78lNrPJa6=#SqC3nTy6l2T
zO;>@V`|&_b&zQQ1if(LX^iD&_Q*(>U*Q~d{!s80j-Cba^Or{>s<QBCJn{f$^ZziXM
zDsBVMc#0EC_jM6(g|WfEU%ipO+7p+v7ty6CNoZEYwc6Fu+t5X5m^f%mWriK25x47t
z-UviLfEn4?4~8`cE|kA;yABDcz~F19ZnUzcB~QImsxLrgbU2N3oP<8jQJRUzjlT8h
z#$lJ4GkL$U(B$1-%OKw~e-e&EcFq68-h24<y#N3IugIPm6^SIAgY1!!nY~xiAcaDc
zQp)I9MJ3}PviFLThKQ1i!;y}tR9Z&L>L_J@ulLt0=l%H~e!nhV-tY51?|SuoKE{3A
z?zj8xPE}4=02gBl>tDTgt$N3OI>_<MSC{Zf=)~8r=V<EytwO?6!~P6XiR1Zn5Lo`9
zKg5bSw4lA9h+*eD3e`*qVvFwww?}O#zrOP1ZxeO1;XPy%k=8H+)98~75O&2rO}|ka
zqlbU&R{oS%16d-kqax@mWLpe;K)4Y4o>(};!FIJ?q&&vYHnQYC^C-D2{rv9>Db&NG
zHW&}gxqsNY8vwx1^HnPu*GN_}&IG5b*<;=Xa@y9m<SynhE=P$MbyX6iN1~PITPrhi
z*<K+dLz1_BvwzveG{{cf=FC}Zsl4Ia0zCo<TZGI7r|!7jp;cq3eC{-$3@d%ZV5e>_
zs`L6xWH}XemOxg2zqWF)Qi{WjD5>qL=Nb^Bf304G$OFWy0k=e2C)5BPX>BXAXpkY>
zl1dM(USnl>A%93VH&&hv62|zQ6t+cG>pQB7f8JfVT=i!ab|PI9zse>eOQ~1~TNWf`
zw*mWA0>^MRvL_o<2ir(#0UvAtllhK1VFvqQA3&^Gf-5zJ;#9M%@`a81$X+}MClZkt
zlyjXaQY~;wEj0a;s(0*wuB`B@ha`|9=m4lTVJBo!c|K6e4QsC{H>{6Jz40=FoyPy(
z>3|>DWguKTInk)uY*78h@|JSr4$%HRSrq+*M`PQKfDUC-0mQ4HtGu7`r$%F2zU01N
zdm6g#9p&lNG&DB=qN<~k<Sq6W17kE1S5k3){`U#=toKYt)%L7NjN!u?t3*^E&XGgx
zq6sOo1bI^F9Xyffe>T@OMOSFCzn*N7@^2n>zWLo$1f<D$48~LS^r|ZG?dMf)Uq_{(
zTU>g7n-FX5qP!}W%14ZeAxS<((PWPcM*@kYc|A99avxboWR(k@4*0ldGDM?ts6bsQ
z$W$-tm*%qcQ>J6of56oj&QYJ_PLqO<2mu&p`|H{B<eIwFhdChkXbsxZ$(bUyk?k0L
z*WuJh4f&J#-{f@X8$jVHmKOxmFu1$0_v^hkznw(0=xzRA@>7cMEoC1&A4RYLG3+2t
zrSc*-(CVHJH?56I5BJD0r97#+h;O6fewI*eC+bb-J5JnVe`kC9#wxwxhk0q;2r}M>
zgLsJSI*UM-`@cna*j2d*MV?T5Dq!l4$HMOHOSmRto^;)~xhQYBqK)#<cwpRw;V3e>
zz4}vB6ebx<U%duCqPeXw8VNPw(CUB3uVF)yGZB>%Et{*c|JpW}mwR~a!<3c^pP%yy
zb>tS2_-We?`+Rq*;K1;@s*g;3xSRz&>Br&G4wL$H98caRgqpQTNDiu2x7{Ym>BIK|
zxxc&i-jrm*;4ab+1-CZs@4IAY^@+*<DxdWw66{D60H<St3}A1J<t(R>$>~-L40m3<
z(hbXJ^cSFvNL02iaH5&%5SpO3h+Xm6kHkupPZacAECZts#_hyzmW#A8HH7LV_t`bp
zzqVrd%?O*M`XZX1yzcbLldr>5+U^2Dt9hQHGY27E`9zIykusaXoZ3&Xl_;c^z>7j7
zKBW43<(JE$`mzb@AX4Dye)dn6NFsODtAo8mUYn`QXSqz_KKKm1M9)P7&2A47bq@>=
zaf%&}>f=G;++QDcU-?m2FsQju=E(AP8_-82OnT`!__b2^mCq$wGfPp|7Oc_$S5%5M
zdV+Azua&YBo#-P|$;b8z2UBSXakFbbhU7uMxSODIATFQ6&NWmVN&O7d+8w1??E)ui
zREG!|RO>nXaAL!Gis6U<O<nh6rZksO9W($994{Pm)EBA`y27rV@~=^F{!%*s=Q&i%
zwQcEaRjS1UHdfnk#RFX~s?VckRQrwx4L$c0jgaI_dL`T$v~Y)Hv-&=}ILRv>FekJZ
zfolp!+O{Gy3KeS5|2ckFl?vSEWI3XKD`cqg%iY~YD@W!m{5pOQn#kXrDcPO)V^XGZ
zJ1OD`Szg4hf1QYrDQ}9|cca<LqeC##A%}891_9it7-i&G>=6uu$E7vaR)lv&s8+`o
zU`2#vq2yINN5wX!czAuXHTgf_yV?dwbxAsTVAA~eQzCXzEV65&l-w0TEA6i`FhkTu
z7x&5CEQdz%T}u{q4q88H!eoT$r1WLNK!i0mu=;wnQC_cQLQEIKPqU1H@C8M?8fwF@
z;}9n?(vggCG?y7=FwgsKKU)XxgBQeB^6b|!x0HXec#&9#xCvm2;{TQRUdrfgiMz>J
ztDUUEOLWwSUvdSu6;}s^COMO^`w|WImxGmAgCa#Lb;T`(?<)$ARNr<ARZV^z?)b`T
zlT|)X{Dd!9b^3}_<I@0zt!=@~r8R-t)vM~Ka;c&GDN17*NP{g?9tS~PFRiWUN>h!I
z%JnM_$dHA#IQx5P9m|)^UsJJLl!}%6%^edZX`tq9Di5nTV8APU-%8cu#t%_oq<zC`
zW!9p`7tZ{A&C`5L9d?NtdA7EUtP;L4X!YP<f2YaJ6N>V#(&>ShHGqRP+h5c0A5Kc5
z!?TYk;1osYET0(UG5}%@qaZrwP$S2j^yP9F&!MTRB4<VF63MMIsXm5^d!_c&_`(He
zvi~pKv*RL(`q$mNJKSPACJL2u*tdcvh}gKgD3cO}QD=GdZCEOi>Xw--qzOAu013bD
zJnZw98=qh5)R~h}nGP-?tw<uA4*OyJ3_=y@XITREUW%^2K8-#&#rK^ERMRAIMwF2S
zq%fejwiF5V$0B))X{$3QbHHk5Y{Y(gx~zhzrnRjIRT1M$!*I1b7@o(TADJ(Kue2<H
zHfq~;opUiAK_jVL)G>XTLm(50=aumc9Q|DeAOrR&GKg?(;51dc#sLwB|2_}sRXMyu
zI#ceyO#IpHUEa<4B9DZAFcKgo`SlO_5?Yc$T%aK$MU(~u%D_<LfHGp4(-IgZ1$LHQ
ztfh|u`6`07!ye6?YJ5`Ed}qL>#?1`OUH$QIV4@W+o8smZ%81Jl(Rdh9dzD5NbUs6V
z2Na{u9Hrj#@I1M?W5>}=Oy@#{j4sf=M+s3jiWGIrv+F4$oq)RMYf2k*AB6;zkyYCh
zc*lax2JTB!J=dE;-2sZ#zHnD$nh+6zh;!~thkGrFE@OPw1T~EqCP#u-qqFM!I8ogM
zX6QH`SVH1<(ZcPsOc^KntPpYQZMM{O%x|o|D%pwc1r8#fL;3NDoz^jZWk$6`Idg;-
znX{6k++?9DXi0li{V5-BKTfNq814-7__#=w%XB2^nFDbfw^QV2g`Oc3hGeo1knzBn
ze?E2qLXaK}`V+kZC}5b0`-julW|nFsz;_uxbYLS}qn!Lh@O(zpP9grvt;gP-g_J)`
z)GmbZEk-6beFZ4Z|G_1e20%A>EdxGSn+b=Nj$YA53XD@EwihKyL2HCAKlFR~OjVEV
z66DpDqn8L-WSkVq&bIEA7w$ZG@I<n0_tGb5=eH^XN{}-x(mKj~ZlrQ|zKH&^(P&!3
zCE1fGn76iKO3t4KK<tt(ON;FVwC+mTWLkbsY&dW14hi!MJq?mhC^;jC;Bh^Y_>wsv
ze`RE3j8KU?D+bGl3(-vM+!dO0MWR?LE(T`_5D<;t^K9)`N0jwu1Dd4KspP$;!3MTa
zlwqaShNLjJXdsD{@EZ6L0lYSJQ3e`{JQ+7)hv<OxubvP#g@8%uxs!<PB9oR;9Ap*v
z%-6odz8yU^_3%Y%d;?<_$`~IRHba)X3?Zbk#_eENODZXnIGoa&zy5$P>;7OTU22k<
z5PduXK@inkiUt#Q^1my<kEG*mHYuDGK1FAZLdh<F|9h*>r!9_n4OlAm`5RwKx{B&a
z%CRLYDGdpuUuKLEgdSJMUp~*adc0M*>>&j#**5syLm7}H%}sEhr_(iAwi>z(V6@}~
z_XrfJcPW+X5U}LgI7bH~&7pA#_Ww({3UW@+(wo*=$T_w9mS0C=D)8V`b5&bTeh%OO
z>RI5j$OUE)mCS~Oz&Ux%;hIL&8a@?>M7k-~e15TarvWu5k#hnChu{p*jf}}O*g&HC
z3y&XPN9ZO!6f`O0ApjAC=)P#tNz9FW@b&9MxK?deY`%C9IKI=A180Dm4^?K}R+$Ge
zTk)%!qJ}GjC<Z+9JPriQ%qmTx404aTJbgc59fWP_>RLANEZWFYqC;LUU35GT9GG-Q
z8MgQ|CkOHQ*o_)CQoT=v7)Je2#-AY|N#ZQr6CMe^_k+KdQ3zfNaYTxIBKtXGhFLe|
z;`j^|1fdP&P03q??9mI@`_w;($v`)STZ9Lb76-V3lx_iMJH7tkU<uWiJ0*SJpe8u5
z>4fSciO{_Hs9@!@4R0qATwDM&Y!yQ}zj%i7^@sPcQ^xtq|KamUsSKw4=PCc2VR_(x
zGc1k%&kV~M5-TXb@=r)Gqx^~=F9D_UtD4E7{O8VcSyW&D{|EnnW+^On@gVs*_RJ-v
z{3cBjNJvOc0g;0nH|8Nof@Xt|6yamzk2{9^y|@hmD_ywKqE$e<CBEcmJMT)}wzK!$
zX!7k)pSvffcG|V@AMbHBHa`3FRErUsBWkV8O}e~wecf*7W6s-ruK&)w`^p7YChD3e
z>uSbN3#{9{Y10$xHB82h>u|7q-=yA49<5#XAta*J&f$6+$6p;ryX>y(r|<VsYqmUf
zaB73AmW}z-o?c$J?%iw9Wp{nP;kGwbzcFu|iSlb<gNB}4{03bGy0nb*YKbw$%Z*vN
zvOkgB(Z@@(-SS|)n;x`tb`J9KQK;9d)!x8BoiN)kAYkL^)13{PjSpF>2bG$yV<EQj
z6HU_Z!Zp$xIIz{;?p7n0F6}~Q4=v(d@x0#q8Zvk>EFxlqUfu2E3@C>6)NWY#=9xuW
zX@yPLY9w`BEssvq8#JgjZPq5pP(krp;Ox9^tUJ0K@-_XV_J?K~?k=>X7_Iv;<iR0h
zf2cWtNzeXmJ$CHajhi>?|Mpv5%*BVYvOl7u2Y_%>Ls#SdsU{Eavin*2&`!JR@!Oz|
z>&-(q)3Y&td_7Dftip{D=NH=99fufrH?nnLZPX2JVN(^BhhJx(pVI@*vw}jS3Di}K
z!A(_<6RA;ip8m1MAjb`fg&M5={6fp}x~E3%4M&q!tXaMK;WZEE+qZA)b?Q``y4JL_
zvphNuX{oM88-|O!`=-N(=lj2++Sq>ZU>(+c+~molSu5DW7fPO#R>Zgz)aY&*;{&H>
z<k4x?m`&~e4%!q;)KS3epD|Ijb?Lc1^E>L7=%p`}Myu`HxBIa2nyC$%G)aLm+mjBO
z^3seZY<FeYUoWxNN2eYhTKwpAkP!-^J_OhKpw>nn@7jQgO`AQtW7`b-zDAxO^q~0G
z_4f8|*t&0B{NgQI1DhZjY%!!!^)65GRqLU#qY=4LKH)_bGex%l`DeOA538$|+o-nL
zk{8R<LWy;p_WCUnH8#ax(`QaiSlGLLzkV%cVF^apqf6(XI_FYG5em3QtNt~3^Y(2o
zx8iAtx7Nnt+vd%y{%axDoT-I1X`J#43Q7T^o<;!kJZ8elTfx0-KhSDs1J^YyqV!wd
z#abhVUz!*9knI`*toCqeWEIJVEDL3)R&2ZCqHi3ErYQ)?6uf-78SaTSydXb6zp%1C
z)knC*bI>}Q;U?+h&DP)J6y`I0YVx*wFP0A}e>f_*?~-O;2m@Yd>Fv7rBG|}PAIIwl
zHbNlW!^kCHfnMcSni>=t-e2EkMbv@(IfC;7LcS3i$L0AwRL3UmXNu2WzcUAyM4BVE
zX$wnl+WE((De5EQp$YV2Mk^AXm$RcfY`?pZHvBp?={GqN6{QKGK^6egT2lYU`0y&m
zfrSD3DW9zOqWZAY9V`c`{=^Q)UL1_2oyo2J{QVPQ?a<R>?LO(`i~^)6Qb14Y@1C&e
zc>Tdp?m7=)&W<NY-t-wW*5lb1@Ib%)Mu#^k;MNE)RHpsF3l`wmxwHK*9$>lnDt;zc
z-EA-A=mNs<1ol0o+y4Ic5A}wCE6Q{0?VjK`Yr(6dGtTxzE_nzz*|<rQZc_nrd45Nu
ztkmmW&cFkMJsMLPWglyU5hOSbZm?hV>9L(_UpF!F7(Zo7V+r)y7<;t<8f>Gbu14Y*
zPfE3k`m~N|=Bj$@eK-*|v@ZfnB@nFoLVgN=F`8dQ$YdMW$A%7b*|;{kx{j9pRLe8{
z>#ks((@jlG6wo&vtVV{^U9UJfqh6?K_KAKYy`c=;rU!lt!H`+4YmFN>4yN)ca3}1}
zC~w`|$)iTqg#8r`xf_G|xFp!JdkPq6Pj(J7lX>Qr_r1UWZp7~?6oF%EZb+~$+IptX
zisj4K`}<Ek{d3jusFe4UtqvbQ-hs0H=kLe-L7iI8@5{`Ojq@%=yN+I1@Zv?SgHt+p
zpE4=x^t9is?L5}BT|{D|9i6+T{-zqWH>lxiqPm~7@?5nlw=D8N0^8tA3?|R>Chun6
zxt1Gt?P|XEYj$2Cb3fP5ID2KmCaaK1of6T(*o;Lh<j#WGHh*w<Sd;uOw^ovu$gbR;
zmx5*Le=%6K1I6!@`y1jmNSBJ(Hdgg~t6Oy0dol~|0|Ex4j!w%AyY+5#Q}uw>{pyRE
zKrfg1N%QlOY*F~&5LXlDW?o&<nab233?6BL>Hj~UzCawsgfYi87{QQ2ntSB&(j-@u
z9VQ(fs!L^XJ=N_-T~BauZRvE6S^c>;31kqmQ8i-PmOJIdXpsL@f&T^3psU_{-~Hsr
z)h;4ZS8THBxCs+(0bcGnc+j!mQ1XK7=(c>Qtc;$#ApOF;IIjgLx?pjW$Y*XJnREWJ
z4QwL0D4;MKFb<=ds_V?Dwy|!n#OLIixVrWdiu{cic{W&MPhL)g6LtKrGyR|_hJgfI
z#=6x5zfD#;jrrqT_!IJxU_n9fcSh;&$c!<sHfh;P1JKGVt}R5D6p};7NOKqTaAY<j
z%enOr>(2k+JXTX>H1ht`P0~_vEQ;+14+f&_QHMsPL_%>5ly-|Ii2dTrsfoO}j@ig~
zimk5qlagu?uqQX4Iz$aqu46jZu@`q!Kp3Ot&+AgW-URns5zkD#4}c`~fTL1SJ-9dT
zuWHY-L_EuvpD410t$b^4U|?VtIQ7twR#Ye!7H5neXV6USv}`-%=Wu=k+Dg1b|01$8
zqm^&81zKV!4>ky|zHq)NyBGPL$)@@(HjHNvy_E|pLhk-E0U)H+yOkqWu3Q;Bbj={N
z3%1OSSiEI{la-Y&9Nvwj$weGlqJ{n!|5Ux}nMH&wXDq1U#M6tu4d8F&=|^e`u6Q^D
z0vvK<Tf(dlbKR8w4Dwh{Y7?ueFa*PPGD1!!K;~_lG5s*}uRp3k?zo)B+?{Fmw?Xav
z!6EgLTW`~@myNodeTt8@SFU_x^XOpJ=g$QUw~40}QKNbD4%-e7Th|t%Tq4J4+RT}@
ziS?`ZV!w^o2$Nh6ywV5(Ubx;*Osu)?d+Gc#I?zXO6v75wj+DdAZ)Zl>-Lbj&^cNSP
z342ouyKy|tk4x8PkXv1H^aM;~Rpmz2;|@}Pkl*o`Q=)y}Lb1He^1W~?7q<XnuFFf-
z7&W+tqV9U1l(yat<(5#jt1}|zN70PqY1TAnd!nMVBP3*5QJ3n6<#<mYz`yZ_B2o)u
zfEB8vI5VruH-OD#_ND+<N>Tl@9d_P(;RzHGhM6gUpEtkFP0uoYz+_)w-Xk||Jlot-
zbpxk+$l;FdyX)>odhu<*=S(A3;`6ss&C{fJI&6V^`dpa1HK7x%3xisZQY5p$YS+B+
z{gVYe{5p0I4urdIsy>ER;dOdwxX*?DRR!Yi^5feoS8-xAUnJ8N=xEl;eN3}0>R`gj
zOz-+YB%)8P5oCsHfG2(DJ(=2Z|KMOPDp*n7cdM2!RzsGrJ0hm)tZaCFkIwjKpsCFT
z-6OYP%9$z_z96x-$>`Brida#(XyK43QE%5Vtqz%U>v7jgc8UcJHEodZr6;;F1<vcf
z>O5b5HL&aPi~9ZhYoemkQ7*L6z=3KC(p!1DLXeHozz>cM43wtenA0DGQtnds`c{R~
zlQDEW=ftIF&hMOX(vx1za0H41+8Ud;nL2)a0+oqS%?jAs^0l;yMD2c0;Sg)ljtPsf
ztO+Elnu>ri9R$Obw}#5t|Kj%Qx9%by)x_fu6MZ>MVSF$k@~AdFxI^=QewlN@@@lg|
z2Os>mi*s(Xi+;u3aZX8K2!GH+-G(mCLQlGaAOmfn0?#2<Jxj?_F_9XP8`ggRvH<pO
za38zIs81>s1P7i-b$p=pderf>%$SkYKVM(<^Ue27+RWtjd}g4Jgy?CsmJHzjA!8kw
z%va2tRzG<0N(p83m)a3SgU<H&06=srEloo~#JUOgPZ4Kf42=|lU8|S$!B&=3Bh6T^
zZVj1)Xbt&+Xf)Vp?WZ?$uk#7qw<aL^jsE+2Kh@9EOc4Z|rGSFjrjQUF#gU5_8;d=c
z)j=Zie@{P2`SiKA96Z{j67Zcp;dZ)y<E{h`_UI0Cxc2ch`(MW3OoYU*CJQ@o2F;=z
zT2XqeO{}4KSX!Zc9V}og`bhrUs6m628`Zx*p9%4|CTCkSg%YsROha+sOpvgERxtTW
z-=SQa62<f70|CvqiXl8XLrr-G2%EZ@g(=`){O_anl^?BphpQte3kz-4hWbzf46F4p
zz>i(L_+aznYKuKKQ4YYF4VaJwxe~yB<bG<3`-}UCiN;74av=Zr+4WVQolbvU8P%L3
zPWxs(%bz=}NUZ>%YSQL&^XQ}34wmgze;}(hcZlriK6|<D$H&9UJ7;3zS2XO~hnjN2
zc=4ydPT1vs%1=Ky90}=gU@Wl{_zMMXe9P&dSBKf9mcBId*BQ{aP_B9U_{{8VTU61i
zI5U!sxqd@-HhAEQPLPco0Dt}3qvClI?2x!Zb+aY;FuP)}fiqE4n1^aogGs`~k|8cE
zs<!XN^Tc16q}L<*1mm>%?}Q<@Hst5GkGTB}Y4(I&s$FvHA-+_<<Tllta5UZIftO!x
zqnI{-eh>g*0XfPT80Wtp{@5qw)jP9m*Ej>u=R3xRA;dHZHNQ3pJf?^W#D)1c+ah(*
z|3dW*dD=j^lHEt1omD8_iK<3J`k@s*cNgB`aDL#@==bWCmRxP1({huQJL}(S;Bo`;
zZ&Dy4v{(3`3Va=IR252MM@`Q={@STrL#3&SmANt3DMCo_NS1uNjDD=^(5f<{o!kB2
z3w2Fwn#OdkC*)xg1WK$_P+VMB@$j{qS>)1|Fh@2bFQ7B0+L}5hsGesRHV>H$F-<b0
zpv_bQG*({X4>W8t!>X-E>`c|;)Pu$QjdJlV>PSasoK;xP-{{DsrC?+4{v^Z5&pBFl
zR2{Y4#>&%^>xb<F-6R)leE-0}=IdR)->iX&lJ_m6K0w(ZHF<Wm7h9&fS(^_WJQz;R
z$sy}<EpnrlDZ6&={6G&t0b$)(Fz^4Y$P?9x3nb%OKzLSl^XJ!LzqMfGP=a+lDPu1x
z1Z@T`-{`1Q?ZbPj9_fHy-QYp)UEmn>gicRbr5HYUZmXeRiaT7Vu5jnCzb?D(R-K~s
zHp<^nS8Tucg6O3edLaQr*L?9S!pfV4S*tOHcMCP%|9sQ{8#$vUre2@kY!Ty4b@z`M
zXVn$>G08e9tx9lS?iCk!US&A-ePr7-`VssN0h^cqR|X|#W@V`<$N~s{B{C``TYr-(
zpNvhjRqn)@(GYyhVqDr|`8HGO^e=)*Y<78Jy=SXGPX6?YqtDmn_R7b1n}`&IS-4Ft
zi9ji(xeodL_j(`!DdgGHO-&d7x{;~`&?J2+fm=Ii+Uhp%x;9)K1)`Jza2E<Gq9oD_
zqO>TgZJu7KJV$;`{<FFyfxm~Zcr`_`PSR*dY4d_`@wFiew9&FrxzO0LN;{sW>xzrx
zoHf<b@+FEQP**6fm>bIbqWRaB{MU!e?0Qo9$|hD?a@q8kHb+~hn^kNJi)g1L0FoZt
z_Wt>&H6gzz+<^}q4y56;!X_wR{n*7V%9YJ+3C^#~+NpPxzst2~gW^dNbaxLi)M(Of
z%yNiYP`%w&%1f;ud!9~%0Mgh(>XdbWzi*dub;N=Iq4cJ~uvNM0z~<Ic?nUuMtcErt
zH6)t44zHnZ-MUSKtgbGvMflZUZXRsUp2ey?Sz1%MC+SW=56){p>PVC<QByv{=-W;t
z;(&U$feZnRr1_gP)t<Z6QLftbYY`C<iQouZ2~w|9fC6#T2tpU32=-JW(t0ka(rX8f
z(J`H%T(|VAN*j30JNenOwv;s!(SiFwc{CAb`;a}PHB>wo$eF}B!y$7t`(1f;r@x99
z+u-+ZAW6)SPnouoQQ>Db7=_)I4$5TaQkRj9+EPOVhP$$$dY5!{l)H4fm1LBpEs|zo
zCIlHO3pPEUgcJS%BDH|r5=1uQzhB#RFtmXHLCw_Fh^CVWl(q*4`@xSa;Mglq5z<t0
zeLmzTctU>Y<&r;~ORF<yvhpn#daEu(BIZ=P_rkiH8@5_c^e@Ikj!}ZOLSfLU6UJDB
z9v`{)gnI(5{rCHW9nyyLNxF0hw2%?xZ*s3`Nj7ZRT`LI6s%=HdV`W%~7ZK=T^KRAV
zXEj#7ZplNKDZNN{c)}v+lzh@g&kI0Dv}7eR|7#<a+X6DQGnWV<B9%{UQmJXt7dUhJ
z>HOki!FiPU5cfQc7V$PiS8IcpDqKGQyU6m!>Q>aw#D4aoAA_&3Ibtp}G~Ab{UV7#Z
zUb?S+Q$Wg%ly=@ci_;-?0G5Gx>^4Z=gx0swW@g$@>wJLThJzr)ELYu-Ye9ys+2%8k
z(5y+O<mZE}Ens+PGXA=K-Jf6I|4m_0(##>BpYIaWB1jv+n~v!u<!iWkxYm}ZarauU
zXH8qUHm(g)T{u*7`NKe;3Iz!jA%)m4mz67%Q3mJ<y-o1Z8E4azA2(ONOYT6`jW|gt
zT_h%koY<c}pEudjO@{rb>VO?o965EWHWBJ=l6q33ZA@UWq3r-zwMn81vMg~^U6H*8
zJH|bXcAj7UHd_mte3!L9KHIIP0Bhu0s7}74K!m(m-nyU0H-7%`j=6SC37hiVAdC$p
zWgk|PwaI=|S^<s;B(-tO@js@V9yNJ#BWMuQXU^OJUQd1D=#3j~k;tMPF&tQmkL5dH
z%<&kAPIU8}|Jq=^VnkDGJG;*B;1n&ev#YajQf<EL)Tv`+WE4;A1*YQyHWd-k2Uk~6
zSXcu=r<mB-TS-ZqdOT@1Jz5!z_{9$4HyHlhm?qT?Tp-9?8^sKh95LESa8(K0)>akN
z0Rj`aI_Bq(uf1q}GlHB{)ru3AYfQ7lZ(X}Gj-wTLHQ#+(Hv(Qbl4dX1k4#g0^5U81
z#Qf5C?;2CI*fdg&Z=id;=x@6xP{1I*xdpJ+Y|HH@9h+-Jwo;vWQ(|B6qs^RMhCS4P
zj(>}t+41L}2Akss9c+v?bpl%i7@2l@YP*KIgIb{EGJ5!M#ky7oM4vQT90k57T`Z(&
zd_Cv(&)vI^qJZ;gQ&jY*(~m}+esnY@rfSbCBv(eAoh6TuovlOJAN~BzR2XJ;X-Uae
zZ|}X&D-Qb&8lpNqj=%4oKyp=+CMwuXwTMOp@;oqfO=}RJ_=;LjNgRW4dbXfWA1W%6
z1gi&};hgM^@ZREAe5j9hLz<(;ELhN<8jS1Pr~Sx3j+{ADm&df}dMs-Dgp-9A=CDRr
zExSjJ47)Ol4Qy7)M%r|(#lA<)l@iOWE}IS>)HP^UPkltwc9eH<NC5B&_I<Y+>O!bJ
zl2Mwpr^%lFP4_%$%El0yow;~%jJtb3Dq50R5lXne1W3&ek>X6<Lpyi=^JXBJrc_pf
zf|}5EjF+f`6LfuFT%7sxQa_F97Rr!G|5im<rGU&_-)*p&4;Z=D0J=6>`_1Tzd8p^S
z>C-iczWrfpj_k5qpV|me4FG;kiq%Qa;!I53GQgen2Mmx)i)}^^|Dtwr>%io}v8P8K
z^_g&T{e+Xv>_WS6B37+Rc=V__LBX`&T2N7Wd!^5u$(@?R3_8k@GChy8Rf{waBtu||
zj{n==yG87G{uFuP(j~2*4fhwleqE3Do~o~3JLu^tU}`WViP^Qoeoh^3QrS>>vX1F+
zc7+q8J#6ieNu8S0JyV8b7c2_;PV>}0tpj7+UN`G*6}u6}Y<xn38m(|*2DrffZH>J?
z>>aGHAKQ`0gSJMgtT9Zc@Wz!Z<x%fC>=>7irdaL3K)VCz6x>Qlsl(E~E$=p{@rb6k
z=cW)N4KU8&XmX%|J_;S4p%}ny+oNG&O$jj3P_ZxbnrKhQQO=3lyC-a-<4WpsMn>BF
zro0K_VpgjTnV4uQO|zfV$foLI=>#_<X=4t?)@oB|<X+OEZ|gu$r!{MgrJM$4SVO^7
z8!$~}`6@YCpK=_)aFkaU3HX3kkaQ$pMbRi8o9tRqGMi~P(dSP3Y?^R#B%^I8Lo{vH
zY<+zEaCsyaxA(pax`#((bf{ZH%5g7VYyyAn=RE~>W@h8#yb!_Soz&E^PEOrv^GJ%C
z637wm^&6q>mB-R5vT@@^Jzk75LET1;M!UH5+%R81Ruy>qIc?wy(#gLbH4v0Y+jr<t
zllg(k<YW}2HgfpNa&Qs?ooe-%r|z<5@oavdqGDQv-BgpYV>?jD#Q2*|^#w@TlI4IX
zx(1o{Z9^somTXE0(NI^D^=#5)EYyRTCiUw5v1`}N!E>et4jeQ{f5?zFAQF?e&0U`Z
z8RPl$=X#u3w)hVDG`TeSM~LNd)AJ-!uAo-uT_N1``hb9kctFuganj>(DD8!b$GE2I
zz)kJDbg4hpxS|CJ4r-4(2s|imTE@7jQ6((=ED+Jj_hi?Ki>dzx`uPnfL_n?k52%MM
zbgy#ifF|7*IGLNbhPSz6&z{Y6;ugEbnp(YQD)EnoRBKIWq~k9e=|negsy^aP!eE+m
z&)f1yHunPQ4|Yvn31S%$vDPr|55xEBmve@(c|bn${ya0WXa4KgS?^K<{RTCue(o)^
zx{M|xPb0Pd2LE*9-Gw6;FYZJ&ar=P-0nBhX#xS3f{cHhZ=w_kMkm>31L*MT6qV9eB
z{^^`+K;@G9%XXfXAjy&#m8zC$kK)MN#E2A_@%?pVvyP8#0$#WwFYnrK9E~5X1}yH|
zn$|BLUU1=}MeY0aX->BU%l=1#b=V+oP5T}_n!w;O@+v=L)P3*0q_6?(74E%m<Hj!?
z<{xuX;zJtK2WzO0_?Gm}<K|a<Iv4Hck1t)<ey^A#I6fR@ysl}+>k6-4M)vpreKPsp
z;<;Bgty#0h?`hDK)6F;JP(9T%G;Bor7=8W0-MjU$HZ~IvQI*=bX;Z?3Z8tZ2c+5CE
z%hJseM^G3u@6uTKMw$O~FgDiZC1Z$Wb|EZcF1xGM-O_$}9$>gDb=y}zoBPG8{aDx@
z-yiGy=Y)<-5f5YqYR3GCdWL`-3#mtwa>J(kOVcuQEC#3%Md?xVCZ6Qpx0o}vvZCU}
zq_+CUvaS2J-mz=f21MxioqWW0F{aC+f|lB1Y5cZr8&h2<l>&9&zAe`4<sAR(Y}c^p
zXsx_BkvVp~9vj=5N5#}dJk5Cgc&?Rm-_}X}n-2S4zRT!cQPBo0oF3~cL;ycOpBZO2
zn&@A(<8VAxB}n=@JI5K+tzUm6QjAsi4sf}cTT&G@{~p19Be1~dZK<*IP4+tZUFeLn
z`S0I1nLj_TQ;22vS3NvT(nDzk_aEV;Dv5)Xtb4TV#Ejn&=SkW#B+jb>RTm@g(=)G;
zyNyOaF74*Jb*}f>dc;Cjrvxq9vx-jhU8S^_F~1>z>hIh(Gk<HrgANaBYfLaAw}xuo
zzdvi49BiSo_KlO9-9I<03+0yyZKAKc_d#Rhz|mb~Ce%r2@b7yyD9b@}-PvW^Hf*Sg
z%Lj4DFDj}@2Mka-P09M$lJ$BnwR>u~OI?SRv&s7iY_T{oOo#C;9cZk{;V>8E=esoN
z(rM{Fy}C9bzLc8$i%`F?$zQfAO`ppC-EMAyV>V!f&EvY>$;z7B(A&_$+FEb^b-faG
zWl*W%-YhULHY=E$hOvTKA;3NU#dX9U%09u9PAz^vXa4*ihw_>cJOLQtB)52DdWfK0
zsadbq>8;`Z8%7*2Vsm~Ihl3l`Z*F|a7)*R|?zI6h+9XaQJEe=HZRMGz$D(F-*&Xx#
z##a|wP`20-DvX;tRV#Zq9ZdS%Jg>NCB~LN52Aw-w#C1TDaBk#EF*P`X`HD^@*3}Qa
ze7#<LUS2zTn6-zlj(OZ`LxPWZv=%M&h7D_n10cRNA=)7+tQ~2-=8?C>(_sr2FOKru
zMik@gJ9gJ(La(FK;whp~0p**gUH!GQMy@+WeH+h3_1w^#boL6=t1DZ4c<7qj!!!PU
zSL@it+uTf2iP3pC)+F`96NBe&{o~RiwFDmd{4$-^t+(&rzm-*RokT2x_T$o1&X_v<
z>btBJ`ufXM-ZQo*S)^-O<buS_{RYmPGsjouj0>nQ+D!CK&JOgC$+<&Z)gV6dR9oSi
zT)XC<aLKZ}FCBg@>TimP4xp*GXVXslT?4g7MyQK)OSENOUP-D1@^mm=yedj++1b7F
zp4D%9iVaK~xPf?t1@w+9F5c|GeDbN^JQjT(aO&~W(2M<A2LcR@G&R*kRa}wp<^4O6
zjO(dOi=r-FyY|(1os&tf5K{nGodjHE<_MnC@e@<IB)%gpl0v^kx_kB<Ble{&o6pxu
z%zgPqk$1<s=MmIYF>9gcQXf2M3{JWHs%68qhTaqFt}Z&{tkbb%5C|}oncG=c=UlWL
z;pO$kkhhAC9*aboQJ(?468xLh34X<$)JF_Q%LtFtlY~+6bD76Ko~N8wd>7>f`rM~s
z!2u9mbD$;RNKKDkBCPyr*eq9Sdu-#G(ctTdYrCvivFx9((Y?TZ=xm|Ph%{+K?hMZy
z{S?K!;mlAYu$aA|r;BmZD9dbfoS04KvMlrZTMQ#Me%|A7Y}<%XQ~w4gbE7OG%@r4W
zY<v>)cqGmffX4Vt@R?-*CLcb2+;I4CJLyw!y^gAC7aEpd)F<{1v2<RgQ3aS(7>q56
zIYn}=PAenu!!0fQEl*X5nMUM36Mc<Q>3bH}v>@ja+Qc81U>A)!zU1FkMc+f8n)Efb
zu<(Z%LRUg`$?6eJFVt8tQITV$7j<Qu#I7q=6m2b}G&AP-h%sYoi~BxwXd5|*X{6VZ
z>rKt&cvQY?baKWjEQfxBlQXtJwy>!9afyH(B_r*sdALqOvdNm>k6*h@)?5LQDwY=$
zDS*+3IY+1EM7%Od-=~Vtk9kkf-w~gc)#|~62lxz!$Nd143)edjs2xARiQo=0=H#_U
z9k$$E*z6J$UDyKC_F-@=wQCIck%?w^gL~7X?fQp+)=7x5aN$Bfi)FVg$ykY~j?cL?
z&n8%>MT^^WD@}B_pFiIfr3gRt61Ga5ap(Sh6CQYW$Zj326@<?!o>|o6n2|@Dj+#I5
zgZy6B4U+*%C>0ds5F80w+zU!ev(J50h4vxpw|Kt4A2fYp-*@ubEC-^t%|O+ciziK_
z;&eMK<8s)6Mb4iyFW-H!y!M>*vp8_%J+?IHX+;1|m=OA<Qon8u?uI}?uKoc53%8F-
z=T&=`zsYJ2-={?ikv^yMcK>d2x{0PZ>iaD(rP;fv_$pJ{G*?$c_fmpUGqd2B%1@tO
zrcNt~QW?7Gdu^9oBOBCc(<bNKGSaNu@zG_ATSZ*i9=3WOqo2-wZQZbTc%22yN=>`$
zHcfCYvd+Y&k<nn#zK!1FkW9>f<w^_q)r7<;CnxTeia+6{c3`r_Jm-2Vm`?hT*j1c1
zVL6_>@`#uI<&G1FvveC#+}cva>eyV<yf?IK*9Kyny(q)tzz9#x42xeD7ys71nU<zz
zr?nSmb@@D{b;-Y~$JQwL(yhbp3AJ`lczo_-g;(6{_xZ@p*!OKMnUnwBxu3Xilh=-L
zeDL;)Kgs9h#fm)bh89CTmM<^t5$!gXfE87X?R)q7;)z)9_MJP|!N`Ta3pCty_aWh%
z#M)$fh$P_~p6lD2EKH)A3H%@#G%88F0nZpdKYCG~<>Bw(8sUvhY&my@0eNILsstc7
zp@;kLD$C&a&L=L;7^tg&v1IpFU`~K#KoV~M-RztBr@&zsLqp?U7zd8&04Zkrym@Qy
ztzf?D?ELf7emep(>=RQKbNsmN`9s6pb#a7*Ao!oc0Ly!(W@dFaBs?tfK0B+6$(S+v
zMB@Z?5`>}}a<ylBja&`<0B{okVGYWkEmF>(J9oj$%Wm;vJYhtZM+QvL0&wh@)t@gv
zj`8<z!0#nx-QE+lX3d(<jaya2Lexq7Y%!@uS)$G@GC{;s`^%o)ySEfLBg9sfdPMr9
zRWkyvF3NlO(B<A&n080NRkB)59k3id=K4N5J=&a})g@GOcTtVy2gj1TvHsJr>L_9W
zS7;;JISO~mzce@2$r#yvFNpU@%hyAD_ijeKMgie%c|$+HA1R<I)!FxLv9I_9yskdX
z?uKh3E??fireN0TN9{XytVK9%TEq}Sz+{F$`e^@P%@7c4iUoQ5BggNZy3*YMrzHPp
zR<6HZ#p=RTvdPajQrzg(^x~N_6r^Hb*b+=01<N1;i5@bi%kEp81e8n)M=r<L;M)05
zI#dK+A>G%j>*D72eEr<F!Z0XFy?*6N`chzDr;1}f1Aq$0NM7c*V^K@*8<G3`2DRQ|
zq90`*3-XlY%DJbU<ytoOUXAV8#=pCVoy#bN4n(H&rc+9b`n71f(?H|nUCkObh&}TT
zO${vGQ+*rUXZ+`SMvWVHXL4%4eUm_!9hxcfGP^*%Z__&P)a(UeY4$A=i=Lp23Owse
zOJ|J7iWSZ6jILe1nvj$<`dbbZ910_VjhhczRaRAuzLasCn3|Fz-(1teXO|z&FG~99
zHpnWZZE7<!v&CnTdiML?WuOHq*5DCM1z@4Z6`<W~WSi*wRj6DgS7+R1MSGsMn+tA8
z_kw8=4-2KZuyQ)_9t@QP0xevy<?x+P`Lpem2=Vj*o$IcTsd^e}rs=&br(H~$&GJ1-
zMcrPdXJkajZo@O-w4%;g-40yXd!Ny@th;x|=L{Y=FyVjjAVF`NHf{8F-M!7U(YVL+
z*f)~28E0Eig~cL(cQ;j6<A1K#>$~sh(*7X}`0l-rX%wU(^3OfBZ%@9Ikw|nw7jq5>
z|MWR?Iy>#TG{Vu*klV{Q42dI;UIRV3Tg*>=B8XOEqlx*}``k6Pwd*n&_1YIt6O)4@
zidx+Sz>6GYYGox!tK{9CJCFTlgfL6ilW`=;t1Iu3|L<?Z(K5~R<?1EH`TEw$(-`5X
zEPE1JTu`z7PR__W_i@>>Wl8%^E{!zL>DFfoK_t-8g-@aT+OO^~(fVAPz0=pH7bkSz
zo2bJ4pzv}oHET8i_9A-#j0`ML(@1j!e3BTB*f}&AH&ty5#q7R(DM%K{D1=J;GP{r$
zoBpAm+;8AwXJ@jV`Bha_@ibe0`ymMs*#yDk+9szhRFyju6GxkUqtQd($VD1-aEilV
z7N_I8NOQw#hz^dO>wGJ#MhB<(!q8sq-DzblJc6lX*KTV9b$YC{tM|-M$GdfNjoiWi
zi@qU3D5Gx-g`~-JVVQm<NYK<8{kXJ3;5MMmh!mjW0v?AVGB~)u_q_&Q{OfOXhgaL*
z;%DS`HbJ?A=LU<zLI~5r@nOu15k>v$?s-x-FuNn^5;AB0p)Jb2j0KGD=X6tH+2puR
zhU1bV>m@I&my|v9T;bv(W}GrRx)m3mjx{r%*!Nkt`3?3>TA6Vp8640V#_twmPiKzg
zGyp5;*#J$)j~r1{=L<Mtf8%w3l&ZO`T|1Om&0@&dqtpD3UP87mCfWhE#kW?El2jp&
z{;??#6??UjDts|`s9gb(bJ*Q7Br7407ZS*`Kk=Y87otLH#mE2DTsLv*ROhlbw4i`H
zZw;Ao*5NWk=<*%%r>MXKWwBsKAP~A<w{8sqN&ar@Y;WIG%k`Qo<_S7VFLuP~i}s#R
zV|j}K##6iR^<Qvlk@J$~>7JxvP1Osg9Pa%q(&b3{`3(Zw*ep0+ojL=&a~%4M?sr4)
ziBQu?U@^?D^D*yysKwAdrsp@Ao{wA9wA+qxgc`)o?LYZD=M0$FeXqq^n#fw!jXv(A
z`{4%lV^XygCQ2Te_M3y%1IK|dl!tzO)2EyH72$PXb|1*(D+X5j!c?jSZZ?Kwo1-NV
z9Ymf2x<7|A>g?LLE;~s7x_6&<-H!&U8l;7xL&qqqE&zfH>RG&4J`fx#t#Fm#`-X<2
zv9|tS5di~yvKV@bq+`*djx=^zSRKbkn%Nco-5XJ#+8lM)@ab1AAMDWxO%JX4EZM)a
zTYnDvnB!?n(|orb>F2IX^+`x0*pUm#4H`5+|HqFx32HcHsei*R&%U-Wq?dL>!yn(T
zW%VMaldoghD9F1#+83`$R7;L97UnD<6CGH85Yk9I)6OF5$;O2EitBfJ53X8yXJPMN
zA0qFouf&WKuEH0@?D~SZczo51v8Z$1?L7nZG$>>#e6}?2Y=fdHh`k`|#EHZgt}pMm
za(H&V6>Z=3_H>(TlfAy0Rp%G0K}UkRy1Jg8UjCj}Pckw_k*EMO5pWPq!T9F;p?zP|
zuORIN4!Nd=g2ZL!><kEKC^V&L_q>an)q3RiVXm2GVlGnv-xp^`UwZ6%b{5?2O~UsB
zn>J!ro#9{C?Vh1LX0calhw|l?g0apoEyYmfnI3fRsZE7NKye6;{hI_1fakmr>yuY#
z+^2W%cTXO@z7jhtep&B*F5jwLzrjKqP1wuLCG-}~s8oHpYYHFK^@NmAXpt|XK<C$9
zqSsK9@^nK*&(?wCrc6l&&e#F)2jYI`#d5)$p+_xo-84!OJixgvx&3AoVc-}wMjRv_
zmMjmY(X!<xW}w*8OKvz1fGM1QJ2=ek2oA_#lmvgs8F}UVuT6I*cNO9h(wdZL7Tn$0
zu{h&Wo^A&ECE>7d74hw0dvz*!7s#vck2}n{v$4^ag4Bsqr*44^5ibjO9O{$Bn{4l*
z71=k|45ofAl{E4bSj~rq16uf$cAY5HeTaXfI61)JAE4P&rg*uh^2yRjs-Jb>6G`oj
zq_7S$1||{4E>Hil8$ig5_Np{fP}!1|V_<>MC`j(Eg?cOP1^<!T0Dn2H%Wl4--L25>
z&MZ87?p%E=nh-{At~_`7<LlmPmmuyyx!SRFr%k^@V0d^Elstu5XH)e!=H~0bIBT8C
zpznw-NP+Ww5gEKRVSJ$5x_}uX2~czQhu*dCdkcmrrF%FxyB`;U&~z9$@M;qCKuXJy
zBdI0j2s83c;g(UWiPCPE7w8q#X5jjM%bu9&Mx?Jj?LODkRo#c3>hJv^RJ&pA>&)<B
zpQBn&oiJhKj2X>9<I2#DxRsD_H0c$3pS_?~KP^eEqCz6&9@7hEYIKA!;0r3<lp$*i
zZq*XJQ;Z2U;`G9}Chdj>V06uO2e0|kcgo?Rf$)9{kkMH7?`o&$)QedT$r!IN<W0|)
zNn-Kxr45B_;n4Aiol9@1&N_MG?n#|C?A@yc3ZOVL`)!BVUEQ01I0QA9uvJbG^hGrg
z61=_&c4~tz_kr9j@)D|-b!fgH$tu7!PA_aY0hcFd+%GUtIDjA+#3vMf!dHJ@F4r72
znxB(@+Jr)uWDxExXlU*C-aDF6sw@miy&6JnFX-k3>t<FFROyLN@bFFKrVwxv3As{B
zzfOcYxyFrabKmdpx2z``S3^GD?*kBE8e#MP%0;f3Vz{-n^{!V>&8fJLSiZbB+Wa5T
zWC9efL9oZh(E;CN`;Hxd2yxC6ArtfKDTsA#l|GFx<*y|ksU}ZDS6EGJ2YpuBgWP_T
zZfwe_#T8!i=8y@Kkva>5dZB4o?~>IuO(Lr#>O)D(nm23qfoWIy^sWvEJJV>}cJ0{I
zTpRuo*(NPYJYn#x1?eUT#Ow>@TS)3c(h=G#PCdKa?AQLJm$~or$!MncAY{N)ha`nU
z5(&12*5>_|RD~2*v@_M>p^`jZc%SqE`duQTNB$^b6~Sb4`1C^)+J7?i1%tKScik1L
zr9sDzl!MgRUYm(^71+^n<HnIHR6(U&QnBYNrK5a4?iS5W*`d#blZsf(<Qv@`&oYBm
zOR?W^_QTd#VHg<-MK^@zRwva~5CaK^f5`XpM^X|KAadBSn~UG~k==nrRQL*5NvLIF
zQ2@@Qa4FJvRXK}#w~`JA<e!hjD5gSd_-S{&#*LFLoZ}X`za7=YSRLX4aeEQ9L+6=!
zxtI<i@R8nDp&BUl+O{j;BXCb~bRSR-l~*IZbSRrg4xI_G?VkcA`u87SivyfKIafEx
z_{BEH4@g(m-UTzQI^{>+k0(DELow$4UE49EN2@8A-0blvzC8NocgKh#7IoCfku`X&
z#r@Wn6^>TKBQv!Bt!T+^rIMHJq@VlLi@%cPF#O4$SO^Ac`P;rG&f~T!ghGAZB||Fj
z!ghu3FO?|@ZhgHEbKd&30&U{DjOLh%K19E-IlfKWZ9`5%s_02Rl^CQYpPqtEEP%X%
z^%#TIo88lUF!4lPa!#y&v}X{2RuNQ-L~M;y>kluVKewhdZrFAmf;ofSl{_{%Ou6(8
zHuQ9~iqBmRjY9bggl?{1%9p^8R708dLwUKwM~52#CfxC+uTl?>1tiCl`TVRX`8zfB
z=IaW>e%n^MyVvsR)Kv{N=bG`8we)q6ruJ8JwkWNrtaM&^-O!!;rKX^a?EEFuFnq|$
z6~-=x5dM90uX%JJ&re!mzBCEqN8@}ifB%<e>zCZixIEA>az3RK%7niG$y4&PSRG&y
zet$+DOi5b4?=4OZ|JLf&ix;_RwiF+fU(;XxvxsTU0|v-pkekR7Kw41G|C0S&*z$DD
z%xYNXB+Vg#$a8%;^(rT3Tz24z8A~FH3~QzCX<EKG!pg#{yT|c)gRi~(X@2G8MWRQ=
z&j!Cc&zU{DrDAyZz0scAPIkxF7Y;Q1dqU*5xmU{V{@JfpuUq?-D<~}SprM-1*Q~e+
z@fA%gP5mQsOHz&SlA@C6FWAPQiiY!|?0@yewIz@b)^T6#@#Uh&kp{J{E2su;k(L37
z3}cDFDOrl5i%^z{mdqj+Pxg1d<UD1>2sNSa5oR$Qaoq?4%7T)T`Xid^?HEUDScVc;
zUd5{gb2q->bwV2SdT$zMxXISTT<gQltWfeM2H;(Nd-mg=tOo<W);a&^wc^U#b|w~*
zt;YBq=y<^0#U-ZbtHnE+mTAklc8}HV6l7_1L5%cIL099c+d7NV`vrZg7d`UO-Gxj4
zeOLA1jNvNNy^}^z=myO#W2jv1cI&?NV&>gw=#dWDQE#K(eWbIbhC(c64sB&F9v;4M
zJ!(Gu<mFDiEWM;GWLI!-=E4hcrE!m9eM%kMb&Fep(DD1KaKkvy5Zl)pIWx|0{IqyU
z2iFdG9XNXm!)QPIK<ZA4c=FSCRc-$>kCAS7eJaPS`BJXtvS#x#+dZvz7{)n1yWG=k
zqG1Cx0o|h~T(LP@`StzHu{0O&Dq>YA$}Q{t>{-RLC06kja}yBjs;{u7m1Nxb@mF8l
zN?#_TSB0-A_t#X=K{D<@qii#XLnt3NQfLqHbm@BN@7!o;)ZulUY{v$*Me+)NFu9Ls
zorc<q{4W_$O^Ua@A65N;>1jw2P{|Sk3RN{};-lqG+_LZmqq<)j^}4rVYhk?;mS$#}
z(a%Av?JZImfO>{h)=<zcl2@{y;Y=9r0s`3{qCGO;^=_S(Eo&(<!xjk9si~%B?JXwj
zf%pk^+IR1M(dv*t-Nnoa`h(u2BBl^0((EFfU%D$ZAzfU_Dt&tG+2GK7$7pJ;`rCQ!
z27KLh5E0?kiTIzSuIuJ;ycwe@e+0_1I1fdJ-HoC6Picdix29tD^^Xq**R5AiO(B^e
zOio{dWUB07Lk~~AS|!s`KA@`8oK}%PsOP}0nsUJC4}>DM6;PjhTzUNijHilO+`<wR
zM<Ya7!W)B6KOEs?kqJ&YI7pTS%=XIX>h?@9z}*Sf!%2+bK=o}sc~_7?5096IylGTB
z(eVA>?w+yJ;w>nUKrFf9nUk7&Y5dPDOepKpm<d6KLhcNo^A;Dg39i#|yVOHcyAvjg
zjPku}3=P_!C@!bsRSkuN2uhX);S<_DrYW${%cN@-UtwBQdF<KXv**JRGL<?}6=E@A
zD2i|zPG^1OY6zL|e7Hn#5H;9*UT(pzW#F0&nsv@eh=?7Gis~jw3kjH)Eq8#o4`4br
zq@69w=0Ik*@i<}7A2n0IX6OBQ@-ju4qwyb0w<TEn(0&a)N;qIJ%2FQ-BP{UokHw_}
zBOf0bvEk|4X9q{`t>=ju51M{_OaUHJ+{cn7zf+%U+N#wSc-+Fgn{2KH0b&8IO%k1V
zpMRs<k3|vJ|DLD6BcFkbgV%m58w(#2p_wM&?6~O^>4Q9S<jBefo`@Qj5$5O(9okmf
z5Q#;bYHR!P0!2ijiCE+i%N0FJ;&#P4U64Rm*odx$x9c4k<X-=0P~704pmR4Why6Ho
z_r~N;3rBvcwKMFm84djOV~%J>K8xDgGSvLsV~Z<sm9Fg^FNJ>3ee!f@x_RUii}a`_
zU0Q~Y+#OZ8P0z4yjT)vkn0Y<vlkvV6m32I(^yxIb=e>nHGd`9X1&zO9_uzS-jkL%d
zeE<0XgjCT4Cd&}H8-nYjbjCIL(qNLc^^vw~s}{&07ZGxV^55z7qq9WJR&=C`fOedh
zYbb&<mLeWU9CZTV{oJ){7NUnFx<hcerhjgQwmHpj*F^kB1Z8AK5<5eQ_5}P#Ms`jW
z;Oa|-N(ZB%@1xhV&@*9F8XDQZp>eP}H$B~e|5lGhtjpGXA_`31H2pb341wono#F;i
z)sXp_*kUxEnKIJvf&JmXCS5$be{&7(D5%CWDdu|N@fpWg@^<-{O@A-`xUyacoBko9
z|0FVYJ5LOL7c)vGkL*>6>{)*scyJ1;z2^~LAbU<mjdO^Tg}M22w?Sho?C5zNZDR5p
z&^MrWGSg}T!eLnyd3&50oL+GfbFeTvJH@Sk*Z_jIH7^}I|BY6i^2@r5tC4cD9fP9b
z_=8r?Mpl`?t?0?(h`^;TBCv=dXC@4T$!1X17~_H|V<zlxjN2h~{as)G)_FUNsY-!O
z`NVj2^n@=`b4J^w@znrRPs%kiazH+zQ>SDzhyg8?v$1|F>aV9~-QrNIF`|2<{K~Hd
z8pr%PQ)b<Qkp^7K&GT5fa$9lP_wV0ZGVGg=;rV~>-~UO`SNZ4U<SqsV3m-;q&mXK3
zFq!1peeXn>!-@EG^@v_1TxjY8AUVrq17l<35~it5of;Hg{^=9iRZ{^?lQa1ElK1cT
z<~q$;x9(<}#=d+Xm~z^+)NN3|^cNEl9Ze~G|9+T#J?5!$08<><aO^@$Mk&fr)y<oc
z2)}Q}{s%tWfMr>Ir?$`dTLms>P<yPFSVznTv`M9&QKDgWp&PR`MNB>@Xb##xG*}0|
z&yCjf-=tY@rb4Yp&ylDQeOSrRLB-I7*KX&KG?xjuNSa$%grqp4TPJ!OtTJ+AGL3k?
zEYg=v^E;u6-pxPBro4w+k9Ut2=0=3(Or$5j-7~d#|1V#^@+>F8g;4)HbNTW=Rk5go
zH});+V=Ov`JOVQ6BKQvG5Y%%(mo8hs4>M3n!Bhpn+rufIKma(1{5vT|H--_XQ}1kl
zlA4-oYkCn?o=Ke^ZI7_e?c<#ru_pJP{qjKvlW)#N{XDEB$lre^a`I&f!%+1%J^wf+
zFes?Ay23}pWza#<`9W9&Y$-4sB!v^R_Ecpw<0(-Cf(!V2`}PSj?(c6B&L`&su#4O%
z<)YWKbT%T-FQV~RuYSG~Eixo1>VTUEX8@bNO8ZL>dIJv$zNo|HymLge9;`>iszr4u
zZ30BS__`~`+F0jjXJmA~mV5Epvw6hB453T4Z`mu%edD#r+ovt2$P`y>5Hov9ii@?M
zq<R>&AyAfSWvzbdHgDd%9~05>gOg$P$Z*aH6SkrP)S^|Z@fLLdoIMo79F;DHhVp)8
z?nvyj0jaOU=^bLJ5%^r#?cMN<rIB42skWxam|Dsk&~pnUS2?NS;nUGhl=pIQXtR8Y
zlhe>wO)e#J-j-g`brJ4&a^brXIvN@UX?y$k?CElAT@_UuVBuZt1gZDfx30R4ePQti
z_SeKh$>N~VB1-K4{A086EuVWi`>oyTZk40GH#fsx!AO&V5Qy;M_Lv+q$IC00Byr5!
zfokz4`OB{&fjn!*jKXkoL6PY{(Doh*N(@eb_#;2fn{%l-K5~MZHtjhkwMLZ|!d=*c
zI?%Nd4?EcZf^G~KZd*xef2IN;|AcZkz(!!(a4}bv=sMFUA;*M&DT8;D70ot)*SGo^
zWmn`jC@?Dni?`I?UgiL<SrZqY=@9kk(8wIFHjd&ou5iN>u?x_cD0}~P=gu@nUl7bN
zgU`4i(L*sa<lc)J{2s*$-oqwGQ%mbSW0DWF?K)`CQ=?L<eo02nZ|N4|q3`DXoOS<6
zW~R~eO@saj7u{opVp3D=L)$(M!;G_|&Yry+ylHUGqe|a<-wXQk>vdaFrw$0mL>8@#
zne%)b?>24P!!u8|93_(MV4eEcU#<Tgy7HP8`w4iEj0S%wwh<)yaQ`<9tz@RC$V#P%
zlsQk#{E4D)q+Pas`B>AZkLy*-rA|&yfIr>;eLlV=PkrXhIBf2#^eMk#K;+;}p4@M}
zWo5>j?4CPRm(}*n>C-7ieZTzi-61#cS_XeJP`T)BM>s-SVV{gOJ$x|g4;=1~e~_1#
zXSAs5QO@_hq(CUgFUc{*IK;+!-Fc#m?3{AkQ`;fMy@qnsML|u}z5v|NbUhDU^T#?h
z?UsP>xX#|UcnE$Yacuw=5gW0p8_k%^YqQ+mY^ZWTlftM@W8YEyNODqg?N1}zgBo4e
z%QCa{yU!#xma#*OuBQ$bzPQ*PKlWin%S0OnZ9cohA!u8QIVMeKSjYlP`#Wp6v;&I6
zbK=(K(KKS)joDIg&!<5-M0Xz|-Dxmoek-tIX3hNLSaINZS-pcst|M{EgI4`}0!zI4
zyEcY;%=|Kb`1nl*3wK-RcI$+4yM2BS#VyE#7B;pA#Shf;$nX6$!XY=u*l}2H>jdM{
zmDXR?J%WZ=51|0rDy|cBTZe$bD&H)F7Mir{-+c!2dW_FhE*v{H<w({I5M@z(=7P%e
zkr8$2(`VO1t>E?Ot;+~d^{bty91z{K*lX8(BiB%Mh8MG&!4JS=MeFTIG1adIH#6$3
z&W$=z;5LXGvS4H6yX~JI<u$pk>)$HspR=FpYWgplHs7MMvgW944&%kkjSk!P50SCi
z1V;=C$ku-U_N^GBu}$t%EUqXo_jF)l`+4xD5<<OXGq{T^U_TivaA4p}Xn9z)`hC*=
z7=P=<a*<<3OD2P=KtMJXh6Rq!z4!IyP!t^t!U)I&BOubR^(HM*N{R)Cd3k&LHe6I^
zbJXE^2d2CVehhQn$@0!$x=$KkugZ*nza+@#>|e%xJdi6%4ko&S7Lc~f`stxxT!220
z;;_En4%s{)$|S{T#Uzt!#b|tbu)p6#nuKOpTIR$Uw^JMrOe_#-Dp)LT1gjk$V(xUG
ze4<Vy&qSjscPe1oUVo1o8<f|>Iwz)P_SDMrx>XhNSy73`{XRyxq=p3s1oZU^imhp0
z`8K|nf6~Zh>iK<^r+%6|Wy%vM4jm8fzYEXsecJqpurQZt#o3HBg2}c7fH2vNIYt^)
zkjX7La_F|T%6O%oKcBakBd0rTttDvpptav(rW_h_7HOggr@=0!PV~gPPV}y2U1c;9
ze-`gouC$SuNwNDXqu*Dvveq38cJc+?^=>6wGvCFp%rAWBSlW15Ze_g&4Qw|X<|F?<
zo@@^Ry7XFFmN>;?3k!>7_sa)cjU)6qO#~N5wzqRitvg^m${7%dKN4P3whGKJ>RB7b
zKUp%A^|U80S@JsU!4a9m&5}{%A5z2W$MBi~4V7ML^K%y?@c@K{u$xgkl?W7bCPAnp
z2mH)W2Qt+T!Qb7i2fI9Z8*ky2S>s0P!rnlZX#S2*o;2S#@hINC!!%Z+>PXMKuiaXo
z_Hkxb)jww&XvHQzBeY+J{NiP^Zu3^n^SOk(>OIfnU3S5zdyKVYSX%!*Knjd2b7?=?
zx}oSkD<+K~%&QI+IXhw;kfgZ3bw!=4#~-Jz%&(PFdvg@A>INBBL0mS_u({8Pnd?5d
zRu))NATT(4`t(?gLGqgf^6h(zBUU+N5KTdevInzCHpBgq+}U=coVjtb&lFuPdXHAA
zUu7?e7~4JhL|Rfk!;Q7F4_5>z_3VAdOFRh+L`F%^p1qr3f%|rr$~D$LXu>rVE$Yg&
zqE4O00N=@!9c;cB<Qvgu-41kVWSuVYO8Za}i#{Wjj*Tpkc>H7LMaDG{O^9Cf(jnWv
z)^K{a?*`9M^u^Mr?4{cv{>tmJIy*tC7g&9}pW7ot_jzoh<>Hi@&y=El`H4AmPW50|
zIizHXv`$i>=)Sj<xDP`^9bbYz8XHxCwJF9%9u0{C>OuDY3hP@g@6YFaU;TFFHQ1Da
z@zKb*6G*nrE-x=HhS4NGqSwnCJ|Tn4|Gsx`ri;smn4wzMSIUW}&<C<z?6hD(LD-Kw
z%dka<4qc?W!x*hx`-tH;Ev*aNOFZ{M4ZlE@5HB(V69@$jTXGD!$`O6iHw%{s4R_8t
zq4EGHpn94AN0LVHw=JQMnon85`}fZGd(v_eX>c|)bTWqYVdulEZPe!xx3$ea_>RST
z=s~k3_Tc#i^+`tW{10E<OMM-0de15PuzNN-Z>(*rKFf2pMie!<>f<c3=!#%#r0*Fm
z%qbzbpTz|kSnDc`3m%3uO}y9tMwEFd4$kht8lGxo>JY`QoebwWHrihV^T%(PIvceK
zqoP_l4A<;3zCI(F@KDI3F1g}~UXM-Q7T1F1S7%vRlxn;@WaMh|uwlicdW{<CruP`o
zuix&q={`-A1A{*tJ9Vm)3^M;hWiG&FL5x=49ipX)hh^AcQt+w}CN{z>{rq*snK}wP
z@<&uAC7zRXc;}EF`7aN>`1-m*3_)e1rEdWe&WiULHZ0GKa$vnRz4XbI(8c^zb=(<E
zUrfU0$Tfrsz4~P6zU<VgQ}3F(!3&K_>L?W2ok^YKA<AiYDy91KWEg`+IP8s<W%E3X
z7B41z%ryfTZK<KrD|vR!`1Iwc8pMbRlr5oJ%7~`7;a0?<qO{6Y(-!$^&g0Dgd2ZB5
zZ}n?f=wc*8Y#<&V9I|Rgwa><j`#zq!E<Y)?c9gl_)~&65OVm7d#6HDqFR{0O2!wos
zaSh{WflK3$8Bh7DyVbk@Tud$4ueeg^n?sMN`Dm^AQp0J=>6skiG5Mv9i-^jNqn0^3
zKCaVl^%0cE<Fy;GRZKDvU&uV%qQcphF-w=a{LI6gQH|fhNH-EFw<)2)WPmajPF|g{
zn?|PjHa_v`${W{n0ya!0HOeG58CAFpjkd>UbK8B$ZTE!uJmsv7BHoSy*=`OqjBG0g
zoa{6p5y46pwfI825+=kRPlLqqFZTxve#Y|Umo+}c?qz$QzO6b#)<QTTgNI3#k{##H
z%^FydPPI~IsEXUulk|d#id=*Zi+9!x-tT#uV74lNF=7HTh(xZ0o|vgr&!1$|^xW#4
zE~PYiFsr|!=Xl!10O}^fa4^SYo|t($7XtDbl#&TRlmB`A_(gaqQ8kGzTN8=OKv2r0
zi4#8tpWMf|ah{N@cw6bmL~Aa0wMm5V7fv;s83-%QLR;p5NidnbApFGX(|saz;=`1N
zz2Qs}@W$<iUJzRhXeS<{)%(L(4$7rVI5`v%_t0Y)li64?sXffEF=ifFi}sEm)bb^)
z1D(^a_L8Eoc6Ww9bdq_7W>~ciZ3nC7N@BnIEN(8d00$1-_rc@O7|ZT5xj8o0T%PdU
zxfvL3jpU5<^g9JE2OinR_Sm+ccVT}z+`$pqmUmN3QjWq>*gbss@Q9EUp2m4cBX}^t
z{>kIv#6hBfERyp$w2uoKDSG;`X9NgbH5sdGdmS7~l0xQ!jUT_ISM&765sf|@ks;wX
zGN{Yf4)x!*?TJj7AhKdl??O<X2s}Ve^m)VaDc?f|^zEzta}OcCDi1x#DFc{VkzD!x
zQLm7{F-8-{kMDJ>lGPV)>7cNlq*_!`IKs>`6v>u?mUh7|y<H;-EK^lPWgajcUFPRT
zru7j_fWjW2Ikcwx%9ZD?U$=z1$8z>`TDEj40Fv+O3qSxeIqEP<)Z0sI)~qSc11uV9
zChm4C)3Kp!3c{m;i0EhdI!XHt5i9N$Wp!puoH==NG*h^F^chIt{)efZBDCOdu6=nJ
z{;EuX$Hof+FmPa&850AUpD?CupIZDriQ;^IPkrqsL8}*-JV$Taqfh@dYS=KHF%qZg
z`d$JsErP|VtBY$Xs!jpLbF#Z)^K5g#WR4z9u-C~y<Z5nVK}OuV;?t)?pL%`$@+IKy
zgJfork&WPTRIuF~BF>ScMj7upazxLzUc-j7@RvLY5JP85>;!zw_11YGLL5dkf=y}c
zn`riC<uyzSLDK0PH-_@z=c1$MRD(|7QN;T)k`&m+mKi=XX@VG75ofH!Xi#i;$4;I8
z;>LoKkPG*@(dl-rp3`Oje10clDj@{+qWE3TT#$1|k1n|Ll^-wpc=+e_2g81@mJw5Q
zMHN~8u23lRW=AcCkranNAautm%89vjX%1b8)IbU40<zAN-+<BBeIGdG^rPTLgZDAH
z)b`18yne5cfr7fcs9~pAtwbe)ds}wj3(k5<biEN)h$1BMy*?kyD$@1Pwy}O2=Ef=%
zhsoH0KhBbUFmkBzs&>AK1!=dAu-_8X0$x??#FD&R8KV6*EBrX?*Ol~*)h6~6oRo=l
zj2SSW9HZsYp#I{#8o`qU){8$umG2}2N|!;2z}o2iJCwzbEk5Rdg=@t+Y+JEOp$O<f
z6l(ja<n=~h->wM6|ARv9yCaM<y=TAoA$i2bi`|3v)F+b%``{J&nD;a~_@KRa?XT?P
z!rph$W?-72eQCba$EE<5L<JC35t`35;|R0_B;gzp0G6egDQ317zZ3i?yIBrX;Wd*M
z&NEl9@AEZec#*ItsANncjw-TjuTb1bG@jUh$dH`ulYn{3d0vxRb6jZM{s$?jb12-6
z<$P>-w+mB)q>)UWx}Qz}qR9U!@9vYzG|sg7KmOJFj)a&gh}w|~fJ5`E%TFSLB3`Y>
zET!1K?FMA06;%d%@0}Ds$Qal(87B4P$J(F@z%QAw(f#JP<#OM7&xyMQ+m6pvMqoov
zVwg~k$bbQH2D3b(6snZl<JohTpz#4%^I;^*t5Eb@t9+oT=*j*gKZ(5d+2h9(ZLI*+
zWWwW`f(G*+f)2#&?9Wg8{H}bUNoMl0P1-#{(qpX51?Ifjt^qN+c=cs#s3kkkd&x+s
z;GN2i-dN;fBqzK?WaA0r3d0{P?7f6hPBITb%+=yjdG6qUR6m{e=+SW*qRV{&H2!3!
z1t8n?V$4Y9!-ZmW<7Tn`Bn#anDUB}$50i<uOE7n!Dxb{uGP{^D7gH$P%>dBdt5j|G
z*`h=%JSStEq#i0m5?~_!Yc_yjU=6e8p8UV|&NHg&Yg_wUV~s6#)QBgF#ArmYfdz@(
zo!Aiz2x3qHDRzZT6%#dzCEFSsDvDjAQWOzMjGI_OFQNh>Dj*`Dpa>%OH~*M(@_u{A
zxF7DgV_Zf)aB_Awd$0dmbItk8XFiV<TkN!cR?4!sRdNu$HPl1stsIDO%{+SOkp8w+
z(tZ%14=m!6=J8;#y9$s86u4if8wbm78j>a#fmH(`Zy)0jEymYx-w2<mj(sTdRjM2X
ziqOra?I5~d1*rK?U$P|B#GM8K8%!Kb(9yN4GraRw@N>NZ)V|`ei&)NC5N`*Fl||a`
zjFPxAF@}C|Q*ld+jZ`A&v<wV0JKA}S^uv7!B-uTXs^i@33!}t428DrVb+5Kxt|Y_3
zI`_z;U9;?x#G{YTzikyNY#GoI$X}ixSNzcIUnyIu$Ft>E%HxNX6hQX~C!-&pE}lyB
zs_$puV_Eq3OBEcq$8uCt%s*7-RnZqW*X7^+C0GnTkZdWRzEIY{_?_}m6OM5YiCG<l
zxI&Xdq;j0MBJo)ROWku?c8~w5wX(Jcc0@!BHiuqRh0~wGh8yz(<Kp6qIH74Ue-;eh
ze{FW%e&{Ou!CpeOo$tGa=Fng$AP4A1j~=bK#+GfKH-jx*_ayvJiHN`IK$>XIiQff~
z?jRi<0em3Oi7ydJY<3^iPgSZAW5GO^S~E{pMf?90{HYuVZz_T+ooUPwpqzeIjpTLP
zmCgH^M31XEHIx0`=XFcV$ypM3A|&KFOP5Ls5nX4(7pwSR>k)smHAQ{Rk%S?Po9b@-
z{`*P{xm@6wyn35oT2J28@Y4}T_(eS(bvS)Z!54~K6**%vmIGqimE@>wLc@Q_XAHSa
zFqDPTAd&fWsaaqd)3)M0sO-QqibvalrV`{VciwrLOz<!Nv&S}%Fb5`DQqg~0d8_9h
zj61z9jFLYcn<=-HaYGE+o*&tupDZioaMyLPe7>lsyKUO(>%Y8NXPS3I(&N~6$vHWp
z`G>qFCyb~*D`!9$Z(?8$P+@mgS<9v8Ws&xF7rRn2^Qn^I3IHVLZ({Bp<{*Dr)lIDb
zVhu77*cXFDAt}XG_S`e~=o0U|E}*@y>UNU<&$8p}wM_SE^0t_kqppZs`7OOE06ef=
zR^jvRd10u$VXTvEc>6Yllk^1NHv@_@FU9FKYtkf9dWT)yFOMXYe*B2W*2+g512`!s
zPV8Ww7Z#c*8`<mSmGR)8f4}m+@=AlP-7L-ry?^p037D~tzOcAUNS%f!&EK!=Ag?T4
z#@K-dJyfy35ws*vj+^b8teay~<wM%Hk=ihJLsz(ct%jHSOHqG;ButhN6{nbvb)I*D
zjpbQ;PaThc3lgPnSy5?<BPX<haAvZNjw-v+vVlD35sMh3sme_7l|0Ph>mEHuix-_3
z0b$&lS-iSmozJ6QKRG&;8_5FA5mMH$3%8h%3rRdeIxvIUFZbi4&K6gr*arE;1KQ^W
zbd+au|C<~FIxg|fA0{CEzYm$pqnUh&wzI+LWiAmHiU)el$mrZ6vqG>SnaPc{IZfH;
z(Q+jGo5Hy*lz>EgI|ScfhxafWQiT3R<EY<E>Qyd@UdpQ5lQ2|>^CwMjQyq7CxU9P8
z(_Q~q^`DguJkspoWO2@{stENvM{(uwUeYV^aseIr`%O10Z>pjStk@Csehh??K{|TB
zFJr3R@o_ycr_Pk{3yEBh#sd4JF`5e5>kRj>gSQylaN{6u@$};=wN*S0?UutQt-+3O
zCPlbO2ONpT^c&W%U7eHQqdxLi^&Ht2_kW$g4*I6Si4^C+Pz=3uDs`s+IAAe~edV&1
zrH@9Mqoqx$VXEUMT!k>@V#HJbem}*)P_HqSp$0Orcx5;+P+oZ7B{VG~=FK27ACJ28
z;2)}-oInh|zoXz!pf|iE5b&W8hC>8sM&ZpmkG~i(rt~*Srb#FsoMOyk@+?39gd7=h
z<RESS7}M<gD<^Cxovs?9r1As_+Ya7f!ByL8D_bpGu)rhKOrBt@0?5bq;#*{(^>?zB
z#rs6PW}kkVd@eHk^#8jp47<o)A2*WfQM`)8s7`V`9eM*N<uC#u{e8sd@ND*$lcZm>
zt#5b*6Zs-BPU3-X0s1{Et$E}QFu#2DeVym#E&PIFk3A`=4b6H0?YgnJULxkThgMt1
zGruoLA@-1|Oh+57Lhm}t%i;}(`-jJ={>pclfytTu^s0yXBTi74QLx|1UFrlyQA$T&
zLVq7%lLt?Ki(@_9a$;<+!yHR{_|P?Q3Y|q<fis|}UaPwe?0e^a+fQZKlTRtZrt0^O
zG4_Eq?9o)QqvGJ(ky~v}`)!Zd2SW1-5kt5h59w@Xh#?HpwV%C|>`A4=kkMZ)IkNl#
z@k^unSL3D}r0-KcP0S@(@wdx=EbR$8RFxL$s~D?5AvIQlTWw8FZ5oPOXUUO2#)Z9`
z*q{?N;_U^Cc_q@=T^zdkl()X*Y4SDsA1vteF|T%oIw7b;_N_feiBOa8_u+)xRCFFS
z;t|jDo-wwSoyu|)?dX`d*GimWxoCjC((qa5ILhU#R6tnK0{PF$Hfq~;B9NxgK?KXt
z$!%vf%l0@-AuCFXkI!4&doAyZ{TA*<<`t=y8qR`&nuD7*>w1+{7A}%e73ntZ_4Ep|
zy63KfC8K$h=@{<$deM;2Th{H$^vIawodzuWcqCLGe(caf74kWcp~R=et14-m?}GAY
z;SY+Hh|OwQf%ufOdL1C9%$nz8Rr_X1Eh*LDPPyj+2$76CgIaN5oolD@5ezG1cZM&x
z)kd6`M6Ed-Nl5jH$+xcOZ5_7quR$3F;_(Plyywgx9zy_IeaTM>V4dssVnl`{+$w@_
z%2p*fGPlXzP6qH+4m^TyVw(i(pr`tNgxh`+s8a-%A`EDM&b-^S)eb8l8DA_+ogp;o
z=Lto|^!mF=ngL1zU^tMK+dj}S{_~n=rq`1s;};8P3Mgv1G}$!}BFE;`yj-&@D5#{k
zzsV*LQ!$Txj>hNiEO%2K5Q7{BC+ww^V5_(@`(F~1o!YiI3qe6zOogFA2UEZ@b+TBL
z*|=XBEBU)b)Ziukye;8sNFpvdGx192!5AcPZm?wT%4J;luDSBAQ_#>)mrn+L)rB)g
zb5*W<Onl3cyj%>;*80+-EdX%WY<tJA&2rC>3JmwqnbOKYGJpXg+v9wy<Q0k`*P|8e
zYWpVC_hiaP&=Wi+88ZjniZCPT+aGPGI3dPY_13R;jacFx&{_NjN#`0jti<UghfFuV
zyI3&q(rO>qWxpytacwhON1HTj^#ivk?2Rk#4oE0~zJ;k3gHUq3&aqvbVKhqE$bb56
znKS5ul-}ar3Bap4yB`<~3##$+luX$_LB9Qbv6Reh<#!UuP$CTJx7I+gW4Uv}Jj%XT
zGQ1%xLiY6e@kVZ<_=)9wcoVrUAdA^T`7GqrUknIUcG8aeT*wF3p^f*PQ?g6s`Q!lr
zBWsVLX0;e-PU`QK2q^Y8HO;uQYqB{+8V!=!VaxKy0g!r}WvVPgqC_*J%l`IC@@agE
zdp34R|NVbS`DzqhI;w`NHVwoK;)kJ{dSZ{*LW_et;~x&=O3cLYo3B{``S2t|u@1I$
zX56C+4W+=ADz?rg%y>1=>CpiA9f_+lY)?o9D3-30+yg0nTQw{3D&O9x*kxV{x%J*5
zLvNGs%L!LXYG9eTC>l&zD^n&>B{%)eNMe!wbXn5YwDgiBy0THMCA$=B^E1$KlCN8v
zGW_k#y%<Hze_E1o{_Q39nlQ-}Gd_;_j+Dak*%oM_k{zwruS*QRx`Al++-#%zv46nd
zinjfIBO(ekg-(Lfuz-0VUNd%G!Nkf~w%H(S)61A<72KKk;bKhX&3?TN4HrFhd2L1G
zODxdj`_aYlIy-;`rM#C2%Me(?ckMOv2%)rt4Aj6p1YEoGO=7+_FhF%={)k#%Rf$8+
zz<0QQgS{N6xoOZIh8J{dwQxk_TAbVZArh^EQ&B}O*%MIPwtf4B6Q4P|y0%;PpkVW+
zO<E4O;fzUxfjELZZP?B0Dy!;6MSfxzNtnI{Tk?L?sxvIHc{*Y+N(s6rro880K^8%o
zA){ga3|(2U9SUs-6g(SJ8vP>pJu5+R#RbrE`gHA{(+A9&Of%ls#KXk;p6Mjt2GKjZ
z)GPU8Z0xYI&q}wN*UwJL8$Ms9*Mqit)U$mqVK^e!KuS#%qgxUwHJ>koIhZ3P9-|?&
z2zIcF7y!7Yxw5zfXZj412gO7l+;?3%?h6c$RR$PcebHYspCd|DqcUvCH`7A|Y>s&L
zae*MmaZJz#f~D@oW1#w)s*LODVtEK>v|v-8^NCa*-8bKS{PcF*6rf~rrc>C3()AP!
zM(ByFZnmH0#8lN?4<-A4RGWCTUhKbOd^cr{aZXxlDAy05W(5@QR?olcqXyx|fca-3
zK+3T|Th8v}@)snkv~3F@@RGG(UEh3FPiH1!qF!9Q6UJNz{E5I7yJCJ!Ic&1eDak+S
z_Za6_!DDYc8Mk@;Pn$cmXxpl9XPtiE&XX}imA~x&vRS<nF(wrYKmqarX`6cnF;^fa
zy$D;S9~iQtot~aqyw%pZ%rm?x$qA)V@(qSB%#)>Te|yRLTofotI9`5ahFkd4n?Gw9
z42%-|qfg3{@v+0Mv@V@Iw?TB<R`p8m^>}1%Zaq-tzh|MNBVMjH9P||F-Jsq&1<6b+
zxNXaV0wQketx>7;9n@n+&Y)GtcV$xV8arCNm)S`3z;MJa5VGa7n~ugpA%FycPUdOJ
z=iq~rz*rR5oSS(@qKvIa?vGTqL;KeGT<RI}VE6tDCTJLgJ$>h1o=ToB`M@D_uQ0^#
z0dKp!Eh;Wfi}9%rW9Ce)1;!sYPN8p5y$~v-<wHw>zVY26cW_+(Hf_t5wETMnzI^_C
zXfeZ`a{w|@^GCcLA^-@x@COi5e!Lq|K^fQMV7TBGAl{TOql%2Abe3K!2ZQFvi)Foc
z(^{cNz`K@{ly}NQnhmPYt~7?*7F$7+;($VKI!7Kwf9$J)18498;ve$Ad|1p2I0Cb{
zI0bK}heUoY#L-iy)Le9e1Vi+sEF#Ywel~w6%@v9BQbFNAT6jwFd5i5un(p5e`0Xns
zxTs~t`VpvL#@xAQ0b}m%?Q1h@mZ{TKz%Ibi(-0#by0pZ%?qlcdBa4KFS)QFzjj47>
zrBkT~az+E#79x?U4iclR89jzmG_mSh4O4Aw9_&qTqo;S0W?;`RzwFbu?<7_ak5`0z
zk}rpCP*&0Yv_xlh?DrR(sA{Z#wO<?OK^!OL$DA+q-o29)sF*AYaT0!#Fh21+r%(q;
z;U91#c;W3aZ3NgC_9hrK3^b4&X<*Eqdw>R$@c2sRDt|0`Bejj;1z3rcCcS9&36#r^
zze()7OUiG6wZT%MDL=8}rT9~R{I@Acm-2hJ{}||v_>VK6_kW!EDtoal>wZe7Ntvzu
z{NEe@-%@<s3jZ5-#fNMoRqP)QgR9@<_yU^N9livMj`o+~8l=(*@=mDr2?S((T-$ow
z*-MvF?wOXw3?IUv;G2#+W;;5bz{Ph<$vg8YpOXQ%xxVA4W8vXxd0xJz4jcR){5BEA
z!>7gFsT}{rq|_pjDJCPe24~`(R3uL~slJI~__Trms@4v6e|VA>vgeiVs{F^;(x;ME
z_EbZ&PGoYDl*f3P_!|#h{bH|yGnXoSjf&-C_kO!E`oDdukw94Pd6i~?aet6uYNuTW
za`!P;b*Ftj{~6P!9RXrfzu2#uC-iXW9_s0^SE1iV%cl|w1XZ~97$5$Y>!Qg$b*%&{
zQ+_IYeK_#(SJnSVOp#{+RDPO<oHS%;I7_y~#UR9fMt+ijidi^|vvK#x-_Wf+Lli||
zpy1*iqZD{>bWu$gJ*dAodfde&;;k#T@Q2FEWafrh)3<>ZK|XCuGpU-p4qH8uPF)I8
z#Lyu7f)DbAi;iwKO6FZCv^{ijk1Fe1-kf09FFJpIcN3tGJ&~Kqxlhv+O`!}(@Cagn
zbRKH)Z>M-Ioi%mpVX3s~SFTH!&*NxO{R_FkI3F9;j*JO!CtY1#oZMsE@a4MVZr}+l
za*HclecBo?7c50-51((<tXT@2mGj6eeUY)kd+n(O2B96ZUmGjHpE7hHMW`fC@@VjX
z<^CoeuVNm*0^}P$K*9Ph?&7OE@W~r2%QthE#PIG#<Y|1o?R1FB88bS~HHP&-63I%<
zw5$%XPX;IZ?5Mu+66p3gwj-|&_zkW$w0{DACj${OX99~$u5Cz+5GllVO=zm*)m6ny
z{QB&;_2fd|xQ=@lk9(!bHE-3nskwl|0no^N7(PvVmmd_$q>1#QxuqF45&L=H(skV?
zXr%<XmAkb0eP8T4mYaR9WS)2WMv8GZ<t!v70rPmEC*WC7S^3V0Y-oIC%gUT5iv|Z(
z$2EU5DYjtt)z#<DM?LQ?m$0$%RQ@v*4S{VFr}$M_!fnoU9Sn)4#r@rR%_a`&;d$+H
z*`W<TiPN-y(*kmd9R@sB=oMaPSJai@2IACRbJ&eG%a%C}O0L&_IO&G}@_o}S4OkR!
z#+qO>jnz_j6kgbx*DN;W7rTBvek*F|AiW*`rcU<#&QE!pEq<dk!VEe!J{a_dj;Nis
zoYNe5KSH%M#3yCKuS@Fd##5b}v^5ztD9xd3Tu1*C1FqSP3&?1u3%DLBsm(gl$q7SO
zC3bAn=JAE|FfO$y)vdYT>{ogWYArb6csCoBSxSuaVW}Tzi_YTF&=nnsv7Zoi_OWD#
z_1CxIL!ad<QbH7>N3ffUTB;n<3C+U|dCeZ}%#F5d^q8ksPl2{4g03wan++D}m{>(V
zrmh`Y+GW!)kHVs&9XE1~Tk4O$*w_D5v1@h6TDYRl-nOtUQEb8WS!}Dpn96$qJZc_t
z#*C{VvQ!)_@g7h3+v^hq1FZI4P_4=RbIt!KYZ!8Ih0nF-e|YH|)|Y(3@H0>->7c;5
zkU|Im(zIX%AdWWPTT>pp?i+w)Zwswky@|FrGELPdsv7wg+2kZsCxcS=q&G>Ys2D$a
zvB;MEdfxlT9oL=zG(M)OUYDR4z3)Nn;oh<~XGb1s%U*pR|FojJ>;!3b`!|I<T=?^N
z{Uj#$R{hqnqZ1I!=<I}?+SCV;SJz}rZKQv=W%R|82@wsA2env}W1t2>>^!go-FD!n
z*|dYAP+<sXNDqL3@~!jR`8f>&W8iF4hfPX8l`(}mA(^jqp&k%GRr(y;5;w9V7v@)K
ze)dSX@yGAAZCC{rZE=}0zPhF)UCvEyRcLMJi;jB?u+08eORYMk+C#$Z<gVYe9In^(
zR)i|#(27jcgOlh~m(9PIbthvB3m-wvApAp7-w;AK2uqmPTla0%OTFtI=zk#NPG;$$
z4Q1nKeE9FV^2tu&+erHlT^B#y_0%|>?prMcj7A5)eKr$CC^bat&f%<l-~6a0owowQ
zsG-m8h%=Mpz-aC{yKC0kavCR^vw4;EOIL<u+*vB`FFh~+jxCx`eU=au%KDU?^*g?8
z)29Aa{_sp&c?Q8vR%<)-vm3scU_Eriofu~`>Ij2__dB-ugU4p^ve^Wm&T$=M?N4}J
zYu>oTQ*Vl-SROyo>FeiP)=Zl8QeIJATAWqgq%DA<&~Afkz#Qw+QL-so;#a-jK)~7H
z2)1nQrI-8TLn0#5D`Tq)H-}x|0B*SRw0~6~r056#K7jG--wqEHr+8I#Y}dmt=qAXn
zuC4Z~EjdtsCx3P`2%#XeVZbslqSRW~!Tb4YBj`Q#wGR6;FQ8@cg}(l`+zhHi;1K~j
zUKcuR+>Z_#jYfRov%E^KX^&BeJJAAP7vac0nM(^_LM<atd4mkp?0I^(X?(3v`0v0U
zawByPGiTF&`!T?7K}hjRv@IQY%Sjx1fgsf<K5X&}iP??%$qzi8##_3XYi3ZP=WKrf
zaji(&T2NlM7s>Ix<-(`qGS6Cmr>~>;dn)LfCtq5i-ki9-DATlF4~I*cE2*)A%0>Xh
zXbO&OSU)^syID+Zs);tES&8RfC#fMy1J-_VJ5#J5IOScI+d4P|!yr8^ahm!~js!FI
zt7yO;Se2e72WJRj8y@KyUf~2Ky*D&PDnB?=i{N{=?=#a0Dl?1g=slOH{wLIM7r;9q
z{;?@iVFA!Ml!{C)4>rx$etxwBMl@nkdjEav{3GAc{0~$%{_1lO0AM$LN<)y8BHA|o
z@`H;DH<!I{Yy``tY_+pC{zG-A_>v_9^*_lDM~6f@G+ftu6>&%^cYZbK*Tre4r`673
zkFTq{0#uE-6NRl`mWPUqHBj3w>TSMxuNlwwGq12cA*XPq=E%O^;@|xHjcL(`(!|Rw
zeq-YGJ~yzQ>XFh!QyJ$5jw|gyb7X6-mb!cb>=5je%=FdjPHji=EXY-&ZP#xotE?&Y
zvtPXVr=R!@MQUq5=7ghX^B)Fp)|Yy3?^NOo_~7KLVk@ijq183n>pG#%vuw^piq=~X
zdaCRk2fsGaQyE_=jTZnZsK4=nFHVx9R@N40*|^TI>Y^)i9NIr+UAxYocGq|M=EA`4
zZ;slJ?L4l@$$Q;<O<1dH6fnx5zV4gunj=Ts2aFmWu4g%U`HP13Nhg2XG`P`_oe_FN
zPLJI=z2&ay_D;i!UzCiBvV3tjEG0f;SwZcdi5E9ioM$+FNxwhApM!(*{Qlr0Tt0qq
zgpW1+7ycoybWgU!JuD(<$ck~=auV|AxByzAQY<^a4;MU-%=y*#7wKph-DqV2)mzKm
zDulntY|VGbkUE#WXJ|*#bphs~N1dQkF8@EH3kA%oyLna4bSkNEGO}2ksP%%vDP1$T
zFwav?Y=hAewX*Oi6O6EFzC$}8YIX6xgz5&p_hf|5n}xXoV>KYAGV+Hme<9P*spPzO
zqdhFGtxwBs&#*89P&;A14@3#Xr#xwV6#AA4F^69Hp?xK|{_~8`AU@DQyl?~_1GMO?
z(Z69_&Ufj>{2YykSvY|+C>MhM(`m#V$2v6VZlO{=yx5$z`qvL<t;26`UofFRgY`+S
z?~wqxTZx3cVKFeyqH46eoxSNkJ$vitH*QPkg<v<G*3Z`?F&PK<A?~t!R#Yblq<_Wt
z1?Gi|QLeF2BV=}(k)WD+C%b^DF51c<M2f4oSLpZ^@0`z@f93$=wiBVD{FlaR9o`$!
z=WPxz2kwbq6!-H2Va=;^NHauo;B{uG<%!ATSfz~W+F`XM)vt!mT(IESGVDBp5-gdo
z2<$p6y-1H%D$>f~xl7HOt96%`<?Rb>z486P-TU?}p3wjEmg6@+uBWODb!aR_%F-Kj
zMvo$&;^9_Hrd%P^amMXsT9S8&yG&%J;Hiq91!_caeY$d}kB;=+8Bgs=tM5c?cr$mK
zy-D-Y7->%BVV;R=GQ$X|&VfkKEQ%{zM8nuR!u6K~J?{3hyf|u*$jmazx)Z1a-FLZ~
zQkeSd{2Ro({6UKl4->>Z;|-*88BVKt&~%J1TQ_YQL<*J4-Gj|3^TjB+6w*>#el_b^
z!qYDC)wZyVa4BcRT3A>7uaVqM0Qg+|$sqiF)XWB?GMYa;8~V}!NeK*2p*Cn8>3)#Q
zouho$i)7|@?$QZb>Zu1u1LhtM&(Jimr0=%%^C8{UosI4NF(5xl$NA6V)%z2?Fx7Lr
z_v^@3=H*j&d{}#XA$Od>{I9E-akOGAf`;v#;?-~&C&RG4osf{hh{*o(S`yGOC>^h>
zh4P0+$^I+&?)<<4IYdnPrT_8F;4y^uv&_~(0<mJx@&)Bv7#XE=9rjZX_4^PFb<z9{
zBMZj!R_l>}k&W0!tyH^5+CUza7a~$nPCG|P+4^x#F05X`X93rK6iBr+uJ-lD^<a^C
z2zlwp-MW1{sxbc%+70L8F2m0K#e{CHzC*~AAomIVIa+}X*LxoHB6BoZ{$#>vZ5f{5
z1FO#-3VXV7{hQg_?7Jl`1r@x6)ur!ivzUHsZ$JH^Y1_M=cxE^DO$g52_=#Fav-_oq
zu6T79VT=8;P4_Mm2{aZp?gC0cjyhVYaSnHZf&bcz27*^0livdnw)*R`U$>;h^!FaW
zw{D^Gw`nm<AUppqXo97lvoYP~wq=b4W^}<Cyspf3G)kv1rlUQE#?x{2Rgxo^LOs>=
z0xZD?uEl@(_16<|Sqm{Jq!(NKq0K4z=vQGiEeN;nsfKLw_RNyQ%jZ~H3PVw?l`b^}
zUqYxS;5|&MHQHkci)hSS-T=d-3Ht115xPm$pPNFxk)i*%2s{Visn5}ZxV~668S1I_
zm~f_;KGBc6K&KBUuxs&0`o62DIvjkMl#m!>>|MDu)NM<%gLvRYXPD8UmdrmHpMcs(
z2Kx3n{wkAdfU=M}jV(aE>*i7k+zdA?$^q-Is+s7Pq+it9Rj4b2UAG(Ha@L1AE;eW?
za5!j1?j`wJxYyYwfWvp#d^6P1>0vGW#8;ir1z6a@AEC1suU+%GsjEKk1DQG5qUe#!
z<bKhm0SU{h-j;-7gtyJj?N%$5s%yiBB!sF*>v?EiZZ7vKda0qdj)=PeMc~N=kMzrW
zun9SPR<^cozU1gkzxUa5XW^l|`&h(B+)QZGr1*k<Z9{_p(B)4yvcJxw>E)aq?Ww^@
zvK_~n>R5s_oOhSxep9cw4qpv)p7Z+p(%b;gtfhm`_iORv9Ceg$O;pC^^#?Pu2ew;?
zUYt+-S4Kwh`DT2z{C8K*{q@fey8GpU@--V4Ivm<Gzt&3rczgC&1_Vn<z!0wVf%zbU
z74Y-JX*pKB$h4%varM-c!^@rZ)_Z2@yJTKAEBVgmS#sqs^_PXFKWVJ;-x^R~VrAl1
zgeB+O`2no-ZytOzX!ncwF%M>+o5=t}A=c(H)WL{Cm}_Q#buw~%R@T;wb&Y!O*^+R@
zocY3>74FgNs^Dxb85~pg%N6JkzB!>iqt=wT+tm_vM>w>&k;tZacyw9KKUV&5S7`}9
ztDpnB&a7?N7&N{+d_CLVH|K)G2G6c1=VsGReO7^;Moo6j{ma)wJMLCbCvd|f_)0?8
zH=EbIuEJ%&?%Apvn+u|cFj{+Hvg4VTd40I>R<>)W9Iw@|lg>A7YdAs6AdVHj-%mQN
zauxP7-4^c-R$V%FyL0ExUq#<MO?HoTAx>*Wv12jRLEnmPjAhPy6hVb0G(49VAa#EQ
z_e6vx?ex2dG2GLf>7Hs5ifOb;;>;o+oMX62)AhtZ&qUZI$hG@2*x%A?5e6Ql|JOM@
zXPM3Wq;X57iv}*W&LqtS)$&~+J6y<NDE`<+dnA2rn4X$ifR18klkb1%WU=&7?Eb38
zz4uV49P;E=b)9r*_#9eCq>tM21d~PpB2_H63=VOX6o-5e=8I?XQZv}Er_jE<RDuN9
zAkS9|=pcQo9lGlIF*1Ov$l1Qqzx6!0bLTPPXK({)Kg`*|WbR_Ld`2|Ob5k~9J>|3<
zK3IX9BA3jd>!I*<1wi_nIG-eRIe?`Wod-^!ke*K1Ej4^tk0x1fdN`7r@PO-tN(Pa?
zK&YRhO}(<hQJA2n70qTC0or+4=B|NG{q2Ma%-Zj``8`6UdnKoWvNDn4V^USCPW{Jl
z;-0+~_6IW`i*mBBdAabPEUM-l(3Hx0Ua|Nm>>~<MncOZQ5L|{yl){5E4Ct8=#~4yc
zqS@uMHSPX;_I>NA${ieFPvIstJU2vu`L3331&K#w$dU{s;WDuhAE-Ya=90VE|5o2E
zJ9^H+p{o{K)k11=Lg;6N1NDyNjEu@hKG*;7>+G-Zvxo7MDIR3VU!CVuu`8pZ)j#k2
zfCXe3!Yo5Y=oC-kc}(@o9W>YYdl(R$Ll?&d?uOJ!+z$cfSvK0V#c4u<A!%`Ac9o0G
zib*+e+EhG5ChZ%j9=7IXDhStc;-seQms77`26mK0fsy`Tm|tj>$aDwRD5NuqY&cFZ
zqcy#@C1}|jjC=`Hz%cK?{6|a8!sw++wt?>Ar+?lL|AB6vo!ud>=s`&IiI55%D_9;g
z$eRGHgmoG2clgyOs@QXPzpt#`G?RFu`l1KK4DZ%xVQq08Y_r|5L4#TN5inQDI4>}7
zw-<#h{-PJk;ty?dHOu`_Gp60U@UM#>gU$`EeaY-BFAp&9sZ!P7qy#B>-yeE?D#eQY
ze0J1qox<$k_fhSX>s0>rzsA(1Rs8Dz+1R!Jd4;lp<rPh)DZxR0{hxht?I+p~rS*4>
SYO3&))ng}*IX}ui;J*MWvOk0X

literal 69408
zcmeFZd05SB`#!wP5t%lX2AfKVMvY3wlA$P#G_Yu%XU)dYs>l$Hs5F-}X^to)&84|i
znxxXG@jdTV`}h6+{T|2f{p)pj?EToS&$>VNeGTVzUgve+J)<bUX4RHe6bfaH!ttZZ
z6v{G33T4SfhUNGh?ZuTk_{Z`~C*_agpZN9nNR~VPvC{ImrVWL{;Yt3t$hULo3Wf47
zMd9e7bB;m3nyxyY8_g*iDm^7!C!)DyN^XPSx;r;rAMB3TS!EWl<zgP2=3A|1a<(x+
zwKu*mzWPACT2B1qSmlU2_w0<08Kp+qExPw9;NXT08;*a<HQ4Z`<nqtiLtT$puHU);
znR}piIH-#|&qn``%=XJaMe>Gco;8}MpQkSle}q!U>=!Nk9}kry`HTzye&C)5^TNNn
z{If`4;a?|qE++3sehr>onmzw-l-4i?$AxF)%da0?Jpb1ozS~=uE&RmRWB)H7Je{LW
z@hFeX$oQKgYoMX=bb4l{y1Ch{K9r5ccRLqX=Yac2?ZmT*{ft5P?%!XxY{iP$h=`93
z5fne4+*yWItNgg_Y;8ZaSvqEHr|o$0;>9WkhBN2S2d1BVx09M&cUkyZhjm6(eSL4g
z_29Cj$Bsqk<>|`%un*bqk~rq(=63JiJwNWz@$q8wbmxqhBD_0x{Cu>vc=i9cJMe#P
zYybZT|6elS*=ByY$6v`{#niO4n23mD92^{P8v1*CdzBp=(t{Ku@q795<=hIJy^dYG
zc5M|ClbWUF>+RdO|FdaRM0dC0F?spoOG%D@jp^&lkA@q0oQ;ENc6N5|77$QZRec;2
z8>_0K687rVj*JTl0@Kqw9x8|ho>ov$IC1)P_05$VA3b>RVr0bjmZxXvZNB-}G}r(y
zrfrXpkKes>=b4KaUz|8`!gjFrUFC)jrlNlP*o+L#l9CeqHx}DHl9HM#uTGUe+NJWy
z-#@ad>fEzauSC@5L&L;03@6`<oyl|TcT=ocLz+?7x8Rt#xaqG&PfPT9xVd+3+jhvx
z%IcnvkJ`C&kKPv)Y}&FVCO`k+->IqM(o&_0a?N}h8JXVxepPjK|N8p+s)~x2c#-Pb
z+Sr5yMa4*)Po>e((PvMe_7*d5+^K-))O>v=BI-P5S`r}Y#hOch9ok|<AHnxnMVeW|
ztJ8sEJ?*}-Y7P$Lue%Bh3$aT|4|$c;)YP8rIQ8^-M@L6xy?AY>bu_+U)22;i6E!uT
zVFPa8EWw62bp<<&eA_?p$yAIFpX>^<@yQ*D5ZDEOPpMu)|2{=dvU?Q&;)wU}51Tc<
zJsT))p{Ar%YE^pj<jKk)Nt=CGAC_?%Z*Gt+mRuLt(Q%oFm$$Fcrc|_Wzm(LuGiRR5
zIFA_?`Eb;yTZESLzqgUOb?cT^hIPcBKYwgz|8|ideOzrU(=JN?N)HlLTUQ&KnN_#U
z%{t_a)|`(?Ox$YScqz&7Tae`BAOF0#<kAD@ap*cxe<ol)w!bsFw!BDeGOxv+q+BW9
z>^f2|Z){>B@~sKqIM&|~`It}Lcy{{l<YaAL0;`BV-q$(Sh5mkp%MjQlN{|`ud3lZc
z=S?3A6f@iM=6FC}Uf$+C8qX8mMq{<!TD~}Z<VeoBo-S&tR;pq7_Ptsu$NBwFI*<J-
z&l<>$b!fgS4;%91j;xHvUKmxrI(^UAx3)HK*3PxxL|5hV<#7Ml73BN$OocO=%+u4o
zD1w=e?=vfR?b&l`c6Qd-!a}ycR7UIV1%9)-H^&mQ|J);o#9mrbQvH!xH;(7z6I!&`
za(v}}`(I_cbCc>h!{7EFsaMk2vpLv(_HSl~-^qO!-!BzVl^i=NC-<dguDdJ6;K#ro
z%iV_$AO44(J?_bq%@+QPk4sHXa_!uCyrR732$N?)@urLWR<2p2dhlTW^DIA#!kao%
z;b*6;GcI8BN9`pi7QcM?vZ|#ekxSmkG3%e8?Xt46$BrFSkjnIU+hjg&eazq+vu49>
z59a)JOS40LD|@P(2XZ3)b(|DLMMZD;ASv~+wM^XR&*<22I$+=Dlbd{ke*ZSlnHg)C
z91fj}!^5`~8I*>^Tewd(d?_vU|9tt+;ak)div;Aj7VFT&mS#(=U#zp*)5GD{zpbOL
zOS5@BH?N^s-a`~=b{*Hw8GO%ZY-k9th2QyW+dt#aowx+YcaDDlfO*4)mk}rZje9C0
zcJJC%9y&Y1JJWCBF4BC)q9s?mA>E=DE^hm`)8Az5SKZPax+S<Q`r5ioaH)u+($`Mw
zo*Qcqlm&>IoKsg9@7A!lPfN(`-Azs%*-<CP($)ypSuBtj=dbH_^;4+3qM}d2+aIg0
zH>>2Y)}hILZT)Ciuv#aGLBvsrU4PJr<15zxwN*wEzI*%MR*vQb?W~cJN~PGBFE5o%
zkfoG})^@`Uc%s<lEG!ar+-L0KHIwx}(Yki2yy}jLO;C1o%k_^LmRXw3;?3YFRKITU
z$5H%~eQ|Tr4UgLK^@AboTKg#Z3Q=)!RWYiuBcFLgS8pl({JGq&Ry)$>o1~}SxX99F
z%X%tK3RLs>UwA3r;-XS|U@8t?!lQsQ)k?Z_OCLuOo0zM`%F3#fu$jImEe8Q#c+0A*
zBf3JI6)m!dZeQT{7YYdqit8-)Hz|L4OgOMaC$4N9X1G^7^Bl8|<BMj;_BApyBbCOc
zrjKk><Oh5>_NTyOw6AtP*6rSSivHNbCB*5N(i`F(T3pp*)#A5Qsl7K&?F}$VC{NFu
zG1J_$SwO9O3YaI=wntgp^4&70&d2I0#?`y)`t9LABo1wLn|Q&WH*KwL{h0^oX3#!^
z{%G2XQ~Q^AC`Dr*?rq{587x@K)|OCbTGQ#Tn~*o%DKPolA@Af{=Pk>YF73{0nMoj9
zP1Yv&w4t^(vN~41PhNKNftW?}9<4%pI1(fvaN4V&t*k!PB-Q1w^`_06x3r5Ge5>*j
z$c+k}?%;i+6|WO#Y09{EZLPO>b5yO4^Jy<;0kJ~1En95oX2$yxvVJQgs5A2|B`ee$
z*#>l|@bTlv%(mNXDfazZ$d?Mrwzet1^XA-S{&qYJkg$5`&&MtcSUb{sW36#@jOv$?
z64643U*BI`DSl$qU!NLxorydvKX2XOVwT#;sj1N_6{Vh<^A@UCu0#msV$*W|eBmSE
zj{o83j*qK~5_>8mO^`-15jMP^RfcHR6hGNp`MW7A=fXDfBucY<KBHbEaTa1Rc!!ei
zfq*AZUPXpD8J~$_7!V39(aN-q)^Y9^)pi;BSm=m1`u(<@N#@TduHIk2MlRi=KgN-H
zk${I)QIBnJMpsB>e0sC-h^=T~N#>7xd?P=+1dM-veWsJJE_5?1YcvAqc_rDI6k!n&
z@1eZ8nSsLXru1jf<l!cTj{^d>G&%LfJ0U}KT4!)`1gB=f0GeEeOIEF3Jz{-J;sGLg
z_usD2+Ba8<FDGemEA;pGXU@$|jpi>|8<msORJ@4}$OYdZpU@Zq{Vj4vpCshXIz4{!
zWJ^23RJ4v`+p^BCF6SgE7fxW3+6y+`I(WH5?&e!vM;;{PjA^+}w0Y{p(Kp-U0kFK@
z*S?;fas-E{B2MR1uCA_r;orY&$jnZ(XCka^`sbgH&Q(^sD=I5f$y)6j%nCzaATN;*
z=+6G}2O7fblehE`@kQKxnvU})*{X27E7U!=CjNZn-(jq0cMy|)vL*>?I!--2IJaV!
zIhaC<X`QfHOKy`_HNEPv{E3=_Wq8Y$Ewn_7%$^-Hz3O?&M~~iQ=6m1s1U`za*PGU0
zE;6|@JuR&lW>|qpdwz6m%qinAeW$;zh99<nsWS-O3p42#wp{M0Kb$-9!%M0!E}1rz
zH#?e0jP#tE+Ru-@^iN2L5cEtvp`{ftJLoCfo!OW0Ff=rjJ32E{E4|U$i!3+e0>6Md
z1Nrg%9#U7a*i27P+m2SN_u-3F@kwqYJ-ve@`Va#W6&H72SdQXaEGPe|=#?-3$jnU-
z6MGX<1AKr7DV|R^AH9Uy5x;^sVILksdnF_^ka#0Gow~Pa#;YkP8yH0X&Kfu%GxTAj
zwFLWr`>X=Prm!#r<w(b$PgjKF?klWG)J+Vu$kjxe(urG5p6N0FW8y!4{1|bc8+V@^
z@^#;n{I@$X|K7b-dv$X!4CIW5_{Y*CLhCRWJebCZ6Ox@Q_Z{hYsI;4(U(v6sp&=Hj
zxz_1dQW3J|K<6s@DM?%<%e68pq_H}+-D0-=CfUV8C8lVOBLft4P7)?7pwN+)UgJF)
zt%}Q6`X$y^XIN+1UAC>svE6$Tr=NX+gZ_h|-l$Lo6P}gmYg>KXq?Ky?_(yv?j{>me
zWT$|8bkSD#3-sJRZ~fBPQwle%yzdsXPMLr12oLpFiFiwulG~q+qWktaE`0lKvQ7DU
zM~@!uE|=e0{in0o`^lP!hzNR+GHVbMP^aUU{4;nN3VVU6Sl<fSiEAN_KUM;@pY)U3
zzu&G@X54VJw;J%_zrA-S8`Ua75(`qpq4~Xsjc5JNJZYcR<n@@oL4y|LnX9>W?2zMA
zPpBHtn{%a_0$Xe4x@3$Ev<!UwLVs-uQkZ*e92_35JT2Z7?#bzNeA~8de&K-7#L?<s
zU%UYn=RY51X$It(K{pZZluT+9sz2P_$fLmIc_eSDX<+0Nmv2;FUQ6-37Zvi69a2<=
zVX7tS#3ur=6@D-DX0t2an;w;$+q|&K`5!yr+1jet<UTh;+#Q?Bs}K_vRYt;oe*<b|
zDYDXhM`HE)U8u5cdm{XMqa#C<Po1K1M<*w%Am63eBxv*b7k4h0NBc5DGDI5Irx=k?
zCrT|`t>D+tUxREjDmynUi|Xl*wH<k`$B+lv0fBt|WTVb6k4Yvop}9^E6=e<;ve^x`
z9uiO^R1Ae>E%DKb>qwt=eevnN?^bLD2*1xwiU8#O<YwmqnT6wIDJDNEaeA|WPMoMo
z&Cb1hPx}$TM>pdicX>N(rE=!2PHzt(HCp$5`TDgAxG2T4?J(7JW^$AO2d(qxb2M1V
zSO51@jEszx1#4yF`V#YWP@?&rJ9jReAcmG)SDgfUG!=Y;)J3sej?!rcfuxE=Cb}1c
zrBI90cu}kCu5F&)un@xw7n3vk_MG5;a-87`54jbVELlRT2mo~5==Aj3h2wPLAp2>c
zp{(o|a`h*xwqx6^!tYZ9d5P!EGmia_!?*U4H)mOdIy4j5W+XXu<`hzZpY4Lf)n{K#
zUhwyc-Q>&V?tH#nL7<artlF*v2Xy?ZnwsMATwV8xg01@m1zQ)+GZ%Q6;<5e->c>=*
zTI1i18HL{oQVgB>WdN8zz7WK8iH&$LgL~;2D%x7R)_vj#)wDE3h8R8Qm+bld%=ag|
z)XIZUgAK{yyxOrIwUdIMmX?<7RDXJ3srzj5|1x@eV)Tx6?{6`tBCm9IbR73Xy-fmC
zFLU0+t;{T>dwLAf+ebW8@1Nre3gy_-STFv}Pb>|1W%IXf%T>>vTZjtzhH}J`cdln)
ziE{p(AwS+|<GXnAVs3@h)YQzsKLba{8Z0W`(4AOtDcSb0ug93R!-1msL`7BoIxHnd
zy2E^X+dQ|b0o5*SCv68=CGD`Jq@)uHH&OwzQ3rCT66o<zP0eXOBIO$%rhhl!)+=3P
zHoC(N3<xO1#KcsP0I-Hl%yGft3<*rl->JXgaM;Wnetv%33dqj{5ZEDj$7E&wXH405
zaYB@>_vcTjmrwWaJ)yLV1<Oe^g*OlCTqLGUNhAiJFRu}=CNolYB(p!Qf%at0<HwK5
zi>JEHI01()q&XHN@<tQu9)1ol=Fd7sGMlopGB=4gO436&39hrVw_ixQ`N!$cU2!@@
zTC-GkF8j$I%}X>=^wnxR1g~1PYT@MdxCx)w8iu-a)Ju2z@OZOx;`Qs-xfM`nPWC9d
z$6%9-du;!E=reLq9&1IOoytU6^nXqtwm~ojL8&qE@uSvskmeymHo!yZUl|$gSoOr3
z@!WA4s_ES9Oy=xlAF6GOBnlhZ02-a>q^)1HXwhWi+@t_OIj9i*!ke3QWX8TzC;yhZ
zC;0d<E$o6s6WybzKTAUs?fPnZigjn-2#vu)Rb*#}d`DsB3z;szhi+9Yg!S><>Zak)
zWaoeV#tj}ie#3i?!e;*0aVf4K?pL@IyKCXvciw5}BlF|XPzzcfB5>yK%)S(ZQceLi
zdUlfFpf8~vyHCn1lfQpD|BR3Ob0g1nN6OAblXUez@P|$?QS{hJMp^4NxC9lKTwPsV
z;l$)<Un-Cm(41q^d_*Ty6n^qE3MvFYf@O+L4@?R_&&}20Q;RDD1(OgG!nv>!6Vmim
z80^{X1<WNCual$o?Af!z31swWlf)b~l%M05<|BIP4RR7z#$O)o(#o=rPssSW<?r9W
z3Vs0r0k*J*o)?~+%JTA_3rWpoGkqtLJUl3S&!?S2z74XrvGE=vMHE3IDOT-r#46S3
z;ENQ`6w33UAodx6nJVHk_@IIx;HZV$Fk6#V5r>q^#OsA;v#%xod!HL$-!s)RmpeNW
zDQiT6*F=bnGc6<suA~!3N3Mf(YuCH~=+Qqjv*UTSIANP(a&jm9P?r;9Bd<u%+Ptw*
z2&SV5-bC;e&mCHmGuazU$92tRWzVO6XDN}8f+93WV;eEKgJhp6yzADlf3bfcTcuQb
zm}YHlO&bE+)>9-qn?w>c+35K<Y-b~Do;Zaylied}V{p(?!4I`srd=O7C6o3-Z#o{?
zK%QQ>m8`{5`Kx|%?r`uxceT21b$agT-e;$FC@7nl#JUX!wsb!^Z+;dep?@r?);xx8
zkyzovUtV4gsGBu{8r^7eq^Ho)$SB%8Z^nLRx=U8w&MtMnxbqky=Sd0jYiztgYG9|q
zcPk#>zpqELApWi69Pc(a)e;*YU*bhCP4d4J&_z*vhRyl$(QC9fHnF#m&ElM)VHEdg
za1ekMWvc4h5pea`*x2IY#{|*ke?Lk-l=~XYA}L&*W=3o605d?KDEU>Ph9$m7uo*c)
zdR|=k7zw8o#lTOWjyJhY^*dpAQZIkrK+Og!A^S`Sj-0W2lit3*;Y*Lnb_v>(<l|Av
z&%<+Ut@NM=sg{DlB@{Y3J7<MB_3om7eSd20({enSLeX7EjzfZq6C3$X_3e3P5K`Wl
z0sExhXeOjkQ&W>p9Q`0%P$l9g+y4v?kAB#w)D47G0V<I;glHl&^(&c_iz!Z{#;NJ)
z&PgZ8vj@R^0A76f8YHPgKq5GfeeSa^)ag$TxU|yEg9-aG13;ug3T^t)QCbd>7nW-r
z>8Tu@X!q^&b)OC+RRFcI&C;;o_L|I}0VcLBuH#0?y4G1(+CSur<iy=ANdKoWF*{qD
z1b!*U!RIw`nqj0E+QPzeAJ7a|o@&>pc12&m7d36o)vn-3c$S4fy^;u9B?4gKQG4a`
z<!&IYYBr1P7kr7?%580Jg-IwTWG1`ACY#4|BY`8L>F|<%s@?SCk@J50R)XMtf@X<5
zr=;v_*Fa&TbG37mwq3O<soztLDtCZ05SZtso8`TL`UHD#-n`kpzvZvb?6B;d3jD7c
zE-(V$tn4nzq82gre@{B{;K74Ql!!4Oy_k`otBX}hIg|g1KmkhF6Obg=mocX3y<1#e
zRYms=#6k3&_!R(KY8S{ZO|Z)~>Uoo@!N3DIw50I>$H0dECzU<tbag{PM08q805o>m
zN{I>z{u^1jE4^W_fAJNU=VoGo-X$$9y5P^=XzW2$RzG;qWATzDo7S!S7qw?+?2=Y9
za>V((+#ru%1)5i&svnu18lYpG1AjX<wmVu`C4r<(weCDo5-6^u`}d3bK8t2)I$5*^
zC$<43H4F&YBsxC+^nCB3FgG^#q_(!Uly%1)A9l%?2p&|Mj0=%DoJc3|pk>RJ?c(M>
z`tIGkRp3F<XG`+EbLWhs<6F?mw7t7`pSXDO;?0757P>9DoWPn1RK^CzzIr7`_%KpR
z78c6UgAVBU6r9bin>P*LU0+66WBspB4}FIf-UfqFRa5h-x7T=q?)O}~RgFpw1CS$a
zfU&VLkquQLyTOPdxgtevagYe!<L>T`z`9#XN(<rVp11eo5`7g@)3{4i>LzAp-oTQ=
zKHmHl#78Nik5TS}8|tYyNvsCpoeDUUgH+nv)ANAa&cUHC*rB-`5jE`fYr&FGSzW^P
z>wgQ1hS>_JuZ7hI6n84Cs61R~uYh`~uBnMZb{icXMJ*jU+O=%?@+~4JH?Ci&m0U{N
z`R~66!CI5tT4(wM^?!I+Jgao7lbF)}6`lnbQ?7RR^{MFUwmg6KK4grEx1h&1X1>c(
z5JDAx_74X~Jj|SD$Bx}#Ztvc`3#wcB(k0NrFP|=3O>c0ZdO~?y61-pCrt4cZheP9w
zwX)M^Eb?a3(oE~N!>klg774#^#tD`@6OkY8Dy>DNxss}t;{sc5wS}XeK_mcFJg2E?
z^vlH9SW)uBn$@e%p$@xr>C)xoI@+^DxC4sqb0AFO)MD|weEYi=_l`@<I0TqU*y)>1
z=Hm78j`b%}($iy8Q&rJ0F*Y|>GcoB~Ip+)loH}N0Wi@oa{m`L9<PEH?-}pUz2-+$J
zeWQr*@SWV;9jA2oJ*f03D#Qc2;^dP(<$gRTFQ+$*q9`*&a+*YBBA%jvuq#q8focd$
zxx(<zl1jYK;|CAkokCP{xzl8>e)Vd`?%li3ojv>F#)`GQgKY(&`#*W#y({$oM5$@@
z#DwF=q9Q)95C}72lcLu@UsgVT{0`y%nVB7weXX*ZK+>Iq!EIq<gL{YHzP+3(XQQp%
zw9-Iq9OgFHBwM1NybJZ8k&#h=;Kl2put1Z&dj0yffNe14Ha#y>pHv%HgGGp96v#ef
zpQTMmvS#-=SAt%F5Bz+ZXS+3@n}GvB1yd2HnY<6hYzZRv*|TSLa#!JC%A1=L-n<dU
zlY8@(f671kPYZBuiIiol2T+gV<PdEAgpyJ%Y35-8WG^EkcJ11wfNeH-{ipo%WntE2
zpX4`he4Kq?v?ot`m4AB)_NltMn!p(xc~yD2Rb$ar{n_8Bq01a`<m6@M8-s}1yLotS
z>Ez`zlbV52xGc6hhLQj6LFNmV*PNW4Q$cm)pkyHBAAFNGj6NMrjmhs(NJzr;v<qo;
zsYEJ91S4{Cg{P;hU%7Gx+>aX20GJB*ii_spao@ao^XT#8h_7EyAqVW_;=0k&9+__{
z#>SX<I`n|+C!9q>f#=%TH*b_ZnK*>gVq$iOsn#S(ySk@aw3yS-E#$*7Y~)our?3Ai
z$skR@(wgAO{I?tb`6v7qBeQYa`&+;?^?B}jaE6x%g(d^dIXBdb#<CSO^W1X6!>$cf
zMXT7=Xc`y+S0ZkV)?7h`(1*!llcJ-dj)U+gOz(>$p37$x4{)MsN29GKhWBvCX=OA!
z@CI;FnZu6@j8jQV@#@v9HwsrQUw%$o+bpGYC%TBl8u|FV(=<61BZuF{bAo|FX>?^e
zL*KxFZ~OKiG~JD@CiBbI7Yh`<`sXsKchq%VQ-9fKrCGEj_{z>|G^CkTwiS4;2p(AJ
z(t@^ug1m)DknojzJCs7=fS^tXNt{LP(fpRUz-?q0d?{j^fhl-iUKd)OsTwwD$l2jf
z$LBBKzNzWywKAxBmlgmAZQOCH4E94WE!$Crk$u)Ob<{H7CoaDBst=4G6@ekVtWX2-
z^i6W|<A)E!e*H4NVqhR5!q!e&o|JqcgoJOhb1v=P$B!JjK{8N(V}=UAySkZK|MTI%
zz`#)yw!NS^9*qo$i(u=7^#d+nzFelx1a=PUiP-S)qo~&Fvm6rP4>Y%lg1cx6Y8nJb
zM5{!SZOzEYKo6Yr#XYN6XrORCl~Duv@#V{xbIQsuXQt?%S5;L7Ho5T)5sbuV9QY*~
zJe@81txV#(LaNdgnl8zjfXrLBCIZ$6PW}|n+ogc!OBH%dq%`Vp&N;t=iJirEsyjbF
z|Ni;a@bij_ishq}Q%XwyCHidDu~thcmNCfPsTM5<Q1nDSe|``?hqzx5GB7ZZh@$}0
zWdhh6AZ54P?0(_1sHmL_v01bbq=KrN+AFM=&Fcpm->Z$9+Mf0F{mpM0uIH1L)g>k+
zCwCj|=znEp^Va9=&<m%hEEgtpjJ4iezKa5Qdrvo;uj|y95e&JI7Oo^fX}8Xx)i4#U
zTDw-=*f=K1&jIWls=9yJ*!Ys`q?P@bQu3L9T1t?eBF$w$I2g3##pi5j{Grvsc3ww^
zGk&w+Q3OF`<yYv)vQl$$a(3_CD>^C4ylGSPn>W^Fr&`)`#!zw-lSd&XYE;2ZtW+$}
zj9XxZ1eKUCCD#M;8j@pdets1yM1(qb;evGc=7QU|Z*TMM|MhEM^HX++7Is28fFN~3
zLE$Wjrwto7K7RT%3X+ERx`2Wp#2@?nn~eW%s?TvwU&Y8+-O?ia>+kyY>jNZhUM0cm
zhNo6Xa-%PcN(MMyypTYZA3rXkWzV|xZEbDsTBi#jQ1D4W6hqp_!on>V6C)!<%u-vu
z=&MvrMi&BqO=3ggQ)CMSq@<^J40w7%LPECrj_*0}r-wDKqI^(XNa*a)(9nrfr|=#@
z@LKPjYU9o|B9zt?o2(}E*=vc}*mNoO*Copb2L~5~Sfpw#NDxjY8<eWa$t}+dwJf}O
z`?fJ!rvWm~X+kv>6~f*8u{ggTpz#}=T|He{fI$lK=Wks`xD^+#)%YsF&+qd@i)EC)
zS~-s{@83xl(&!~39jVTzw3wGK4<m#?P;u`3`39@dqBhH@n3xlAfFc?5Bn_H}B>3T^
zq$DD*_&-@36vD?qQ%b%88Y^UP?SD5$YiHSWjGO=Ci*w`p8Z2VqgJ?FPy(AIT1%YHk
z*RmqbZtxE7-LnS)<?nO1T*XKgm^)0P8i1Ea0k9wdL5LUx!WKxhGR37~Oe7?D`2Ua4
zgK_g77TY+~PFUbV-a&MXCjhUiDk>u4%J(i?33iY!s)!aV-OV8ZSU)BwFV6=|iFAm?
z5bS}!1?oY6{%hPO3Jb5at1W5r1@CvjHkh8#t)v3nlAJ?O5YbwoGqU;h`zVCe$IqTc
zBg1IhmLK!;Nr*Ue;X)h)8t`62_z>Fn93yYX#_UOEKA&d#fRy#SxVWfix&!au&+p@L
z+(Zv?JOAP2OhBCjea#xl8niDmGo`pb#mB~~xVq*5VO60>fA#9s#g_Xg<qck^q@^)k
zU1B2*nv~$tJ9h-Hx1gFlcJwF<H6(H4#*LfU*{5$Muu%l9!8@U-PhF@xAd@(2VKMN!
ztE;2KusQo`fVf4-E2Yqc=g*(7kJ~IUBP~*DP!hn?)7^3x3+QjkQU@%DAg640W(wd)
zPg)=SBE+z=YfDxvFZlxoOjK(>v@L_<OStH2z5F8dw()*P(7G;?`E4sz-Pjm6ALXNx
zlFoX2dmAHu+Snuyjo#f{Fk~sgUN{^?1CtNNnJj;bnwO>(j|174yOxo$eX|aQ<q(XY
zEP;S&?Ec2tDNbhQ$g9IYji4h5fK|N=3zL7Xtk(hZfQN^N02DNRpKt>(M<edz1qx5K
z{QUXzkgo}<r-h19LEeDT;PPb_uXGY;fPBCH`N&F_9tsvtOii)HZ@t}apH*E~7Y(@v
z8e;@?;%$wMjLrh*fM%~ot&o-`26+`zK!){pA}8wUk?cA6{j%^bY3b>=Z3C9nk(?zS
z55)xqpCcrjpp8<WVG{{_x_p#f2F<kl`XvDo=rNk$9KL+|^zZ%mZ<)dFB0w6U0|E8m
zZD=Y4wWR(*9>9AjMpin~iaUijSGf+Ci1To0`@XBJ*d{Kn4)zum)94q`i@!)eo5JFG
zYR5x5j_!Lca!m%Votc>lg`x4%Tr)GXqNe(W2BKfAxb_;=Vw1Ulleuh9J<5sYsh$#P
zW*;IOj?;B3q5f$7`s3(Gt|eL*w5S1EM1N%<k5r(47>~7+v;ttjrLn8k8Q(f|DAVAH
zg}eLQeVtq!^$kvg^ag_CKpK@ehM<Bb;*+PHflr<k2A`JT$*#w8^H-w2;99Myq%`++
z7=j!ucKNd0BJ+?T>Ntdul08lUgVhZU&K1cw3!vlzRe9$Q16o~djGyr5S}i?(VPQXU
zRf>zUl9DfS2u&vw&4q#xljJ(vu{r~`99yF-ii;JNc|-s233*vVNwLAbyLPcwaTfR6
z6YPgrHwmJRzsE2(7SN;f=TGB`a0N8_QRT4>^<5jFZ)?6$r%~2Acxg&X%0nkwLE9O5
z`68<&t<)PAOIQCu*?7@HXg1~BH$@Q4Za=Sv{M)c`8d0WLKbbYKYYKT<?cR!3Rp^G`
zK7#r^o0*x{X39)!z=cKhzm`q^9k#<_`S|#}(irm@?)dtuySQX;-mAsk(`^zApdRLL
z0WJ)#yh5gM_~Pk9hpq{2-nfyPV*7NTksol=rVSf}U)EkPM=u*7M8_aW<DZQi!=65U
z`cSF`s<Y%|n={ULsLsos$3prqFV+DQaPu?vvLM7jEN~Mv08gK?)R~G=x3aOBTsUHC
zx+h(7iA21#88eCFO<4}1gDt5Da-?D<B_7mp2ypwZKVB@s>Gkm?q%np7xS7A+6mu(}
zI3qxdNb?W}(Tr09-E@Pq7J*<vk>xkrRKJavx9Dkq;DZO-ygmy5sJ&TE=j$j3iU7~O
z5`s@P+?kzq<63&ae!XPrcAF4agbdh$(ge)8n88Qs1K=>S5P*x^1RFE6fr~|Q-P4t;
z^)%*#h2Y^P*wMvO?>l$Ovab%)&WQ8#n=j>CW*LKEk&2Gl-V1NfRFv1U5IlOgCC7QZ
zH~-p_8^5)r(ozf1afEy6i6csv?P-D4>KM?t=-7^B7(Ex-c^;(?upG287@UDUtG|Db
z`uS5IHpki`E+)p*8|EM#^0EDP;2Kl0z$?jhsQV&O6>?3FR<Ul~T58I!$KKP|SDEWN
zmEt^Z4p8WPJj#YJ3?#jf&BcNr1qPm6Yo)hZFt7w|K2m=0XLfH_kgw(3iE05KL1!o&
zt7bClZ?%;os1GJ$_WkuCywd?qO-(>odi4N29T0G$d%xVgBzo==YJNx|%gWB2ekyqJ
z`!KQt`o-R_Ym))Pqh7o?1h1t_`~o|Zu+sopRVYz`CpgA~kYI?k^1y-S@6NeCUS3?3
z<t|Snw?^`o?d9ixQqaaSHhp&^@0Mg+gRf6tw6?Y)p?IVt8xWQ@;DBqUPz_4XmoK-&
zB1*rzPkwue1BL~n3i`yt$iPsB#_;3dV6&TfkX~kouq>ei6pvC=nV*eQzu>(2_)NYJ
z<5+2B>AAUXpK6d2(KEcbb)aYE1!ZOH#x2-yg7-l4ni3uA@#AMMU*=}qwJ7Y4!R=s!
zMG@FAyqv}N;r1Ut2wz1fEPj?egqy!JUPRLe-q3ojcMW30e|x%Dl&)E`hE$Yi&pt>>
zZ|H{wqa912ATu)?6>dvP8?4!TP+UePc<lq1)g+NqEN|bub&F^TQ8r5dwgM3b+gM<l
z?Ck6a3_(VRAFG}}zu%OPA-%yY<-yK9d)Td|1qI6(+>yLjn5DQ=zDbv^Zn}PbF?A?>
z<LCr(A^~{oyMY1zY+K{;5C=(H>r@I7RY6{tx2n3D^_m4pIU-fOnTtB>zZ@;23lL1o
zV1N{y&TeF8AcpCzeJAEELgy0Lz~(G~k|RD@P3A~vz_YxRgy?9YTnDM(BG=!!xd%9;
zY>&fHwpf2fauXA?2X+C;_2I)&q+21A6%xyk{p3QwzJLF|)HGuqa6jZgh6I!DcSt~J
z>Iq|PO&PkGf-})dF?<m2?wN}ZVry%FDpZw~g~M}hH1w;1B?2}W0Ug8At`@6G%5Ep8
z%yK@7y?edaddn{&9RrHqqdzGZU)PzU{^`Bnj%Z2@7JCZfA0H-Fu}e!$jeTR6T~?D^
zODigp)smd+*I!(k8&X!A3{9el^NRabjg8V#EA)XmUwge&P=@ju?BI)MA_jCIEqztk
zi(?VX*`nLIL<tn`-m*T$eQx&9`O~LM?iV%m+e7trGRwYSxDee7l!A{zNT6sZp|jXi
zjq46Q01uwkPp*5*gHq+$5g!K!hqbHRWSo^4q%W{kmzTfr@$ostnCuAX&a~21rp2aj
zN8xVIf){`OShtIH!Nd?tf(L~_cm@bjZ<Dn_fw)Ut-2WhRepGu#EU%cD8sQCvnjr_A
zCvjp@QqF+^0e6#szve9VWgft@?`seT+qJ+X2@4C;63xWe5;uLL%<m>`)ykD;A(=#@
z;5NIBR508_e6en+{lwr!=wa4oO;kSMlD|FDJIuhqAo#vnX#k$Xklwlxv4NzblFPAC
zQ4$xkmMMVSsHU-s>~o#8EqN+*1c!HdYzrmS?w98|PN)L_qN@NI&uMAJELp(>Y2+z5
z*AdQ^%1Y`qn^W9;JbL$Z6v%+^OsJnWOuH?j$J~604phEqae>HNE%`G;yK>_gC)+IN
z0T&Ih$wjW;5Jsq*OCzsNuG!i!+4g-|YxfkR#o*11kY`3An#RaIye>gtC?Lc}lLP%q
zBEaJl5_({{IREL-t}cU-X6<apB%x+tJsben;NE#s1#!_+x(Kfi#l*f2{uj@m{~Ph@
z)s|}(Cr;cWrydaCRG4V8=U)Q3D1x7o(oNhlM`6c+RHO|B_G@{`WPCyb=bJNLo}kAF
zj_>R9b(Y?~{V+U1u6U&qCFuVB=SaNzhkr2IphUqYy}Pl(t~GxVB05fVM5N$?0wM4g
zm<r?Gs^~Ap#VW^-e_}Ydh7zU=cATJhQf5=3sylu9G|`nGJ9caRB&W3F;M=+1z$JL>
zDVWBMLeLE;+zEdMVK`KetZf!Yk1pLBH^WGeTpkb6wf_hGKENw_8S=Yl%K6vNmDi!3
z?6A-Boyrc$&BMoz_z+@DOss~%r5#)oUV%`^$P9pC6qe<so-0G@I##UZtOBWJJ2BX{
zk^lP7_0T;N9_g9zF)&y})Alr&krMh4<R~eAfBc{$wf2x(I>p7sb`XAWjmS6-wptlX
zXI|h(T^Ah_WBe%zWddHrsP4@b2s3}M=7f~{(48&c$mV5k-ptHQQU6q*=16>c`uN+@
z9^^%J^+@E4cFuH-J%xdvuuMYD!g+dndN8G+df`H&QH~sq37;%{e=THe8=M9e<e<S=
z@#x_6oYVqW266~AXxEY@6p14MxhT~)f#O-lxS6FLn+{A}=xL4j-FG0I?f36c5Y`Y%
z^b(0zlS7M)vcKT&`B2M3+V>|;o>bA&s=rT;YwzB@|0{H1mZH91i<GQMNmk$M@*o_O
zaF;-VD%WrqHbMo!A)HHK$quiN*7XUSBIN;hVNu>5bKmE90|Em*XWV}{683MNL2Ey9
z2Tt?;0)=H677f_531mOw$Qmh9UV`du|3<@~phMUg-$Kl5@kly`P6ts2rO=VVSAP8P
zVW+ID`zI@erpZYsuEyBJM0G^z_0z6u-)qrBLHz^%n8{fYc8J`I-}oADSBzXFk!}e$
z_sxb;r6ov#sDDWfm6TNe-8pS?a#F>qk`p+dP@P~F30?unfrDaFqcmR?@}d+Jq}v=;
z;6r^Tx>Etnv0Jq1aDYN{+U(VNGuCJvq#ho}-~!`>GTdO#F9HX0b8|P&(&jOEZ`q3@
zp_*R2zf!g03!4ESwr_twAtf!{_#GHdakRqq9-I>6A7DYHq@=2VO-RQI`EQ#u8qlO0
zR9_zhe4#QaiB8!GMMX6bUu2Yn9_v|IuLdNp!dw_p0D_Mw1%f$4B7?HBk8@+cJ?6Dy
z0jE%dl6n`crO^2*5$BZH*gg?=U?o9Lhc^%<KvJ>^6$M&B=v1Lg@n81~83{5BAnb5N
zo%TNThNz6IS3kla1lEplXkm&aVLl+#L@lYfvm!s=10DY)GA0J922(m{0l+d;KvX3S
z^z@vCp`efwHta*=0dJf%CvCP-ksEn|pd`E)){TCN_kCGVQnRz!!OIr9{}e%{r>AzH
zEZ+$p`7LAq)e%IKjEp)4m-8rqkwgs#o_RaHVZ5p!f20j(j3#lYPq8F_a@`)xCnyNN
z*LnDv8vy5K0#d+@-0{r-xd`n{n(PAjrj;vK0<QQ~8%JzH+m!bjeIqHeJUe%G-$dOC
zXbRsYtvWn8Y~~=Ut>-1?>DXAd{;i-oOEvnD=J@^lXfktmu}tv0tlKgpZyf{gkFmT>
zsD8luQ7>j8oNX3g0>Rw2)=?{z2lReBn#M3L6R;UXhnM^3aU#u-0|JsW7|IE89jklT
zdlfZjto!V6XnmSlAQoDM`~asQAf4!k!ANxo%^A(0w-*bTCNun11d&>*hYD*LJb=J9
zQhWO3$^AV==ygS*^NCVcXbgXx#r##$=*UP^Vq)oc=X&mu(b3OsmaFTf%{D_IMytyR
znThO!PP-7Pi5I&F+9|C6#>Y!%_4J;@Vjw*FhlRy3#ft$=Qg8_6I@^IYffNeeZ@`Qe
zW2iFwrbn$OzO{R@FHxQBnw3SPd1p_XoiM~d?%hOF>-m*Giq40{&yJ<%?c(K?pPOl!
z+Z$M7zvBzXCTW2WvI9bRj|qM(fck_FMS9P{{JsJNFM?jcnfF58OKAS6(9uk&1BhC|
zFC3#W-7mfPl@A{#f=AM41xKQE;O|kBnz-@-xAFamhpx?gS@2~1V)@A6=&Au%<2FKz
z7Ipg{Ythhn6ax$n?n7R>B=VG5;ro$sdag6_m%TjZH3H#40oxUyrz<!e^BZ{PpOas-
zesJ<Rii*B#D>wzU*rlP{CraG#u-Nk4=|9DOK53S~*Hu4D0zuhEfzAQdL{5>6NB|Mt
zo?(R@VV0_f%nwymC`+<4A)6UOX)x(KH?bB;|AZgvsgz3}SA_-VLj@vgxOoU&+~~Sw
z1271E59B!?aBI8$?VKxZb}drZY?;~W9{MooSI&yMmD|>h{d1w}e8lDlKPaiHUdhHr
z7UIj`JpE?X%6)Ncsa40;Z(`(M7jb9Zzbh3O6h0|_;=Fv{>f69tw~Q03FYXAU)eaB6
z<*aRU_8!jI@NZZ4uuN#I{M#wlZ{NOw`BL<?3M_d=Uj3ocfwPF8V0};k*_9t-HUjX>
z`KFf(WfrLp)0>@6l9Wtd5kF5R?Ffqlc_OmH-X2TccA}6(8zv?pp(g5Vm@Tv<##K>g
zi)0)SyNFEHb-c-e41#@6qLTU<J2WEk+jjETX7HR%nRbM3H#3Vzn{%H@%^uQ{0Nu@N
zhbctzYqieO>>Foly*K*Urntom<W0Xh0PTf2l`t>Al1oRxCiOtnQ@Kg~LV3Fc`hb`i
z+ASiI+~5eF1ExU~C}jFf?!IG(jeB0Axw>pC!I2=U{ZPndZ_KGOM-U%rXqiO=(5R`<
z0jveM+ucxy`W~2!3JMN(W^b^oyo(kck-qoWL=p{HIeb5iH`GO0M@NUqIjTXDnSu;R
zw;yPZZ-HwBNME&1?3(>kEHL_>QNR{mGCG$<hDvI{*f?J({XABeT4zcRDWLPr+YyC!
z!`R3^ld%L4{JeIgZXuhXh)pO%KhT@?6^ScE)och|m&_O=6H^WLHxjVys*}?fQ!z{s
zAH#b%LHC;ijYTjzi0Hoz*f*jjKq@(eG?~HtCFnvt1hR!J4XALmCo*)4$d!#7Hjqh{
zCd`g}fkqCpU$?5dsHw63huq#QLWh6<{vB!CMAVMxJazQ-#X_pE3(aasgF%!!bv%%N
z1o%LN>zGRE8)YYYBr<#l(a8l)PEKM0cw_=mHbI=OimxV39Vlzg9yoA-s8h&j4wg)&
zXPm%OItTMiW9x{fI<wtpKaDDGRv(^+2{1B7w9mXz0{I}MBGut9l!ht#C0mee3_ttv
z+`D`C0b(AgR?Ppz8H8MwJL%6)1Uy1J>+R>avvp+MAk!ttas8auunT1Udc#RIH#a9l
zgWFU?3z;|5LJJlekw&x71i>-rW5u<Ej{D)MP}cj-D1|=(JEb6H9D9A^S>QxySdaS=
zRU*cbSQ?;AplfxBR1rRO9)`MW-@cKku8h{}jJ5@9#97?!?d+<d-j5>z4*z+JsPUuU
zxl`utQb9pMwjVdCcVlES9XbJ9z<FeM_R&T{FQl1jrc(v72M>cK?1d$g(KQz>R|mv|
zQA~t=uB=pX|68mJO8v5_jg3tuMh^GDAIQuXrXefve;fNUh^P{>uF*_ah|dehQxYX4
zB#!&JxVTJ?Wb^*`l(>n$Lo8D4931_AqN}-G16%dVSR}2(!wU|zjN6sumz0(cBzKt@
z89fMB`Mov^G?YSxWLHq--p2muCyr2|%-;*}qo*hOq(IKOiZ;%VIhwEa4*3eLSg`^l
zr6>KMQHu$?U4=P<N49T29<zYFGBb?R`Q0++<9@I#r0tr%p`wQY0lN{Q@Q&HF`EO#s
zK{hh>w2L4jMMcF#udpy`r~oW)$HQpdZ?4-*Asxc6+nml<*0|<i?q-n53jOg(=tmKq
zNaHRxCr7)&-MJm)>A1wtxa4mYHl;F=WT1gGu!-Oqqy(yyQ0p^4dwX{)Krz<3i=$<e
z*w*+J=Ke$NMSU3mIRi6EoEotg5wWnaQ1EQfVn3}e*fcu2KcKl@0Nr1Z%7oU2cq!*V
z(U5IPoEvm~qfi5~+l3|!n4CRZzDtI#am6JI%!`}1nPMso38z|U{O|1P^6-L-Ke1=w
zWa0<!UG*(^KM%+3(<zX8wslwhS+8~9MR()TNXz{{mqo?IPWWYKXFD}T^4&BuH0-tu
zZLVjsT~?d3>^tuUQ-@pM<=eZuA~CaLE^e`36ZDavkK4nXf&KO4qsNo=ODdl1J@3?{
zqoPvTm8^q8x3jxD3h4TP%)niV;AfTe-PCJ^SCWD>r&z)aN}))XJ&Oj`fpsD3d8lX=
zi0l&*?T)s#ZNP0n{(#Rj6GgJWM;!szW(MYF6UlH8STB&%C)vbI%gAUc8PCJO*FY+B
zf4OamIrLeisngilsMzZyEiKJGBi=0&fO!yw$mZogY*SIoIxWdau&k-8`}?T0#qKwe
z>*EciI){&Y-uT@<8qpuHrm=RiFMCP9BznC-9M6OrTgEGIINU+)<<Cu&c!Uhe&dKRS
z8Tj>!wAvXb=+k%+KYIO6@|R=8&=w;W;0b$?n<b;2z(&!UGx}=c_eWECc0suX!x`c{
zjUIM*^Ld^TWW>!5wzg1EC@Bx@-L7~PbUSws;`aVdT#`Y4=^1Wo<#?HCZd`Nz;ZKZY
z1XWzc<TPK;;OaImyAuBX_+{6i7M+_Nm^<O8mT&<*z^~K;O3)2pY>TnIw7tmJpW77@
z=|a=*T%?eJ>D}i>1>AX*KcX}T(slV6?{erccocvxrQ%t$L7F_WoiI70qvQ6ai@phz
z{IWy&`;U6FODg+~K$D8<zLbHJjg4&-*v<s%o*d382q}qt34DXFQ%=397}9+&Qo$2$
zFuX4sgz(ERoW%Ru+1Jz}A2M>fy~i@}8j*pb%5Mq@OG{G=b(@SP^9EcjkC{qc$ix?q
z0-28iVFCgh8ZFY@Cu0YI&jIK+piWcjHK|s|*!T!!Q*1K&7%v}zh}Y=jy-j}MftVji
z`8M$TcLJF@@k@cY)fGM@cVB>m>7&=Zdm{i8yZHF}QbM5nA-`@ck#Y1ylZ)G%RYZ=A
za+6ta&IXIDelfBbd6_gcz$|kA>L+$Fty$Awo;OYqVQ+OT8GGpKPj3-(n&Cg67S}+l
z$7rD{7)~<JK=NR`cGg%_>3TM{*anNdoQCfNw*lj!(Qkq@Bja-yGhR#9N|%{5&6o}S
z_(&gaMQf3>fe63>a)QQe_Gr95h&oOyS>Kz}`B#$8xHBdZm8Ry0Tq%Ns>j#<6B8ZXU
zlN1aikkK|oEo@ySgfB!Ii<~yHWw4p3yusR$sdzNGA$w3xNJt>V=|u5P)VwlUvSjEh
zsfe?ZxESh84r#|T%B4GM@pv$a<rw6Qh1|RwucC^U06O-Y&@i-WFEj$#ib76oa+!mb
zwRJ`4+_ZsNQ>N-1MB(U8??M7WTqF%@vVhBC4Hn9iBv?XZ)B{Yeiuw=rF3HBEErO!Z
z^$$C~lAs*4x>KNFB$`#cl(oIR8fyJQ->r#YOSjB8!Ab=!(KjMW&**3c5KyE;ib$}a
zrmU$^LrD#m<23l=3J^l1&L(YB==j3IjzAd-G2Fh*vDHWuJ7TNr`!m2*;5V36{6_xD
z+6$%~zmYE>@q3V*oD3}2`QyiSadGjC%uF=tP9vOKJ2?1zdU{^a)^^U;M{G6-`7G9G
zA+jRq6)^{JMw4B#b0AQ~o&76$;VF~TK>#z>wzks7YddGhJY}a3>AKIPavLF(y9x(x
zp-`TmgtD5j^mGbIiV*no<vbWtbt;wW+u8c=1`=6sYEGS=o>*Wf!l#lFk!O7_E<V}Z
z-27)~NNBUJt}dAM&(J{;jpoP*_dm-iK}2RoIu8|XPo5JT*ngqC2ke1FC-vDox<9no
zfB++;S7v6thPt{s1g@`&DjFM22dw)sn;nCNfIXElf8WUF^>I6bG!&t`1qIJQI$J0O
z5&^S#7{k7Q<HiwLC`w5PA&=Gj{#q+O2De(^9Mi!XziFtbwvLmvEph2LcDC%OudY_*
z=jU&#=a!I^{9IS3iA_h9VQp!7|KlYgEk0D(l2TG%@EwMl_9>cgwENjIwIprr?PJ2i
z3Jk}~s%KxA^cXW*UCz3LhDTV$hX=*u=MRs(LdKrIGeq`hEiV>FZopw=RDl&Ch=^YL
zzyR;~)X)&9nU7({c@7}|emR;>;9>!*SPxx-){Kc=;yM05nqDKC=a(W)VXOa6OtADN
zs4derW6acIaddQK@k&k}zQ1Z`yc%I|aIeUrFrH)T&i6oFg}ELh-35Bn#rf;NwQ6Hk
z$Z}B5LKoq~*&rh={SR88q^*4H*e=LIh^PWdIyTnCU1v9LRwDP7faEz8hOLJ+5$#e9
zd{{23;rn;T`qH+Z9-|RjfA6-wzCJ#jd03*U`csqAz-gv`fKLIrVV~-jmV;K~1IGQ>
zsh4lwRA_9K49iT{e21#+97I_tM4e1AtEU3?VqR4oN$AQCAJ$DTrrh4QSf_O#xn*ID
zA0DUo$<V|^g}lIt6SRoFQmiDhTI!vtSTR)H;21HKeGc;IBwOz(7~iF&k+w~3Q*bA%
zjcIS+z76|b2_^$J$RS~Q5HC|$G3xLlJ-0~v+8p2nO$Z2JJ(nE0L4~N3&_jbB7(A6{
z)hS~N%NJCbL>7lpV2~c9lh@3%W+$Der+i}p)PrgSNCA4g4;l_K31|*<+BiEp?%cWa
zlZG6P<pp^%i~w5#Z@668|8HAs>vQCDTu@=s_|~c^4Y35i#^!E8bb1kSadAaobpuQE
zP3Iu*r&;0HNokJ4v~U9lM@h?i%Ih##8#Hm_8qlLdDQ#ax1^)hK`FII^iiYBG08(Va
zsFPK_eXD^3E<_LGhmL_Rr!PVVq1oo|hPie7HdH3ZiT)gyHf3cw`_5=zxWKJIC_Hi}
zh{=r^2u!HU2cHU2GtD3+QAfWGrc62ts3c(I(b?G-!~)|Ri1dNf!w9QFfT1yG!{mr4
zg<41}$7dn*`F2W|(oMyKQMOg02j)GBKa?h!a}dfhM=DNEMSpoW-ww}LAYTMdYJD?j
z2LIl@G7*AcV9FURk3dhnVZ$Y9CMKpQo}Tiu7gpZhfJXd0YulYY{7t0ryq4A@OlK`+
zfItLNHg((*lirTMq-)9=Gb&kz$Hr=~phxKCB~goZLpO(--QZUMMNiPg(C`e}Ldfsv
zM4Q~qKtqbxjuWAV@Vfi<o#T?D5ota-Qk<To)CP+`^rFM^9J>Ea`fZ$&8`A-7YXkK-
zydonbVL#5e8-#?V*7l)pW+Xn?F3K3?^d!{rJ2BQm776Jn;=C{GPi#&7n=4-BwY6wC
zqRmf$%fwK-sN1w779d>Aoc|h9w@WwQl23j9eAe7N0eX3*33hA;i7pUxk);ub2AIc-
zk<t}DbL7YoS)Grq^EZL5rXq^wq$#5tB^_b29A)1-GYt)m^nd-I`LgVUzd#Sx(b*X>
zTq>(u;coFG*<nD3D56kHljeP9$?^0$Q__yn(Fr*)(Yi#z&)3(Nq{fvScNo6ENyyDh
za;zTP(7lcUe8I;=t~a8U9t|gy=L>2kr3R=~gc~A%$E*WZ4bbW|-O|(An$PQ~NRhQ7
zQD@s%Jb=XO(<MHuZ2|t2CW8=GKuvEEx^y&sUAjq0iK3teR5WOB@<~fykezt9mYc&|
z{*C*b8yp-+>m<}zz>Wj<$dJKesKCL0DfnHuaABYQFAYM$1A110+3|KOqX=q1n#;Dx
z97@dfy&;AIt}rXia*<}h?%iJ)Y_W%bKPYackE&3XXTbd)KYBz2N(e(u?;EjDodGFV
zP?Zrydm0l&XMPI?;Y|sI#oaYkD1*2;m~2Iag$a*=`=OxfvkmNQ0N0MeQ$-8`1&0^R
zL_B=>@IR$8ur1f%GQsB%BJ9MQIFJ8U#nOnn1n09c;IRpsR%kiH@`)zsnmNppot^#L
z!O>P#G6*f0Pp}3&6U&1zl{6dD58@Jru}9zfdVA~le&}Ci30YC+z(DD*Ezg<3Vh|-L
zCNg&Q4Ca+X4ipg?NmSt2!6yB;xXR5Cx=fH9{yJ`duYjCIHEaUu3;NM9;LxA_Woo6N
zJ&ny4Kk&G0Q?Ht;YH-;Q;vE=`v!<qT>${02sJc2el&>ghzTIMvTKN^}Jbpk!sDJG3
zn>T-MPSv|&mP0pvoj5o52ghHXogcklcpT76(%7f}m4$$0MrQs;7y%%`BVSxpG_}F|
z8gyOI$0-uImO$>rON2pk57FshDOFs8q@t|6Ay03SB?<(hYsTgq?ZH+6&B(rJ$B`mD
zMnbsCRkmi;svoHZ6#8TvWrS-!q?1&E07oVRNc7jyF^^OF$yC|W_3Kw05qG|wnGMjq
zmj8@siSQs~E^Yg|E4YmXT-Z5OT!c@9Y|8wa1<E6GTNarj#8BD;MBrDap1mZRH>gCA
zbGgqfvOI-*#2^fjn|cZe4B@`*RNgCbcs5c3y<3R>8@>bwo{^OmdhJU&^Sn|-h14m!
zx=?tp0#N9No=09=F=B-|v%I`K87@M<10krX1_<QO*w~JVMaw0U>r7o}SE0WnONGP`
z5;wv$fbz9B>9GWYAlKWliJL#-#smJ-v4HT}0bY8)_DYXykN~Q|GU}5JeAoZ1sj6Zh
zm+<;{X?igkr$PF}ia|I@E>R65*TcXX<;$UhLhdE#4g~-J4#u3j$KJ-o^rpN&-!sS-
zk8sNMVP=+!Z#WRMX0;GJ9VHyBryPVM(IaAtlpJxE!+@HTlT&(N8hKer->WddyHjBc
zSRT?6CJGG1AOQBX+6aoo67&c0Wmj%y5Cy1_QIC~xvc?`1AL7>=Z{A#rplaQoylq~(
zP;eXY2yp^9`5c;((9vSGtbJVmTn3fFyV21B!XbzglSdc?L9$dvYhwM6iWSTrH*VZm
z5L_Sg+8hn&hej(A10&F<;JPPs*S7%h5c$N5H#;6ds3*1L`ia7#qG9({f_K2n!^n1F
z7MgqvTWW1X28n#H27$CYb&7^tcpg1{YWVdl&sPpk`k4zx5orpN6X~O(`0y=)Lb0%&
z^n7y<NkM|4i9Q*baKj40qR{?@7_*F}oy(V^_~hfIa;SGit$rMBvMg8wnf|`_%^dsf
ziQN<up=5ogU2K|pQw*6l#3O2<Zm`85NBOr!EK9L9WFiLAFcKX^L{#9pMbd>xR>H#c
zFbDNDk)6Xvb!=8s=e1*ka6KFYsVLCsrMh-tol;yMkT7L_2R>J6>+d(anJ%BOgx=v3
zK(56lX%$k7@>u~dqses(2BEHL^X{8M<h;ZP?E8Di?%u!=k;^B@1QU+}w5+7RP6mFw
zx5|c47+#~o7egLky=D$voUZ$QWCw5#x^INCoYzbJtFB%vaz<!5BV83T5{Ql)P}n$w
zDMF+V>A~6e?XUAkz5rv)w{`1QP;A9TMe^9(!b}s``GYVnfrpD$qi|C|D+(F7GQwXc
z21F96vyu3S@D+(6(^dH*G|@yINNz7e{br2NDc8!je*K~SE1rki&ItefB<O%2>Fw^8
z_-O?=Bv<?-;UHx9<EvG@-`!JBfDcJr#FiC)ts4-K543AQTChlk7@SmISQ0wGO`KCG
z4ZIQ)6A4-L<41)>D#arQ6e*;0sBIPKGy516gqIs*Z9r;0V*6i%aVI?>bcuKr5FzBg
z@3-qc2R6RGPuSy7BK|7|$DtVzUEBmw9F>!bNQVwKC=0Hy5E-^^JS;&U#-<ctEh^xx
zt-_MH{1b9h<;PO{eeD^vkq|-9(w^7MmxqNFqRpcQYuB!ArYy=IcXV<xM!yLI?N4yP
z^K}Umt{2jVm)D?brnwfQ0fi<AnP7)igSRUzsYkxakla|th#2kXw}I;`9H1J&#HXRF
zrKqTg!A9PDO9b>Q6~>EPuaoAg-W`I0=;a^e{3P~3;ZJUI#Gn?2tGY_fa2aV&Pu<?v
z=uiIIT*H~6e|s-wL<SWB006)RQ7RKQ5%r2a#PRT!ofs+jUW?{BE|~xUKyC!+1noog
zptxiQWV}>AyI>sXKMY?LFxXaBR{HMWU;FMIxt$UCsG_?(Ow_n)C!*`n0K?4uD%U}g
z4>YGju17q1l60>iaazbgh?ZyIo1ro-<@MVz{LI$B7BG|Cs)sa$+Cjf|aXtk%Q0)U!
zg!Suy!capfgNW5=X|DWiqoJ{J1>!fE4F&Tnc`EdPHj-l{nle2r?`?xt7vSvs(`S$&
z*&j}}fi8rsj;qYu*Vq5y!%EheEd}&DP4X#FP?PbwdHv#XB4-1(AOgITCzYIk#$zdm
zwMg^S_43w+gw)8$E~_Qy=P8Sr%*@G@-o=z4R)RGk>cgEp=)AtdOdA<h1Qf<bHkmtV
z;>E!_8sf?+KSFDB{R)<_+P+n!=sB4hgs{yvi&e|jbw<e7%GUNA*voSBZ7^`a!mr<M
zHPO?}cTK!I!SE<i@ObD6qF%o~gT)pF@2y}1vw&X5P7ntmCxFZaYvzNczpwAwnN5o*
zeoW~$cYs<DdCBc=fKhx1&8TGRlc=6}YG!}l0wN~Ux8Q0-377=jz*wFtMlfOV!3j&d
zgON>ukTLk@E$D!&EpWO;()>ssVRGK&{DB_BA&;H)p&q2`tR;@X!BBOODHc#RMBa`8
zySZJd%7dU(w6;T0%6bw438I8(Yb#%?j#e^1zH5hKP!tr7^s#+tt)SA2KnKtmvU>ts
zk}S}w5&|DKr%igN7^@_><MTsalacOnYg=x&{AVb$2p}cnHpo9v<e#yz85Z(Q(%7?5
z1+VW$kr4Fa-Zhnixz`9pGtc2h<UWbjk-&*~kD$RJL_m@hF)W3v0|*6ecl<bmIQJ4t
z&{>c|QUlq;w#`?&KEu$U%G&Snn?4(c{tM){EK{6FW-V@0@ucIV;3=yr3@avzc?Oqf
zGk0<UFT&>m5z966W5+&P`J!(qR=<a~Mf3#RX?o_vbNT>-WdMl5tej_`@CK@?)hjvr
zIy&ej{X(UIVHd)_ynp{zXaz;@AY>$@WWvC;+ep|}6f3*Y{2(g>7$S40XqZ#y@CP9j
zG0gthcEUgxrz>>C@z?hx5F6>WoqFrg7bK{BeYgJQ%NSZ~eH*_Sr3OiXD1KSJe0@vd
zJKeJMM9BLNZ;#POf?(#cVN;kuTRD+%pngHMM@EYP^<a0X&V8MSaVghwVN6EO4_5V}
z8>@~vd`K2)p^pJ)dNxm_rKYlq+w@5t0PF(qUQ234u$gGPSr+U*F2(-yH&p02DS%gH
z5U0QuzRsbG@kpXv!^j94ph%m5pB;A0x4B6Y^q>O8lflma0v-GkI0H!@!3&}+ViNE<
zq6OQy@yK~xPxBg*=jTKdyJJT$xh4|!hVSuy4Rx?KQHPDCYsJf*48uKXVEzKNygBxh
z_6w2UK%tAq$uBzvi6gi`4nXMuvNajNBg=Jjn_bytD~2-QY4CpQj^jkZ1_lK|5bUGW
zlIkU6^g|z%0d$fZ$Ok2}RlID?%9TQsX?C|VQ{Ww~f&jo=gX9`Qpecw(o&&d#c~}{l
z9C?#4pix&`zd4C*<>3hNW`%A>h+4=Zo{Lv(?dt!Rd<J530z0vZxRr&J>5$H@<k-4y
z-E-s@<Bm@c$aX;Xfml45VXW6mpHVJoM&=?&h0LoEGzF;W)Rf*w4aa+Lk#S0q^X1%&
z(R2}wz32d(SJz;}SA%<3@I%*>rV~TX3)MA_+KPr4f*K$DN$%58@HANE;5u?ZNT{4)
zN?Jyy-&(sH9@PWnh7XA3x@(`r+}g@uE+Bq5hTMI&ty`5^F0?um%$Wb03~rNAU8o%<
zA!ETcQ!m1P*NzHR5+WTW6!MZEbXu&qc+=y>UO*){v=nKHaLjg^fD|dW5#7K4JR*OS
zPY=jzeK$As0}356-&nq#SQ5UH;C_&i&$7cWMgRpv!h-7J3@#4EXtG5dW(JhH*^$OE
zb_mvrTqH1*L)GWEbRqts_x%%CA<Q78;tm-c(@3T+1a&03fvz`#&n?Zwy(SkC#kPp(
z(~xDUG=N?s+~-Ab)lm5?ro}jmaOo8(dI_s7L}Z<~x@o>bCR1+<e8a(m6!B)MMPTNY
zK_+4)5JciC*zF{#Ay2>vd8>HSz5M<Cg}nAdHA(7_d-q;;F1Nf#-^u&v4L!ouO}IMp
zT&VjjxgpNIY6gB!Ks8a=ko%7z5@Joj*f2ZD5#Lve!7)$|ZK0x-3F@BY1-LOuPpIWN
zFNOg+DJa~<0c0-At7sFH1ermYCUO}BF4fqH^F+=>z%uCA(*F@r1lJ;9u>=%=Bqo(9
zawg%8A3b6hJtfNt-8Cv3Sko7V8tSD>9sT_>9eJ<vpV5ok+d*LM;&t<M(9-M{twP;f
zK?KYMHp4ZcGF+{`61TJ<8~TzdXru}@D#o4hv<f0ypJza!^o*Hj{fZ=+IHbboC-^u<
zB~g82Zi{}e#q#A}=5N~|moxyCy12S__V@2a^ZBZy<Lak%`NWsOx6KQL7s>;o1I4`-
zI9Z@G;U^pW?@qbS=N6DYH$k0~nH|EWqTm4LU7C^4N}r;?eG|0>83aM&vob`+*(Xy8
z)<Y)Eh=g2B?1?x#lqiQl?vp6%?cFcd4E6yXF>>f2afqn`$-o{UAVk!}4RP~Y0B0Fp
z)b8a^S{=cRHKNzt)^>3Zc)X^~Xu={#gJ=E(W>RCmc{4PL0Uh*WV=$J9o+02ox$_<j
z1=04Bnhg+sp;LsChyWT~r4W;pH1udmLl`nHhOyP)P)Je><0|b!c{^A9@**(y|6%XX
z<9bfpH-7wW&5UHs+(QOoG?-y1OV+ISpkXkKWe5>MBvV;a*=KH9Ci~d3W^8FJrBJpQ
zOR|Qf2$fK1p~UZbyx#A*@6Y4+_&t9Ae*gJ;%pWtR<@LT^*L7a!d7Q^_oYq5V8x=iS
zKUq?NOlUk=QzM@@VNbWEmF-90@0(`LuGtL0FK131{cU0SOP9Wg*hdbJDndJjC>wpa
zA-La9QiSHOg&|7)hbM!3N7v3D(A>laKN9EiBc6iF@#@Y|p_h7<Py9?>09#*fOL<19
z=Ke`F=W`%q<#AG|lUzIY=NW6Rz$U25dz9Q44f)C!;mzTbBl++=)H-v*G<pD`<}QwY
zFKTj9o~1>LGn)H-+dbjSf>*CX08LfBJbwI@dPr~HoKH#`r1b37t1pk0E8j*}ltV+@
zFl4JV6KmVUWJ`9*DfjNlf%~btqaeG}w}Ckn1yhwf9?bvu?TJpT{P}U46L0q3vgdFY
z0;{aguU;K_$c@V`zVS$a_Cpi-_%y~=@=YXFSMQ&s6{J4UlqrsPJAz!EKY4OX$Jl?F
z%GWWTA(Z@l0*&Y)$7kmMm~+298H-G5+hQFAvNw8>j5nk4LR0f@@#4ko{pZ7Mg6bN)
zZR;N7Sw)yhZBc*vT=$x@{ry{bB-GRoK+$-({{oUMlUtw>72`6WJ-b0vlq=fT0AA+`
zpmTUapMn)oewwFYw<w-<P^nO)nsO$T$rn@<uVm`hJoz^YT4$S&2{YV4-Q0{@YU*|5
zTe9uK`?wz`F>(?av}CcE1k}C&GT8fb0XK1Ra!NBLL{YPsq}TqC*zv>LK6Le`B|C8`
zi~XnKt!Awmf9k7E=T61cM++2B>HJjQQR0KGUcK5Yx&9A-Y~Qm#`g^eh{vQ38`6D_+
zDP0c=7zBDH-UtnCyiGhKCyC<$?{OuHrm&~_xo)|}LznEN*(&4G`*`qr?X%_3B{&zl
z4)eL(@?L9$Mbo>CHfyF2pJ*n`z5EM}!sVc8oAdG1aLjZGi+3Zs=qK8Kp;+E&bHtE+
zMYt2@LC2F6F69VyM^B8;YBe`zbA(nl7PmtL47h!RBf5|Wq{KHMF4pjhnqwEEpOk2r
zUB->4-I)YKwwkW}lz8+^e(6hEP2ggky4$d2TG&B5$NPIup^S({*Z7gW=x7Iui=9vQ
z2A_oASsHia?K|Wdzjwtb+}@#wiDsMP75e1CgDo;QLk_&^?5GIKm--v4_}OZFU&Orw
z-i%+1l@j(X`71WR`!^)Nsb>G+g=Y6LSGv|kPzHB``Aw~{8LwwA2nQVDJAWEJu(tP&
zr3sI|>)2j4+Ng2kITnVKC)}g0R}6OJ?J8R7*1JiN<nD&V<qyZP5rchi4j43E{vhPa
z?tWSylVNt~e}$}wjo4Jip=y2`tJv;ER9BD_%ky~~DX80Rh{y7o1nsQh&fzy){s4m3
zPc_!jA<iiBAQ3%n8u+$n<oaFor-seU+9C5ms&KD!#+ET1S6St@wiy11;kUYb_XXR<
zOf=9pk^J7Vm^{AJPjl!6?vpJB0s{OcjzIsO`;T@TG<?Pmw)Q&XQ<SNdbB}GaP;CU+
z3DQz_jE6v!w`W9&^z>8BHoo4?w0O4N^p?S;JmHAOuLX0v5>U`wDpKT8OU<sYy{l>g
zPoF`VosSGxH+sB0LkVPdR0PSE=QA7s{ptw=v40=u9pusHY{4+~VGrY!HTJo2OzRki
zhM%m)V+NPgJqHdX+V<?xBhk6=Uf1wpdVbr|#&gXj9Oq-vl!Z=X`OOKd?B(Ih@sU7{
zkIK!@pRT2HUho~GDC+ZN|BY=#{9tM)c^i}F-!(s3rxAvOX9YRF=yg#hRCizF`pa)#
zM;})`3Z7v3vOr0N+$xOo^13CqwWdWl#S`RVYEBcjY+1O(&Bg;h!i+mVd&Gw!>R8@>
z^!Eyx`f>{-EnP)OC9E5cd-GOa9(!Fb^O%nR385n75fm^Ektfo44B@Am9j+5ojUPFm
z%n%#3%2}{-oZ`tLjGs@s`p+w?ue0WGe*5<A<M3-Ft#XKO0##8ab-%zMb04E$9MoKr
zhoH`Ki9eh){5tKirtmjhNY1az9tsPexCHZ?eWreXQ0Gt*nb^0NCm(`z?>>=gq3HVN
z_x7M6?aVH@2O7LMQQ8z|S`!kY?Y47h)GYBQ`Qd8h)oNr|J>RY&GdJ5%8qoLUYp*~s
zq}e?dPb$Njwca|Cs<4e}<yiA&LE3ifQ{Qzn65y)Pf7lI;8aKXVS-tX~Dh8ETgFzO;
zNK2c?YdC^T1lK5{m_O<q!$G#xDT<4d!qlYC`n2ByH0lz<O%?!4Q*yzn78Bi1jI}e-
z<uF{Ev3v`TKrt(K$p0M*hvoVzHxF3C(RhNbqPH4+d6V-{+^BvhVX%BuYhT@Xx35+k
z;g9R&%pUj-gxFKV$y*t3$pCxQ1sUeEUB-Z;rY6!C5J?t4z83LVxydtg$rmjz{PSr)
z)2B^1fSFhcM)bFEQdklp@(+x8yjR!!^Lv@|Kxxq#{7GJ|1Muq#)}RIU=SGsgf4-`l
z@l~a0(#R@ylm9Chx?(*-$O+}{I)RGhpEVh*Pn&oJw*(o9J~cKx<OZE3n#ec)dAfFb
zx}qP+{9ha|91dh$!(ARwa-;F^@UUzW+mE$}GaKNcaTQDKn~8~aTp*_;Vr;)#9M$?O
z{<iG8g@!|M(OTML5HR=3&mPPv#2z@#`Ayn9s*AsTjDQ@pWxn=jhZ|%7sqdb&lGBu<
z+}hzp|NXy{w(v{i`Fc?L>G=px)!il2#>SJk67a~l@QdtnoOrx4qFm)O>L45SQHVW{
zFCw;JpJ%RSTjque7HKV%y1|v3iFNRj2+&nM8ENW~5hIN=Hqtmb{Xu7Uh$#tF;uJ3(
zD7>&iG!rcc*1=UV+xuf_^pQk#qFM-Hb8HvuZqP8lRlKx~W}|fLYQ{?|3(g4bQ!G6?
zULp4IDa0d(=Kj||U)aj{!cf|}&hK(hi+j6-+N-RL*U5AL`-}Bvy-Smh%D3_BjvZ5N
zV=$}tV~Wj%Q4zPp9rP*U4j93B%0$W$OydsFeJ}<NO}jT`5%L80j0^fIj;!MU@o|vH
z71(*c>hAn1*)P-ad1kg>HZ3xK<G;v=Wm-yJ7aSoS%8B9^E69~CdowLoo8-+Vk>p6x
zf1;$aM?an=8n?U@==WI;&YPxssqZ^Njnf912?9bjYwtpmpr22<J<jr#nZ~v4!rDUK
zUWZ|pS1`AxM2KLf2h4%Ol&6H&;TdhsA1<zGDv_My5iX@~haVffbm_ZYW%W6&U5|N|
zGg%Ks+pPYq{Q0+iHIELcXyc;zTaHaoy0(b5QdO-1y|ds5!7TcXHa15rzi~rrK0w@o
z*ZD~!vgJ(+((C91*-%7TY#l8xDrkk!AzLRBecAA4*Z%Nkk>eBkNhG90@tyeR`8RxF
z{ItGQwUS=6YnN<W19U-t$zb!A6v5P6Z_2F)t4I`izeCCi>NsP|X<CVFRcXCHv&^vw
zGuC7@q@1U+fhJ!V9jsX&I7ueKHF$1$v9rZ=aj=&^jx6CN;SN^r8C!HSoNmPUbLL%=
z*^Z4FRCfo;75V(a{aoT$ty(qRP4ZbD(`?<d!Ddtb*;I*f2QZP}&Ay$rW=_Cgv~=Yk
zisIZUz8g&ux!oGsmU-a7`;kYcl`<u!!dnhLCwJ%Jv-1bgjBGb>{)ws7tIBt!607o}
zZR^&Wvv8e@pmUVL)w9(|3+oxOQv<@Vcj4tqGNDA02FsnJ&GHuDo{>%Nb&GoBYgmqk
zpm)TFqCrYGh+qeNj6~Ozr=+y&aQ?zDbY+%G{9qVFY}@TfSk6}{=NJVs!^q8KAT04h
z2-)qbLBszH+Z-^2E}Z4I4|=ZwB7?oBPQ8W0p>Yecs||O4!#<m+P08t=@$r;qa;?M1
zPARx!=_090nAji>`!NxN|9Oob<x4dU8aajotC<Nc=!&^N6R}-(xFA0(?;I=RN^$^q
zY$pOR+i#tSI(lfjl>_Ov{qs=Gjkl7H;sjkpq1?tO%j{hMf^^;#Sm`p~Ioid4GjUrc
z@%#?g?4<BkqH8I=$y!*$%k7qta=Q(t3UX+ybeCl^9t*VB-Je?cAHQ1T_5Ip}-n_;p
z7~ba)HbcZ1*9n=)D1gkv<~MjtZ_ty7qqJnZ%#MLILIg-WMcQ@s+sJ!ckuJ9z)Y>1M
zA`_#zTb#LYG6#2mO$wmp<%gC<%*drisYu1Tw1DS)c6p+cQJ^|a`eRc-(Rc-ikVEXn
zVYBmZM&rJ$LIM=*a>6Q`${<U+3-T$5NT~m2a6pAb**VLzk*m$}lO4Pr=2bJjk3|(K
zi{0zjcJzFlo?b7@J10pBFPyLvIh0+J7Y4wE$b+As4HA6~YJ9MF5;MD)5~RnatJL_E
zkqn!bG9<~=8js3rvd-0nBR*SF_u8}V35+1ax|SrS!`!<GeDag`ci7MX`r4?PMCdT4
z???rMg1_AWqX;)w%H1E(285?POuXgmeL7Cc3%Y#%e4=(Pvbk(|;3Lv}?wXah8Xl13
zEnMtbPIf@RnhpRT=}><o2QyjQaPYmYfvaZH<Hjp_L>7A<7V850EbT+udB$u903ks}
z8ynSfrPXXySnP+Cc%-L(T*A^2F2B>OT3v-=)6zu*ou!(c@$R+ZW|JQtSkhCjrod4k
z-Uh%!zH)2KnqfONFQDx4bH7tpa?S_PCVt`cf*7ik^Vh{DE65l7n=3r)4tmzPfPb3G
zAd>F~63Tp(QX)eDTNzLdb!<sy_m0;pwYt|Sdtp6n?Zu9lT545D$hAr9V`5??xgu7d
zBdI=U#s~}zoyj{y1%pOO0M{px_=|B)`n3z3O_&E9kX<mzyLWv1cb^uZupE(DYbI7>
zlZhOFNaY(4%c;q_0dQ~1*3)wVZT!Z#pxkn~c!K)>3s`@|yNA_e(MGSTmM~JbCpi>*
zHr!zOlf&!(QXHHb*}SXZaJko$^Wv;=Z_;|GO9AiFYiskXZusW|)T$4_bZ}Y~Q~A~G
ztqvW-_|>!xK|GT(GVB`C&&bnT0J-eXpU6qH|7GvhB<5MOC3em;&tHCuAKLYWm1^X1
z2qh0ZR{*B{E%bbRIyyK{EyguzS<PCNKWsYE@%`n}VW%L6bCi)Z9+~!a(Du{e9gUD<
zCUi#nd}rH3I3~?8Ak$fE=9Y>FrUhdqgj2#5G_eXh3FroZ;(UMe$XsuT(15r0GBk!a
zICcN74`t~FzKY<D&#aT?><k4&R?#x3X8+kGZ`-!bkHeNNMJy0(t#6*yfETJI6ro<-
zG=R~RBP@yTej!evTdf{-nKY|j+SyM3j$5qa<6M`x#rM&<)j~R0yHw6^l=6vXN=;6M
z$_;PfDu;cmb{NT=bHQ(m7hivN=b#W!NjFvg?aDhNgI7r%AvVZ~i@lO}y8%FY$9E4Z
z`w`H=rc-SO*i>em*{_T-OJtnc;0n{P)+l{{%e<gz><DMaJa$C@J~5gGBnJn<CKlD^
zpP`{~w;Rx<#UlaaVT}}p76UUm_krI&>yUjIPgoHdmey*QxB0`nuydEMMMuTNWb-p_
zN&5iaJ~htjrg1Q>KWFq?VwvXCi-9{&Fnw_sts3aW+mlMr6w6zg=9CD)aF<~9xW&#t
zK_SBXB9MywAiGy6A?~*_fQazL5SnzA>N#uQtc8yAs+Fn29JoY+=($tY!}AtaKi1W%
z`E;v}V{<pza{D@elDnxoKTcMmV*fx?gMn-}dtxlp|9()nhTf^P`E%z!LY)mVE4k+B
z(_PIP1v)xxAEEhlLpivBx?Hr|sJbkCG!hWpgsy&65*qxU|1=x>)v{XcZdI_I@}x@c
zs5)Qr(!0hjcCVZE-Q)_>7ul4)e{xw+;hEfonx<!GAQsWKp^nt46P3$iZHygn)289w
z(~@D8VQw^mo}tI~c)Z3$5!}{&*)0?(y|b;1%Gi`Si>>O=DeCA^4M38Z%3e_k3QT`W
z8*nOoDCJy;=;Xka&B<raey2VWpO{80RU1Yhbt+e0)*;Zif=!9jxHjE_NHs03yp&1D
z%<OovN=Sg|9}ED`6s}LsM(l$*g30L$z!s@EHf(r=LJttYLfnD3E}|A)b0XaHfUd2K
z?cn`Nzh(JL5OL~M$V$(LeQf-X1I-T}H?D8}myx|1e{Hhe&NrC<)or<fmFho&IveJM
z#^0C!nm29!^Z)(D?`zMRW;blW?3@oD{CbCc*6d32x^tGVvo1UO)5ZP1i}}yYtN$4k
zaU}HSgpykmZXTIAC_8#;^pT_^6KXvjaCPymQXhRe@8eyc|7l&l!kk|7?d?Z@>C>hB
zFCjhwfw>n?Y<KBf<(FpdYkNNTJLQox`_=uQJ0;B6cqKq{nfVhm<$uo?|9_q@?0+9-
z!~Z_a|NF&m|BtT5hFT;~X{WdlWkQaVx9lLzx+YY;P?vh5#bmfFlT-%;2FB6^qB=p*
zD>wi=P)>OgEK#;naXcb_Q;P>n9hf*=c}q|9GWJCMuh^$+P1uw4-q3o>g$wduMYi$%
zxn;GJ;JIK3T89YVIX0GCRPN$*c)=0GzcC+MI<raTO70s!tJhc^{RHdzN2O24K~A<t
z^h#mqExIXV3z<l*T2<*zHF}m%M;pSz02idIuRE{IZgYC@klQE7F#%^2sH+@XL<LEb
zbX~jlYDWT=74niI1Cyk}))xPUj_O+_Cl5tcAaIc3nvW1}u?S$T7|zI-IXt`%UcgUZ
zytqYwRqR&aHM%N$%v!RxQyvPlL^IZ5B6n^j5ctIV$7d~xAlp<8&Rq*7ZveLOz3N||
z@pguB@Ez6xip1H{*V<zCkw&(fU9rqCLJe9`jXjh`oMKJp_xzxBm$x0<^n_pU_wLOU
zx><6+rSwMQGXRp!@~TxRQFy1<l@z;B^7qA*3^G&8X)j#Sb}jk-va!O!)3n0fnELU5
z%!Ru@TknZKEDlL{T0s=Yb?QgxW;G7inP$(Tm`M|-LSHgyfNL7OwTEUA!1!1`(1y;U
zSd&0IU%QC?3vpfpA@~Sw-`zESrR^S~L);wVY(6taGieMMW?jzlj1WfhP;Hp+?|j&8
zm;k|(UiQn5j0DV)%@jxt>#^UsY0C!mA(#5TPwG?*5v%RYyj|~sucRVcT=iScK_MUg
zl8ULjCFr`#dpGzmd~#%gCXNuE;ftJ=H%Tjqn?y?xh#9kHE&F9vH5dqMR0qNYc4?b(
zzb;9lNCNK}RkVBO&iK|xxD_mnOKO;H>ip0ZfjHKz`7Gl_t}x~umj<#@DJJ|CMiROC
zqMD!KnPZtLOV{nz_K%xlK&KHZ-~Rm9zj4C{Boaekb|-AZ5&TVvd}Kg{#^GuJ4W(JY
za4u6DP-p$Ad3tp1=}Fe653$T`e*J!f_A5e1rVGpnd=asiJF>Qa?i_hPj2rkg$TI8&
z{revJa+;VC^v$^=@JVOC`n!RsaonT}pZ`!3in@l4Xt*WEyk_6WnXmKnR~>wvb!B|h
zGqFFXxsn21B(31j(kA7W`xfP&<Am66e81c~s`6mR8IS|ckBabrFzU!!f4FXpu%o6U
zzkZhaOK!_IZ{IRdW)K-9-<xqBLe1d$G5iI2y3PK~Qy*Ddx6knTh`P3OTycIa=b1l2
zV!uJQ*zD2e?EuXuK;xg%`8tBD<W`rK{R^{+WXjJD@gY%DRlYMhVz43G%^sS{K!}I)
zZ5(b5CN7z|0GB?R?pT$iWpCZIvv)qp9c5qk*k@Lya^7^0D}J1~wUjC0GhP}E=`3)C
z>Q4T_B&MLCVd{Ibwl=%G-@S%VePLB2IYkiG#wQ<{WPq>jqWKJ(2@FEr5+;UFM~oj@
z-oIz`035E8xy5~QDXnAzZ2+O3k*NVgLFVQapu`Z=-s+yW=8e_Q3+~rF0(H)X3)j>$
zX5v-lbsM29Xlkn_a||HQn76zAhqWgcNBmm-n5`%Oa&Tn%rrl0i-~Vl1^<(L^n@^8r
zUU2ESHrV=J<=s<g71bTt4LX3d-k&!bXd4^sJrc4n`Dq`hME&76$g4v9E<$ODWb}+U
zm_Q9GPf3lR0@Yf4z`&>+>MBMJ(%d@mrar(NwO?h8ga@RSkv<1XOq<=_uMm-5rO5MT
z%;I>clgwa^%tPCK!LUKfrt^|S<+ERMi@=inA-LF)R@EHlWbEpUU`pH*+aY6S;ZbxQ
zA#6+g+dnYQ0q9JFv&Sx6T8fNSNH)p>jdB-+AUfl|p9!u-DLDaQkZMJa7w`!gVBQ|1
zz)#z=&8lfo-OQ8i1hKO;d5F;~{~1WLQyOXLtj|NA=8hq3QVIN^p-4-<yHoy6QPBio
zXCaz84nI8@)V1>P;P3L%&JN~ZHdJz+?)d4<+}<_w(rSIkJX!ph1+twRRM>)@Bz_Ua
zEt-Us(Aj!kiMWZVRs$CK;NKn&Jk}%pMA2@D8CK6<=jGjqElG~mTgEYB&Ae~kF|^_T
z?{-}{1$|)2l@ITh(de^Uf-Q6=uWMBNi`Gvl4jb){>cVVkv75qa6i<L?6|x;>(*!!2
zD){!(P9MhLH1`gf<+Sj<ytoOx7XqG0u-PrGRUJ%dKID8T=}6mivt7(*8o)&~mjP)l
zObmqucPPp0?-G{W=3PlnV3Md`cnyI%&JBf10?|W#xuETYU1`fT{uGk9Ba*iLwZRdD
zc#U>iUu$sPCfi|L2*IPyxTTTES{P$EqDDBXS)-A=LNiG-Y|L||gFXNdyY&OomEcam
zCl#V?I)I_3{Q2Q@LJ(?wpiA&hIJz1;r|~@+KUFUi%l#X`zPqVfMY_EKGIKfA$jRf@
z!bNhsCPfX<&pwh1_e(%60C<UX2<svsuo`I9U6zqREc@ZEOXM#iNL^#V5i?Cf@fzMY
z<nZ@|YouT+=It>r?b!Nwluq-lVv1k=|07g~?}quI<piYfFwUR-s5VwhEGex9ER5LA
zECtO2?7;8poK7*P&B_JU;_2~q_@zC5=3nG<x8lpmRa;k+V%Nt93}VH$o~7(J29!8b
z`MUJugcEk=H99r;A!=`W<FI18ubJyuA2KtAAB7`0%hD!RBjy2s$h7*-ovh4$-b+>M
zP%gTV5)MFBKCe-<H$A#5Z(WHmYg_OxB{R|Z^2DZZ)C!^_qbc%(J5@R@j=Y*89BWN6
z(*!JI!0$5Wg0*l@{eTP`ef8DZldFuItbPqr#uZ&;Va7}_ke)B(Q!BhR|9H;qK42j>
zdu`I;ek^3R@^`w!152_;_~m8(=+Y1Fx}AS+XQ3%zIZTgOIj@m<(W|nb>mS0_P`eMZ
z-15s~o5xpU_L5a?|GuoFekOzsW2l@O9-4TiNtQ)LGhMDhK@*`>^U0#2?kpMln|8D*
zsxP@qufcMu-ONPzdj2cqbSJb?9@68^1%;mlJf#V(&;!ZkhvNWxu*mq`*A;P;@MzBg
zZLY0rl($iCmzGkQHkq%%UUaUyMIj8#!@VR&HbASJ4EsK6@$2|e5yq%-4XtN;T^Rgk
zZ%m9a#&l9-%1-FwHh*+Bep=jHhJd+sQ475EVQl2Ekr=4KYYJVi)$4x)c~53$GwF9}
zc72U>K|7~wzYpEae8+h+I~muNcL##l+FT*y9L(-(=D0A*s*Bo9vikX&_h#FaTwKmi
zsld7O?k+UeCfAQ|GkKR{obFR(G{&T8n9R26;DH^DS);-~@$@Ux$jHlkmE{GQT<fp}
zIWEG-d|~`3wFnsvk}1kN*JKQiTlL7gtp*tdv~9Rb%cwZmn&Q=nF-YU8RlK}KKnk<f
zF=O-9G!FxGJErQ>Z6(}$**@JQv1zue<7CXa1l%j+0D1ViEzt7^y)>HUKdO-`lM!uI
zN`kLkG^U72@dC&=-Jo!7qQ2KR#7!4-F{lg#aE3~zUenA+9*G;>MgEMN0^iTDxLHl{
zUS<4-2!lRmMG=+6TiB;0KxvjEErcB!W2s+BT3l;T9euf5FN2?9{VQJJV5;49kZ83!
z#UnJ@Cy4q(aqNdZ+>>qgchc9#trb5s&h{_T?c?-9qKTZ3Uc6}H0?IZMU$ne=wcmBx
zJiA~WrP;ha#w!T^T(2P55u%*Xa;O^`Pw)N?fmspA@;*HjHO10K#@I%Ov3kJ{-P8=z
zYRWDM;<$Tza5PwGhJ6QxWZMH>^f*D=)cJIXN3Qgg7k{<hvi4sRL0$S$T58CnMskyi
zZ=YoBxnBK@Cn>|tE;mES?xZVUAAFk^o{Jk#$%D#)iDc5@U_)&!t3S?IotiHZgLkd$
zyd7sR6nH7NO>r@UHM5PW^WQ70G*6s#_(R*qc9Dj2W9ucalK!Mb_J4atdb0U6n)T3(
z*JF{tScyH$K~FTkAq<AviTr(&$7|Xjap3W@XN_Gr&&K^AV;Kscsg}1}(U^%kB_Snc
zm@%#tVBZ*~GM%^<K#%4r*&G%e+brAP)vzDv^$7|6OfvBNRLC{xFIvTSWW%dHC}5qk
zfeTU(l;&@I-qMcidZ`I@KcIM$rXJ(<u*d$-u(0kjr+abib#HkaoE%_;)r4w7jOrtj
z8)W=E9eM~Vw&}o*#+>gDZ<c*XgpT;%UtutPN>{8@>85NCb-PJa?6?VVQvIdC;^b7S
z?yNZTSn?58YJjhZ!)ql^lqA}@A(4@h_J88!NxV&@#!+7WzeRnsTZ_8Zo{L5HaGHL|
z-Majaf6<suTvE7G#%6As_2NemRLchSH#TUmX3(dL?pz7jse}OXT+4WLWqHssp3=`f
ztRrJtq>ILR57ATOhWWEDe)F0ZKWlER0YQ(v`|&@Yz#vk%U=Kdt;=3mD7G-D85b<@0
z-$k#rlySDkm`wy)d1N{xNH#V4k<!MP8c@WX+O0}yz;gX8n<3E9>RW>j8Y|829G|d(
zxtFkb18|AV?D#&tqwllZj#idQt7eYwm!w>p$@-p%DK@a}6$$~8G^1q=P;c4Ib#$_V
zst|ZYn=`U&T7W+x)CsXC@8rHSsI3gOjE;$^<@&YuPcN(t*+qM4B^YVBdXA4*4;8ZE
zqEdk-h=h8=_mPj;i2BanF$jmbJ5a1;H|Y5GH;pQ!5~-d}`&^q=ya}<(n4M!|p37O^
z*mJq7WC#=)Te446ClLeK!W*rf<i5;#r!E@Q4nF3f>o7whCCDd)pi4<GJdkt3{0H}7
zTIAUt9gEkAm{zOJFK6Z$Ey}nv8P76#A#ra&rhNbYYE346M-9`=XR4}(lZUKCqzina
z`<$|14^6fwX%keIyFH{#!W#50ZQj*QN9@NpFH<B8S}zef!XPw1TSUkf3|!(*+*`wb
zCIPD&IgY6$kP)Ccpptb`FsoTEzd*q~=A<L&H^T{L!!(jyeYNy@vkPNVA$$IA4{qd1
zk4Uj5k8L}EtJkv3@vU@Fy-!kWJk`KYv2Nj~c1}nLOa3r)-$2b_gs6OY6Vu4&8X$1%
z?#A1kTM_oZn2FMRi|)k9vfpSkQ)Y#7I-};=>1V`O|L^n-Za1)QH->2wU~!82qjTpR
zHCWWy1K${k6Q<>Zyfi*Ya5Q;c3CB{C92h1#vg~nEPBjDlk|#xG#lFvuMV0d4b{AuX
zgB77nHbObN+s*j5O803eUe^LT>=8n?*x~&fcVkfdwq+f5<HBOO(cO)$?;V8;tXOPG
z>WdQjWKIx&0=b~UyTIxylt#gYu?O26=%?3XJ5-@0ZfzBSKQWDj$dC)wZ`!n}d#aeK
zVs&aPbcc7jU1_RY4K&gRcY|JR>jR(Lpf4)+v$!Y>M?Yg$%vu>Krf{HULj484dlC1&
zK3QCZ`Iku|N??qkeZIor(b12HKS4IFG%NDnFN;Mg1Bz>FenxL=y;g@OEI{TTp8KzA
zjw|%*D<b|ygo79;;}rE!t3^9w8&k`;H(gHOxvC7-|7R{k;Zp!wQZgy_3D5zT{mR9n
zXM)tJ%}lK~s&8#$->x0ER&Gr3J0OviWCXqsi>R|6EHok<GNa=6A$qAst@Z7<er`MX
zH74%|#u=O<$&gk|S-r-kBdlp&z#vi0r!%fpHFk>#x|ILV<BUU<_HOHUY_GMSP-X8w
zQ8Yog2jHS2lH6KYcb9%<VT>{Mm>4Vg2|MuYwB1_Rv}{8<`PqN)Q%6#VJ-r`$D_h=P
zY&aUL_8R%C3Qvp=0NzaT0_bgb3pb8I`lsLGc@;Mw7mRTByCEtdEFJdE>bLIlhr(B3
zh!qW53l?l8jJ1>3Y=oP6G`STGJCo_NO>j}5sviPbfRQ|b4RJSW3gWWe1}6rg`Zc^u
z)gA=6Gn&^7119*?bamF+1q-Y&_4)?xy)ltjwq3LowLv7nqxdk!LOV^o9ZJAyh?}1h
zK6Qr4j9W+jqMu=F^iX<zPaX{I1N|mcSLf~Y7ByZFml(&Ts16h5Pl|?2Dm}X2*JXzX
zQf)QRNc%sC7GUZ%lLF9~Nqo6w(k>hxfTymL#p|9)UrUEf{_c3O^@pA#(x`$Y8XTLw
zHF=7-(Z&_$y^wRLG3F|!sI4ek45MB$LgQ{R3x(u?@goFr<ELP~hSJPtza7i56G&LT
zCA9S=pLzOJnJ|~o&Fj~<@xL~<_4}SSwV?ByoIxal6P=QmFBnF5ov18>G@1=Wehv)P
zyU}m(@sT1${cbW@WamN`@fz*Fv)9NJliR08#Iaq-_Cdu5>&_RHK{9~5CuYY8XM+yt
z5_bANB4W_NY&|u${Sj-2GGAr#A(noT%3TcDQAfz86KI!GSZz?tuq<<Tgnit7O(Pd+
zL1=y;hyTY;0Glc>WU~x;IWFha3rt~plKD%i4IYZOrV*OjSu(Ts^t4iZws9PCi4}qm
z3<{x0kvSpA3|=F>#DHORGL}Y-Jv@e^TyUt0>_xL}=t0xBmaZI9U<9UgZKKbG%UK6m
z5E;RU>3ex3xoexX<n&1}eErGnX28;AH~h&u&Hv#>5U{vGX~05@n?&~z#2$cyTQ(~t
zeuY8wKxNSyi@Sl6FJ8K|lJ1PSjvKnCa?+X{0~J({0S^A{@?BEcHR1X9sxCge_Umk<
zK9YpdP(~#`ZnHyTzmrI+%&u&}(N04g02|Z2YJ-3rnq;2pyP4#Qe|m3hy92tvJ7x|2
z5K>@-?EC7G3!k?7&^~>xRJ#?xe{F0(qaU#N=@*qT2cuy^C5q+Q_=)T|hyo+BGs>r(
zGDcan?(UKBOmkNy2GnR!c|+nJKdNRD?HNIF`U`ucWCKO*A!o&t<NPx|+==pL&X6|1
z)~%<ePn<aM-GwWUF~G9;+rv8G*I5ihfV3f+Fo?1v=X}ZE1A77C|F3T@N-mJPRME{>
z_M!j99E&QO-P1OpFzuWUknsnkD)n&iV4%ydiI{YcYp>{7(XH8o6l$`#8Qx>B+{cu*
zC+nbQfFL%g#g(64o%ZOwQlWoKwGLfHMjnv4F!Hz_0~`2YCWiTBmoH-whHs;n{qFnb
zzo7b%-cHiRWsLPjoP_IFYSeq@_5edN4#OLG?>P|9KY8)w#Zb7eQV)Pw?o;EANh-y!
z+xtCT`O>H33Nm}#g{3o;r4o%c+I5!-%5^hjByr_Fp^t|KXoA!sw8fAb9yU*j8p@B4
zpQ131GYZ7>djD-kUAv~w09M<57&)B{3XU*DkW!7zenImKUh9cm*9Hy3Iv?y6hBXXu
z1`yDNl$6>3+Fd)Dkg&v|V1$}+<b4>NVi2(ak)zs!@K{EAyg!2Q-^}ocUrP)T4H?AQ
zeetiq*2yc^<P+V^K@~3nH>AFaZ<V)XvVT`9eX89es8xp<d&;V3W3Uh(@qb%d>J{py
zVjc;s|HQ<^Db5Egn1X(9(Xyqq;gC+Hn8m0jw4GbGPhli!uJYNlvNcyx!mB6aZuaqj
zKJBrQ=~nqZU8d>`>>I{WrgK;8FYa5LnM*0^rDWwW$FGG|WuDwI`Ck5+=(3i))Zsr=
zrj9Ez#U5i^f{k7KT_REd@)AyJwrDPisih(Dl?wq%Dle%i0d%mNx=7>Y$2Db4acyWe
zuvaW2PezglSr6IkYmbqL=6m9xSLIXEjb&ZVI;`s@GZ_p1>wb(9PixNQF8wXp3Hn<Q
zSnIa2m=B)sm1^#Eeof;LE~Wf54*JjWF_fNZoxt{};ciObw5@hY4=5HCPDGd0yf<Gq
zIcQJ%8x20Jcto@v9ADyVY-1g1(6xKYm1#SWLGeJ|a<DVoLuA5~@o+xJu)V}ZfA9-2
zF|D!VKQG6i@BD`|g7Nb=E@2+U1M#H<QDtd0a}((OW=e{4<~=R|_J0<sX^3yp{xIp`
z>3h<SL|+dVtX95f>3E@(rXa^NjCvWDQKw|Yh!`pj8GImhnK$-LL0+Crc;wmuY+q#y
z30V5nn=rk+`j;~WGYY)}`rb^~oI*p1CO58MZ~jom3+R*O(Bf3pu2w8>)gm?)Z#ty+
zmAr2oHM&kh>kFZmJ8aM1y%Q5>aGGQv2mNLH^#@bvzldxd%2Xns8{t;A*~(T;@eYJ#
zH5Mh(@GUBcXNIvwYTlm$8?T}A8~dB*jc!{O4z)$2R?Kc%G;6P||HR}yoYI{p+>JLS
z{Prye%SZF(B<+K(rQuO@Xy?X!<W*-cmCQ{LO?ww!Wn16O<?P0B$M(1JS<Ohr*sHG|
z$mn*+FpSI)Kn~ysh!eA|$w#)$%@0~-+O)#X$8*R9LWd5o%uc5L)j7lZUR6y0CIq)>
z*G}4R9J@v6(a3oMCn({hpTez74jn4po4Xn3cXM8{@$lohoEO^M8rDs8Bwc#dxt`qc
zEMu=jdRs-)CDKa*Dwn|`ZTlLUjz#lJ-WFVc^Gc`Ury(vkW!%!}nsFn|&*^A3Uf?)q
z(K%Y<8@;b9NTN;c4ihI5BIYIFqIv!5w(j84S39Q@Zq=_RWCov2BQN+lx2!1?TUqjh
zZ|F>Oc_5*!`EoGreJN>lAao}&TXk#bn=S;0_%CO+O7UF*SjVkPSv%>_-nKrDCju+^
zlP-zZ#2dg6L88-R3r}u0&X&VTDs@lBQmA#e{rcnp<U~I)Jc4SASBt95yGLdt7vLWS
zdb-j}9QpO^UfE2eH;zx@V$bVVHt3_hR43@`o6S8O_<YSwCiiTmZ>5eQ&p7~%+lGz5
z)y&NYfRJ2XPBgrJX3`@RW`i~+)sf|$0{=_83({uO*%6PN<WB2~QU`bJL9VVfYqxOg
z((KL^*VWZc+YbhO8n_={1VFaEc+0(hfA2Kab)P(m>q30b)tZ*MwA~BaRXG(+=`P@C
zX!?~w5qNYf(vBTb<WKp>$I%>8{%=1H2BC9|)B2AB9P7)gh?+8FJFZeH2nWrNIAf;3
z5qp<58FPuTqHQSwX@d(ypm)FYO^&Rwuxl7sgPW>8e9s^E+qN!wkjL%DcG<yr)Xn)N
zK=CnDE=%rTVIk66UbJOKtwcL6VI6(pj!pCau3B&01lnxAcy2r?;fkLC9<le|3v-}`
zJEOpe`#0gWly|mC8Z^ai<VwzIr7xz@$3a9EVd2F9VuhuC2~rEB8E>6JnSXGhJ6lZe
z`N@^d;mYcOf3^+&+Tj=_`><Kh4@WcGWDKtacNkjQ)wkrVo8v|@&JDbcs=7zpk|E>j
zJllx=8-8Lfzv^*GbPjmD({Yzp^RC@5c|~TE^ov|Us9iFA8nz0j_b<Os>zBqCk<9vw
zR#0b{0HMhc-_jtUrGkU&_U?A3V38kqFX;sZ424_b>z)&LYLgnEVp{k;z47Dzs&6I`
z=`4wmpboYP$MSCE<ju6u!{Dt`DqLuNg3-JGy~Ws|lB<gwf+s-rX>&~Oh62ZwaDlgk
zw+JqIl0iCp;irdL8M%ZQ1meUPlhNKD=+bzt+do`@<^fhq!^x@#4!jjQo3?eg+coU$
zye9G?xaS(YD;E^LMRS^b8D@ZUspGRL1Zw_>$XK<B>n`rmqRdgQ&DUSHScrHX&mX8_
znfPve3iA|E0%-ahv$0=8?AG`_x>wD<xh;I2^cud*?b23rSGuC<O<N8-<V)3ST&cFp
zcytWnYE9XBy9cnHCM}bZ)_YS4+Z$40lK#Ps%hluC6UKg7n1&gEVlUg5seb`{fAI8)
z?6fMG38+@VPqz(fnC8az5w`#`(qTN1YIO|oKjVJ<*fGskYvc1ovkysds-OM7LDQMJ
z0d4&Mg(r2;Yb`GvFX_vP!lAP%-`WPei&C;h;{uOUizOgj47-B2B=!(8cJkMi?r;*e
z%NQ_Ej|_F;<>B?~lgl0LX0l}({RDAdi5BoSy>C}-do8Q-WqR<F=0|fMHg2}S<suX;
zP0v6q@YTlm(^`B|fUm>kzH4v8kYYmjc>F9wwxj)UpE?=sGWlSwUn&I|78yZ5%1Vn&
zm63k#=TH2f+6BQRbg7qBPNy<`4~5U}BS)s*E@5@d&PSP{X|1T9KyB|$_{i-ZiR?hE
zJ{Sdgm?b4YSTcPghms}T*ltbb*q|!&@fH&~AY#;ltKl;lAulTv#AVQ_JZcU#U#<f0
zC{^THIJOYW!0Y$RZu?K{B#xTxXSC(0LUx@wbEa-F7qYLITmuGFnu#4j6sUIhiN%TV
ze+XDq?lyOru((yL`KDHdIhrW^>8D>pY<(TD#N@pnd~K<n^KLV9ObHGk?rFR1Cr@@N
zG^o4;<QQDDQfn(fCHAozwB&?ha?qodJ**5C=~PVFi<t=%d&!Ird!@AHp8DGCBC3E*
z+1&i`l+J>mO>UQbaECTe!CUw&n;tRfA$97K#6oP0;x}J${kFK*z&$tD&Td|lu&lxq
za%o>vO$r5iremL&qS7#YLQ854FMjlB!{Nfw{5WG)C}O<eQ9FQp^#Pim7>uMe!v4=K
zTTYJ4I6(WW^Rya%)24Z9!6_>2zowDRWO*ki%m){_mZ{%lxcbU&uCj{c8u;m_8_Q>t
zV^i95wL@!8e@SGk*Fn}=6rEfm%|>r#9t~ruG}!FZPn&J8G(R%3RdYX+{}@uO_=n?2
zBu^LE)lLHkLVw^X%E$q?lSxB1Sj7zz_Yjq?y1wAx3Yb+2#!|5r1s#PBs0q}Z2YVO#
zJS<_7a;K1xQ^UHxxocz<4kOVz$j^+=!04Q=1%=Ei+qpA-?ceP_d1v$PDCzX=3`X_H
zza7`K1I2h{l)wDW*{8hsxN!}eXPbJoP3e9NhiDNS$q7e}d=P54*@G}3M6F6@eS7=F
z?lo=e5pI<`X;>&`@oMRl1Xg)dwQ5yB)#~ReH?~7vmctC6MpyyiZ$Z%S52r2sv)onY
z*L@AjH0{TM5b=%GgfBjC?7K~Gae8l+dHL=-P@Vxe6tUnV3$4nQwf;EZit9TDwk*37
z8&{TE?k6rHJd>3o%N~iG@7HsE$Q~@vM+<m~;`@y{Y?3#E3UqmDq|jxSb)UsLmz#`W
zbFKAMH^)dW7_~V;Wqu?hL0PG4k_r#>sCDiH2Rosf+-Is>Ah3afLa<u+EC=8TB<1S+
zk#*kGT3eEQ>z30)ADn1ONnB{=+5L58RD|`<dCxC7to_+J?#)PUP$)*Em-P~zlRs1<
zR7momJC}`sGJE2`_LTG9q6UdzIiSh<F1)HFDE-0v5Ja`Ug)tekI(R!F_B7)D$$nT}
zvJ>v{n!nVtjWKpYkdN>gZg0#3D#xgfy1g#8%*X>bC)6r926&c!>+*vQZHfyot&8R6
zhhjCdplMns_BDY?KG{Sn67SaI9_$Ddc!6B0#uhIoVo$0T<Wh(0QCdHNw3Jss9TeC-
ziwkpD*n-e<pZ5WZF~^RcZBxq2QX91tMafMDX0Ll-w%1JS;{AUPJj@#<uv1?~g7O_K
z-^F8nNxLx@R@OZjV0DBbb>CDv1G=n=a`_9H0XT-y`<+RrNugO=8tb5?q$)kT*WWD!
zvlVgsRi=X@uqm<iIa9%(``2pUM~TEKt_pGxm$oIH^}eZi^i!}^&;4-0-fa6DEVng&
zIj-!D7=nHNJ$Pv6IXNI2OZE}y&)kUxtf^P<UYuWR>zXaH3on^&U%Z&@Kcc?F1I-BJ
z7_kR#!48#A5?%2s%Sdv#;S(uV#hew|(!E;T&z0WA{ti7>k}nz}ato#wQ~8<}H>dOQ
z7%e9`47vXIa;paL-b1luX43wi`f`OyjZ=E#Xc}Y_&V|>tfK&n4$oiuItM~2u1@<9>
zLQIxPpz1avwmI;cO>$u?kV25CyyDZ_1vClKHCrQ6T5(;}csZTVkF}>v(ioUbNW^3P
zT*HCZZd<273<dUO)$*X!OzxYuo*b9sX#K|nPqVoiH3Ic<h3-@Xyne0-K?W2q<Fi_v
zG#{n(UkQ>BrD1I^+*mRTbk2QCWZ=e9u0w}1-qY#X>rH@lva1HjD410qzr&L`>lU8?
zXX3ywZuKAALSlp;1wMHVewFgHpI5cR7fS(Wumo*6t6-uY{Hw1PRF5__*+rpKT&yw6
zQZcHcz_ccD(ebG<DN!*V!1RVNp1U7|=-K`aEOUWNzs$>%rzC0rqG#D0b#d>~c+Jzd
z54}9eX+5iixTE8<04%o1O0b)on>r_)B>YNt%hDr~0LtXDhOHEsqim(f@`>pl&R2kW
ziO%t>moM+suyouO4zBs){&xyaHr+z8M%!ZvZj@HleS3o+ccACYL`IL0-jROtqZC4I
zO!<y$oH|8Z6B;LKg_}p${#v(~><_1@MP*~l91o(F?7pES*)k^?ihE3IQqr`(Z)()6
z**?>=bTe1MCh@6P=KTDWjG^B9*>()`Skhs&Dl3CpUUrUEkmF?<+m<QjB67N1gMhCA
z6MQ-ZIDIJ0z~NgZ;Gsh@*N=hW2$veVXaq<HhouOWxQ<b^)cWOzH=_(V+eFJ-#D91k
zWb2jgOlIp4{zt!zlvoI4g~?cFGQ7agnU6>Yu_#4edmwz^VwFA)nrFb&ja#*fd_8sJ
z+9)zb@&(M)_LCtCeZ`Rsw)}{v?+AkYeC<Qf0(iVzF}?g{&Yto?6`KOaiq$A*PnWQ2
zV`y3>o@Dm0J{0K4jvZqX+i){k3JH&3+AIyt-gbiDxB>qz?d4w9IDSD*f#SB}bpfh4
z0pF3X1<X~l9XT20-%CGpF2g9zcE;_n@o>_gfEHB>$`e``3UM1E+~>f+9v=7a_Nq#b
zfLIU9^i|}SD!3p~l3Ty4=<(Qwd6zA{zO%cHjdK(4$0i4-UhA5j9c<mN?d)oOP#b|%
z{bMc0+kmMvFQLY`B&luP8<UBm9>r$(v(J!Pr2cr^a;3GFcWk$3C7zMV8fMxs7kOdx
z_pXBmO}k}^+P3X<)a8KIR4X{|XrcyLjWBBs-`rM#sS9UE)h%n+UD!MFK~i9wqBPnc
zL%?uG6wROwqfxlkJpkPR_Uid#n?jYJKJ5vz>952DRH@2zWI=@(X3AZ?>&#V2H*eB?
z_v)KjgV$}*VW84a#F&n%g}JKjcPF)6m;15CkHwXveFlUc8Wd!L8)S2b0l^x$$quS_
zgHTW-ul2Rk-q${qJ5Rob=A-({)*M!JCYA5V3cLFV)N8Y5y{?~ET73hiJc4ofK0Nc0
z`@q56>o^574Y6p|=H}dZ#MOCYfBylqTx{`5yOuKbnh1hrcCU1bC*fv9wgl<!HN56r
znA6S&s%*#1x0aE^`WWC&*Ta~g9Qj&A5o)2X+-dbpy<J?Y7p#R#%Ue&VP|9kob)|Q2
ziPRn&%QfB@;Lb>z9q=6jM><oITXK)9>u=+2Y+HyR3?QBN2PX9ufPzcxo7Hdwr6OT4
zMBXqj^B5wkz~FNW7nUc>-F&@Jl6b0Y!sZT*hOH}s;Kh`blJ+C2n9_Q2$;}(mu18p4
zB;(~O-=z#_JFZHrti0ef5}vxX&lKS*{UJ=L7AdIu4>Yj4J^rKq$AU9WCihJ=kQ|%)
zv?h~~yPet5&#`&lrJuJ-yW;6td@6_hEjZ}aMRmYv3{`nUAisF}w0j~bZG9jl61N&7
zV$8PKprtLHs#RO_W97vJT=m!J{Rz;HiAlcAnyciXB7-RQ6x5bk)!WIz>9reUj1M2q
z;x9);TfMl$@P+h&4>+DcAU88V)MOvg3wZ5)mWuQw94l_OTO`v<B1ny>c6V{1XjZLV
zG9h$B>aAO28y8Hi4+P5TCb)esS9`Ko8s-#TQ5VPBngd$=@i`riaZ=dqHC}3hCu}qN
zz*zcSjauUCR%7j$yX;LGH%&`Q_6^J)WV7H(JO?$i#xt`d$ld3cU%0$haB$G_EdQZx
zt=l&9_h~TqEo#7t=4L0ynv17hQ@=oYC#!{^TG*CC2$)=RgvgRx<(M3n!X4aJ_|_DY
zvY&S^Y0aTSKXqSbr=dSwp^J`}2}9Ij$xsf<mB`=iYEtcYX;$;PJpS7t7n<U>KOm~{
zbGdc{9z>`+HGaI~5StBmW`^$D_k-T><jJ)U&0f{_&BVO}vzOnzc&?eI5)~DFQMb5L
zVEUX}HU8V|s^dJRl-s8=+Qm>1{V1&C(ZR`#Y9rDEmFmgJ?_N%4ynEd%?%W0RUI@<X
zf(ot8tLqSqG@l{18uUA&Xaf+qSlZuOmml(l^`TZc<Fy!Kx8W$R-L}uGC?auQIu9k(
z%@`k8XX6V;)_ByVPwmE48W}P85tSf6%yyqW?=P4c;(fN2TfkAO@^mH+>SxbBK6azL
zX{{^o;p##UMx@0@k$qw=B9wwFT_^<tDQFWRzPHW}YEaL-`CIp)>%8RWy&1@vo8N70
z?LR3pnY~_qpPN@?@{*~Qz+w6L-8`*Of8DJOsNS9o%aWr1@e=x?tJT1kQ-63GR0#R)
z8lf={efvc`c&<M7|4OV;vC!nyBw}>UVM`xeyrtuSZ=kF6tK7J3eq!@3kX&K)bQ?Yk
z`23;h7wQt=wX&n`WK+VC${f@cD@YN?rc%^cuf0(!Cob%M`tK=TtvW>1?s=luu-bv|
zB1E9GK6-X9R1Yv-&3_i(aq)<ET|hLdMUP&Tr6e&Y^-h7v7yKzx-!E?Q)kt^+BGQKV
z;s@~*ri!*F>%u}HN!||1nGHwYa4$nlb|a#k!6+hf*p_PA?FS(6Dlgt+tb!@>)6f!D
zOrRj#uSs$jRC~GBk56_X+zCj}{M7tw9TI-n6_hmg;CQHG5{m&w!1&cXBS5H3?5}p>
zvGA1($N4H$h|%{-Lp=@*>>HZ|QRd9JyZ6veP*6O-e|#up6M!UzRO&xpOcvBpt#``@
zBJmztnpg(@!LSLL4G7`@{o<ToH``zYSGhthm?X=N!qmVKfYoS*Z^$<Rz0cFV85@?(
z?6avo(-C`7b^Hksi7H_iza`;l6V7LYP1C!Ch)7rIqiG*j(Puw!#UnCVpfGkB>I4F4
z5^_WX7gv8&>ixjuM@0C=-A}X#AV9BUFT2r;>lwlp5K&CLcKQgaHqg${3CLY=C55<{
zpPl51hz63LCNt@~=+D@+$;r;U;|yLcx(;Qeyy{691eCxCYJ$8vJkQm<+_tgxd&PNT
ztf8F%4l^$v0k;4;v0B9p4$jTK^nT}uyA=yAy^oL(0YOmvae>^e78mJE%qwrg<=QyX
z(?Z;k98TjJ?eb=m!qAE9`0#C5?~T_))Z!;scQ@AoH{seljB@br@U%ooJ}kXTcU$!>
zK6*<MYwDXGMIUa*ygsly``Cx{aG$g~br#KD@!ik=F4g(pKlMHM$Fb&(!Y;>OSQmZq
zj|pR=qdj6gj>f$3d><8f@ZJwCpWo^5bMxL8hmGAacKp$nAI<&po5g=Ms?hLZ=Teou
z8AKiax>lc{qraFoFUu|(em)~+Lt(YP56)&3zwd$8s%z{y|Mie_WJaPMX7U?QXRJu4
z%6mT$3kdA<ca_SSBpuNEw;3rN&dI(DDox&*&-%!DzOr-Pa_+L=5=(DGbXjT?>*>!~
zzAjqA<P+waSK}I^vv#g8x_9MH!5zQ08zn4~PA2lDrRNCS>82=gl&PC<`k7CrL;EFb
zgQn48rIs2%JrnSHeX4ryfOkZ6g$7j@?o&P35z=*S#!W#vFBWr4N_q6{_T6_`gJO$T
zF`bQrMFT_<Rq0jW!)i+k3A%mzb}7$UMOk5f?0^Z(kW{jWExN>DDDL`EVqM3|ZVw%;
zn2oRTJOC>ls99hTxwitbwAt8Omh-Wk&Em5D)&9CKWh2fR_zcBBMtSDcsK7N_c&I(*
z()+I*-}EP6pn+1r^v1kqxGs2pk%SuTZU7U_2F}8|rh4udBEcG(GRQ^5E>6rfvF9LS
zW-&{B8)sVZZVI+yx|(=uErK+1zctvDD9Kn_;|CmtAC-jm33&Bc`aSVI;s0F4esd?;
zR->=K&f4%I<*~d{=w1D*cI|MH-HE@zTu)I5K||Db%@T=KL|`t+O-Y{Sx8s&s!?9F&
zqfYOji%vW=YPmiAG5XU4@a<RH@QOmpTR*eULg(?U`4gCF)USr?&E&x1p|e@=lHsY3
zWrMOTAg^BfY-*`tem4h*u_T%pwdc04=G&{KCCv)#G+47w3eL}l5cfVWtYa1)A}7Y>
zj~z2c|Muebls22ASswd8QwNsw*vbu$;%t@uKg8z(@x`<zlAgf}Ca?LCcbHY!Rt-mX
z-lStW85v=<-{&q8!4#*nl~^>$MJ4_?(@2bSFPIpg1`;EpRpG4X%v_Vf(3EzCc!G$0
zAf>}`3_>z)PK!iD9)9fAe<<yA8J#ERy`8}l$c@uxQC9{?&((+IKgY&-f?JIT4=eLx
z1PZ{Hh3Q@3@~^%#adm9$hjPbXbYhyX5x;_dcXxQb;DOtjvd8vUkK$ru`p&h`V^8pi
zuy+AN2&>d9w(xjPe!EQ+;k@Qu<ZhGvnV%4{otJ$iHDI@9#&irg#u=aX<I`csrca68
zQuy`;XX0y@jC-{iE;u+uSYxDj(pk(QV^T}@i#00WBk5a;k0;=zbj{-u6Hp-NNrn!V
zUUa;W;#7`w<X6=l(vGw3i=3I?epC?y*y78Wlz@bj>9h(S`?h8D{HapmjsR8?*J09;
zJ-*C+w!Y*+WnRRh)Mzq2F1y*_52nJT-RA4nI<WqV-+~JNRj+>9MoL#{*N$UVC-J&i
z7?|V`$~R73_+s5I@J3v#c0w5xy@0O8>b&sb$3gcxK7R|9qhu*~Y>&2gk_vue^+J5_
zE;!Gqr4DX!vrgp4pE;A)E}(-S7pY4KIQ7)KNp`<xA1=<GlO4ikhW%Y%F10dt&HFZ5
z3Tp4tV5n=}xzBsWbXZ+deOJj`1{?uva0SEL_`TGK^yH|qf5rmJO1n_(9G~B2KZCl$
zlFau|TyC>brH7Ma<IDP;!i$U|(JS<9`*v*srb3|{@v9YHCdA{Dhz{k|b-UM>zq@;a
z!K8aqU(T89F{r}K=sWw%q+DFED=9hVL(|2t|MbLHKIHx9{Wk5Bza;~m(t^%KJ;mXG
zCr;e^Q&PsL&?|qhdlwS-`KZDMT#tNyuf>rYK?9v!8g#ukdFY1c$sNvVikyx>sX`Ko
z5xLcm(FUhwgU%x=G^y74xBh#Lb03%Z*t>P`-f`PoCxzYLQRQE`-}lq>iGqWhvqX-U
z&=Sd@xWLjn-VgjR``cmBzH#TWi+$i3h#O1xZ^!}2_JI7)Gx>6KGwGpmX?tBRwsO*!
zqr=uezdPdB8C#y#YIv!AQuOvfY?49#yLUhLbDwR~E>G={GU7h9WzV*;v7cmYJ9yA1
zuG}Nn2hZLfs^mSa7*_dVT4=Kql78P#;Y7cFa$E^*{I=oQh36pUjc%5|E+QaegV&Lk
z#v8Fcv7vWH=;Mm>9{a!AyZHL(>!&FVS2EW`+8pM?Z?}WD*ruk<@z)9MJp4A&baM%j
za7yqIWMRaX-Apqh0{SmszC5j<sN~L(NkcbBEFP9~^ww7!Cl;qY-55i!!6#tSpr-fV
zN6jo-`NlU7M)@56{mo9UUQjESFQUFXv(A>2X^$UI+J<m*bPf7_%)iSYOJ&sC*IAFO
zllmVW`)`}PhLN5l56l@-b?@Mahh0`>#t+Ehj=GkU{tP`9YPZbh0TJ2Gtk;Tza|3xH
zE+JB{k*!@nu{y;oNv6Ps2mkuZ-bK!AV^42l_-j|C=k}rsczCy}cAraQOkSMxO;oI3
zxd8QLw4B!p>HoFAmz@M(y7IQIC6{W`tyi38S7*w@3tmp}`|Z8MG)HSDP~TTB5L2hL
zYXI#j<d9cNbik}Xdarfp93rLh9=MYgasc==l06eKQ*6V+PbmSx>G!Y&WE&`NjIirp
zhh6^@Kumn)tiof)ay|~hY)miZ+TTuP6kh@QGYj5I)2ci`O_FX7-)!9QR&$`g50SzI
z7ykI^&xJ{@iG8<&GMvWyPb7|;l4ft>QND35xb3c9)%F$5r?tYaNLPM{Offem<x|-0
z3;a+RxW-<JL@C^>Jnh+_>r0f8>Sh9RwF^&Da|JXf*J0GmTsufm*m~azDn1un{d(Ak
zXX19@m#7ifw~(q$%G*8r(eK?`OuE~Gym0)r?Jjqjt|n1b!O<to7vDu{Fg$nG@7twF
z>I(qJjV5_d^z+47NXROX<oo<2e}|{eYodGS$Cat9Ij_1+bnnwHD469+PeA5fF-v3?
zC=`Rz)Q%As9WsJzwJR9f^~vJ#q+&{UNw#-(_*e0Wh<~^M#iC=d?`fq9(dXlf&Ls0a
zo9XvHyu0#YL*f6u`E!&}$RusV?EF>KE}&Q9Vk>u)dnp8k>tTMK$=d8@=l!S!Q<E}c
z`dZH?do*~xIPgOsaZv~E8?u8HrA`+wML8l_NZ{4d!$G*5u?dzUPohKkq-2>{${1ur
zr{wZVn|VLx9KDf$`HKu1QS3xlJ&Gy;(gzJZf{p<^bxZOKjpF$E*=1&oX=rmk#g8YK
zE_vmu(+?5>ty+@n>%)vam{EVn6(~MnHHHF3jJ~N;C%k&}w<Iex*QRzJei}fTRs5r4
z;E7xNhwTA7gSF`?TdO5Y#NeZuGbS%v>IJsO`LVwgB(?ZgT=iLLbF9>`P-YFgqR_WZ
zwKE)@CU075Xx#JXDU&NdUi?4@N}^hqYf}1Amj(76F3=5)XwCh8Yh=Jn!Z9=IkBSTn
zeYfFCEk`*k)91DKZe0_Uo`1OdrY4gd-bhZ~11v=7a$oW2NLu*cqeJ+I)PjG2aly2g
zTqYi}Cg6e0b7BH&pUn+2wVe$fmC2NGGRyZVtHV{oKR_l)q&>lrsF<cD3$KEM0ElH!
zmm$>vLd-=-1U@NdID0{Z=4ctn$_h3vNU_jPSVznsWV*yR|6>$m&$ju<bgZDhU37X2
zDHc|ZqI`2tnMVUI3F}$ouQ#d2I`qB@bjC2!Os4cfId1v>9w^r_XAUtePN!uH{;C$@
zc`KKE{mR=fNSa~$`Xy`P23$mF5*kv(&<((yOSa{MdTu!Vpf4mU0oIT>2nYgo_|5b}
z!ki&N_%7dF(>$!u6g-^>r$fOz6H+T8VeW&|@lR&PB3!4ao;PGYc`755f-m3faPnrC
zlQ;XFguKE1z%9zmN|xQ}u0zzRen~?IWtS(yEjm5&kY6S}Xw<-0U!45>X|z*PuZKRB
z_F4*4z~mQ)&SAqN1Qyg1<W3WRHAK`xKISjsS+os9{?WA$+Cj}#wGd!c2SJ%T(ek`A
zjbzIJ#zWIzK3qKbc`mL28^Xvtu3rC28|xZCCcx`fbOL7kp1-HZ?$~kmc9ZzV;7}uY
zWBjvi_MAOC=Iq%wXV1+!ko>O}rQH}<0jR30mlIQt!-Z*5UM*c}v+eNwJ&)(_DVYzV
z%XX1>3u#IE^T>UV!If0<78Ml*6waD4WBjGHjzAIp2@(<b5cCGAJ~S7b*b5K5CEPG`
zmRfa5Ir#IXlnI9?kmY>QZt7m{rZ743)UPki0M%j4uQw}?co$wlr{MfAt!uI?ZgN{K
zq4Vqlln`GGg!1-Qr<c8s7F~!wqCQOd@?Xpzt3C|<0Ko9abC12~oBsImnE5x1TgYks
z<$Y%tcuqa!dBnG8&x_+bPKNBGaZ*Zd!eie@_nAaXLX;=Mtz~2;JXf~+=IbS;O$nR0
zY^mfd^K59DPkDHz`bOGBjJ1mDko@N)f1qy-fIendWav#yaqc>=)E6M00I$)L_B#~%
zd))tnslf2l&Cp@D$+vm-$Vm2XJ*NY;dhM?r&m3t2c8fF-%XJpU$|tGDmf37+@|_!#
zP4@@^<^p>>Qb(_1h?p7}O_A*h0Z1l<QB%^#6>`!6riYT_TIlK9h@eEW-C<!FL$4e*
z&l(M=o>6qxb=<hj<+J9Fi833Sd0$n-m~{pt_ZrHtr46BrVA1I*2}zNB9|m-gv~Uu<
zcOTY|SQQk$!d|OX+_J<T?L_A3_3PH%@d%`Pty19BqJ=TRbMJk|Z3Kj86%|QC<oE=K
zAsV_^n08i@s&LJ0+_A-_<4iSe7I8}+f@JTWj7wql%b_s9zGo~Kx*ULieORjWA=$2@
z{e2ZCH#8@&Q1q~Hl&hRMdf%@7^xO?)Tq(mvEQ>37l1!lZdEc#49x{o<wZG8ctFqmr
zKVE)byLPI_2+;=_1a;|=Qc{{%(07XckBd)WaAz;s(!<w2wQw;Y(bC|dV9D7^0JCnX
z5Q3OWq)o60;Jf1BnJEit(BUXBebkq9`(WQ)x{z<~<(k6$=<3vqdp^0{`p*yaA?Ms7
zvR74XPPV_yz$=jbajjXY5*jQm?>cO_aCkaOgTs^}BBh~T5OMi=Yd*8N%{75bFQoU8
zS3P?5y0P75#Y*_WKzp?0FHUr-e0O%KxU)~ZTkBhKQxFe;{v*xF;#;rj^et&nm&W-@
z2>%BT9O%5u*25V1Xo@-m27(pJ?FthMb5e7vrrG4Rei!MeF+9l<D}49e(c=)`>a{;z
z4yP#h_Fp`E_{SJ&{b8|;hp8`cghtQ31i#ibcVj*{X3P$W-o;+;k69OVf)+DgjYcJn
zq+nd%oR3Ti-I2*?Kp=IR?VK0v<vz1$(=My>WG6F08Jo+d8$I^LC&tJZlWS=j%2%Bp
z+!($yqg};TB<$vqLW-oprfEbVu}U-S+9!-|R=r|2Oy0}3(=LyKY~5^Nz+^A4sqG?>
zaZzM)3R}c$h%E&n5WI9*>J8?Z97Z2Ha%2K=D1P9;fdx2f12>C=LyRKLi)j7W;mow5
zY`wS&X})b`*AZDgYX2umha0ZCp4C1(JG$yAjRK2cp~Qk9q~|q0djQG1tWP4HD$~d9
z7JFZFvDjV*4_<$Mj2@x9j1eTRwCBOb^CTKz4M9&M*2nHFW7~qT$MfExlvQ(&<-Km#
zky)4;7A{s3BsR!`G(OV)wRo|)32$4U2t-%>WI)$@<jej{3@-B#(}7`|CpUIY8hW>3
zTE9<GSD7j1E=5V-zY<rDo{YG(d`Hwq9~z=^8A0B=!nav#h$9Symo=p(S--a_73F3j
zVv|!r<=eew%Y#Gb77^c=u?OK?PX&$Ppig&WI1$4(xX@=;b@W*f#RR^0x6U;#PG`z)
z-f_Rn$DIATH96=#=RJ)`xGzeBrTeAs@h{|Q@azWv8Kx_clz)j2<%Y(wNgjMHwsb0q
zulN9s@0Cy6+k!3^8WS{vqWf`_`^@r(k_2QJeD=$2Teps@=kt-FtTL6CX$LD=bm^4e
zf}`B`8NS2z+?;UfnBS!Rh4sGtvck%siuR;5k_PPk`Jlavh1AShmkLyGkKSMF_v`0O
zrioOgpf?R!u<`j#P1e*+YEwkae*Nd1$wI#A{uu+*|7hVtxSaRX*9(n_7pAn*Z4X!T
z3QRtla0>PgQ|B|MmG|=XZBg>tRQY9o{s8!4?1HZ(7bT^!=u)KOiHwT`6qAD+20A`2
zkO5ah{1@wxCTb(5In*mG71td4By<JKa5OTZ%G39vDTH|N&mkN=I`>0IdxHsQ^0<tZ
zpm_{d{Ig{4Eq5m`3iLL^l2TJ+uZ$YIzhifFO3CXdr_R|T?>dOTykyy2!MCecO@oES
zNM=!|_~X@MDPg`hC{3W6s;;R#!5}t8e{QPQnW!rm9ynP&_q}_c9a<a;t4|Djnh9w2
z>=VC>ePdIFgZ8F8StRM%(Q=VtUaT!IDv2$;cj%x#u7esaAP2R(S-#ua{$*0tG!&|;
z$>EGT)XZMqlh=a;3pKdp`Q$)p;~`#*YW&WL({*>B>2xqjBl}*3h90L{wV2CKKkQ__
zrG{S|^d+w-A2GywkkR~$vX^IX%-TD1$|(772prZs$uFNRbR?|&Q5bBOPL*pig-`j9
zl4Zp*kC?HARX_hc3lJ+lf_5~Zl;AFTG11V3O(hpZ8ll@ZmT@R6{5dstgf!F`0N@dP
z<a$983Q<u7dFrju#YChVb=danH!sw9su#l3Lnn(}ybhw3l4zC(<QAkPGWS-#t`tBS
z*R*z}974#GCg9f~8)thxj_5I*62hLJ#I&~&a%iM6BH7i(<qR+S22?qg@T<jVAO1ee
z)Wnb5&Y)O6b#ijDnZ6gxG4}S7`zfS)?T^oUZN%uHkq{m(R`BXmX$m(;INng#JCvT$
zmYwfaaqY%vLfp$=TLdkxk32)CTpS*18g;wP>M*vt&(Q4=X6Hv~)*peja>%RCnGAKu
zDyOvV9qs_0dP%23-`T$3R$a{TRv3rqbZtVpYUK<f%tFk{#?C(?Bqcn&)%}i$ecBB3
zp{VONeb&=tmId5bHD0whO-yOe2t3^C8l0SpL6giUpWlu+UBQ1nlm+*8w6ZMmCPbLJ
z=%&=LyyD;SmJ0JKV_UGHcNmYY?zP;*Sy+*h%D3;p^-txggUmBRGHU+o&4>NWyuM|m
z*)X}JCl3B35@j3BI;ye^-^;XO>qYrZQi%Y=s&C&<mUS&{JCy;Q`+!wEiO;{Z-6$+T
zgyxSwRy2x?`#V1;6E~B3ym}~)^>Tms{`-fiozjNZWV|0ct_Z9~PshbRJ<gr*CN&5R
zHl%0^7F>16G_!UaVs1Ke{t@c5ZlOa_y-?|G%FjmzKeWaIBwvx8)(l^1>eh_R$};^G
zqm#a5dfeV)%AP>m86P2L@u*3Coby-G?S%dmCOjEO0s1g`X&X&lyK>oUY)9E>wpAIs
z?2HVLL#}JN?x=*{?)a#^*?PBH@b#>pU|Wev1yPk`Cw-E;dwNhwR{4i~N7*Cf>t2(y
z7F0BSc=*Q-vHmng0#H!<*`n@_`S#W=GH{E0lbTIKgrSpCVUZ`DjNKPy-@2v^e1tD1
zRg<b3GsJ*peusPhosb}2ipFNnnDyD8BQw+XC<=L8rgrAgSU6GCVzQHF*gMF?=9`B#
z$$}*Ui$WEo$|t^<83dPpEfwVR2@(gj3X)uXmT-TGPb}#YHzvZ|Hw)H-Sb(P=<~0g9
zhgKoRXvE;51ocHZWuqL}TM#Xl+<6AyEUaV`g;IRxw)O1yprD}6d#kqAEDm`0%RRfj
zP?f@<pkw<WZV|kzk2nYQnR8|gUd=A@89dmwZWEKMrmu+#P12xQ0u-_em6l6$$k3rn
z+~J#uD6i{3gr;q%(gd@{3)$tS=t^mOjTF%*(ktcZ5r+q@C@9%VXES$PY}(3UgTbuq
z6e8^uys>{OSQ0}Ov`3B{X?|iMB`0|mr)l!U{XZYgJ5w%bOI=!!gIuR9*dO)qW8xq=
z(KsGSXNfx(g3Y9wLs@aMC5`-B4ydh3pdvJ(hb0T*y}n|Hz2Me^uB|bqYFh8!Jt5BB
z93p~;M*}j4Ie)b9TSN2MvoHQ{1YQw5#MR?Qhht!~C_Okq10R@*BC}XpV5(&)R7Al{
z?&V;LOJs(tg(x{YXI5%|?lJ%~Nz%;FQH^t`CU?;;CkNC#4VY>xQS<Srn!`>)koKFk
z^=?XQtvWAN$pX(|^{j9*KX}82iFvy>wHP15mTGza^D@;8L!!6}Np#IA<JhnhkO&ZM
zu5$+^WV&+VDOMxNhNhGp8IW68m|-=`Wcvg&5M`X=iG@gJDSE0hqe%IFH(#nIyT!nv
zM+o}!7W!JV?5#gJV)qxN>LrniqdT8?s2XAk=jTns$cd$D_+eA~pSk)u(KXLZm~+yY
zc<eCuwn4x)@-$G;sX6zbI~70m<tGbzw<u3Rkabb)uV7kpfFNfv>s-MN2Fs9_ruoE(
zu%Zt8j2k~*UYqjuaJi<`{y;1Ag{1oIJ!=EDAs0SGZ(AQr*~1(!NKm>V$>o(qd{)2L
za>{l-f>p!>w_GE~v36G;jqSVHz2m3h!(SAfo*IaVKq)?@<x1$x219e*&)=GYt`jrV
zVT)f@I~+x`Dr${>Djmg!j$SYsZebB%Y96LxE^Vv7f$3w+S|`-Q)1C3E43N;Dpq%6r
zHAeIg%ihw+g|3x`JMl-i71sH=Ef@6sRJT=}!vpWuvgS))v<Z3;s7+w@3}T!{qu1}X
zHdGHrQFrCa<WIgSZwekQlY(z&zz4}jl|t|CM@%ia5R2#g=&3|Xz>}{>+v*9D8bc3?
zpV)Kq<Wc|W`5nQU1Enh?-E-%T9gh!X%tyjPE+=4>tNimCHF}0t{8hX_xAP)rAfAax
z{9($JmysPt*`f(CS3%Xccy;O&S7(5=DQYFtAcjb$aq;2`dm;(%Zyaq(JrtE|a_Pa4
zvxZ>$hs#n^moHjQSjo`M(etO#a|FuEKYu*!6t|gr|GGpbQelC3hB-GSXtBrQtJbRZ
zLukc3O(in9{R$+ieNTXg02)fY$fOOQxOAkWLT${_QON8Rd;h8J4~EtSsTi#zU7J1I
z9?i@kFBJ7!^#tbp@IG5E?TF10JCCwzbg)o380N$hnz12FooW~u_@Q%bwGA6jFREN)
zaZL0P1f175x#J?uo(2d=_d!eBp~vTCP8;+8f)=$VYyllpF)794IiUb{NP7tw+idPV
z`5NrlaTSjlhlz|CiL8D?B}LPwa#B`O&&e);YHrTKhr~?Xw|&LzN1sulT8dBd6Xu=|
z+~b#+W7Qyb&3G~ZtPvuJTUpFEOP#qNhjY|1FBd;0_jonClJ`>MRxdB}1F6b|lq(Ow
z8pI(j0(u}7TC16tuOf<e2*IW4i^c9P4K|Pdlg8^N=UHaUfo8O`j~@2(Zo5(U9xoJp
ze@yT>77BP{a9J-<Ct)MtU*$_qblVLQvkO;r9!bSo1i7+*_EA7Z!lQ&cnw!lAVx4nu
zRPmmZ`x)()xbGhMwiQ~7)2_T>IG<Bz)y)4144W4bKdL#J4Tfw7^;`%V3%b>U;X+b^
zhRA|d{3kQ4m_&-tjd3HxHK|uB{c>BGZNk|ZK)ItTpP{X;o1H3`WV5HC0FiY4M1n(N
zQLpc0PtSJSJTWmd=8e0!I98qhL$6*d&yG5J{$`mqO#()7yGF&4a!Ng>Jn-%z7b@7A
zxwiqa#Nnr|VLL-GuYLw+CEL4bPgBOe{k}o9$v&6u{4O_iMwbn+m3&l`Y6B7(vpIVq
zg)+uy+{_$Y$!543H0xb6q{kk7yES!v1JV<6DrM^38V1m8l`3EAmabj9ZeEt2^#+%E
zDrN38>g*8snp{_%)>gD`X54NSLt#~l-~)H@PYih{#GU(`={qgyR_U3ErsOB{WTSR^
zD97Z#!@wK#-q07GxF9Enf4!R#y5Kk&0rcbD1j!D?P!WZ`L)rzm5(||vj<@A_&%9V#
z>7<&Kese0r##S2+r;@GO5&-JrcdD@{5ElwGJ&g|8*^n_BvIe4MOGx+~6{gK^#a2!H
zjEp8j-!~%Q?G72Vp$jcFw5f(+vQE;V6BZptd6X+<O8c7Qbq(HxV2Q|=;zZ(|q2ban
zr;~f#oHMsK^`jXJ56sRU(zxn>TNWAxPL2yVBbF{l3lJ4|)OEu4&HWTO-NBb1c%RSm
z-AywS&{*0@LMVHIe;8W2WFIY8u0;bFzkk086UC?q@7L~1icAHjfa2sdkU*{hDWnfv
zhi~4Ovpr=T%!E@&a{kG(xR-MZ9lLRvQ^hceLo;01Z?8MyW>p6jof+ntR%4ohhYt`m
zTcZXWe)L3Eb|9#B^~=7`U%YS{Tc)xfWF4-W_`Q#K1i?^k+uj}bLub!&Zqp5v_f1t{
zmasMCUQL_%{G^tX*HA^P6t$E!=*q0Vq2c9m>gK|_%zY;mdliHFa#Xbw1VmC-2+As+
z2kU^aFJN-*cggrD{$895bkV6(kmKUkz;oi=!e)}f#*iV7I<z{MOs&U8Q<EmfMh)=w
z^&JOiPY%E*m>(OE+_bEfl_yUk_gR63|DX2GJS@k&@Bf!&Fe74+W|%N)3{u&*R5P|o
zjO<FQEo+E~wAf<E7RjEyq(w-|Qbd-FkTfJTO{k=zODWa!`kYsD-}C(OyPxCt-|sn&
zpTlw7*L^tWI?wBSewWYkUY2`)h-gKsyK3(FQ~I2PJ_RMbiviF{AN+h6Y!0GK92z26
z_rUcqs*z3CC;efU(5ug$k(}Vkujagd3A<y{jqmAw9L-(B%Z@I&+)!fxocMBKKM55@
z46mv}&9;!gcqHTk5wo&ibfXnH!=ZMlgo`n!up@MJ-{g%|*{zlrm}n5iUMCRMiBS#_
zRJhG58CW@N3I@LTw1w#OZq8=*4wauGnYPeUi1^ZAl$3;eme<EAhC0+m^5kl;EBB6c
zRIXP1h8Nyu&lQ0I>DfMU3C^QR?zQDMI$A>70@p)U?{P?xZ0)yLmW$JM$ZLx35nhXQ
z+u~EJdHsL8o3;AxyTVmpNyYr-GesD!`R%+V5=V+3wMcML?~biv%M<{5?+m-66dH7_
z1hZpeD6x}o^IRT43@FtQDnW3cORZj%lPtR4B6&i2bC7ihd3w8k%EJul$B8b&zFx&w
z8<N$!R+2>uy~Q4FbW#sYQ{P)F`mqb0mG5^>;dxj|B1wMb^Lr~Q0<l%Zdx{po@B2a!
z`q!SIh?ut4mWWArDfLaz2m_?1lg4>f@iv%UL4c${>jR(1%fk%oB{R%qUR!f-yP|}e
zE8;;A3V(4vX?w7+2E~1jCa>Chom_5mL3yp)GU%PbeJwz)xJKB^v0kL&@t<Dplddvd
zA<xjC<?>>U5}tS}imZu*U8tYroHvGAEHaJY5iQ#&BJDwj%5(fh*q0}kZXh>+Pi<br
zE+OeL#9NNyvRd)TUdsFFaFlYr=%`<*$nP4vh<!m!9UHecI+lR2GK~+VTJN!wJeic&
z!TZw;63Y0(4<A0HA-N@5gIm7dd-HEBfRqFWYs4n+D;l1r*Vxlwx)@(B4WJH}ecqVl
zOKcVKI%|I_J^!#WSDfA`Qe<oHw0e65l$IRLnXM@r7m@aeYXw5uuWPR?SE{OOPLTCL
zZzspyp6LOONYo@zS{JImUYMflR%`eDpzvATIKetl!FZMo@mo~z<y;4;`1h5HE9k-p
z#Aa`B&&xp5F)Nv~Pjt3hycMKI;&Np}YmIL50Ys(4&>f54+^}K8g7cZK0*D^nRT=&Z
zD&@U?kw!T{P+Dm5ZcEBGNsgF!N(2sM63}l^*N*=gW&A&2iCqe&MZA~UT#_%aB?@*k
z%PplFz9Yr87xz#m@8X%Und*<33ktZWWLL=~oTBb7;;d|bAOHSl<>IrnWX_>YDMGXo
z_jY&fu}65qS(yG2iN4odyl}QMpe_bjW=J)K@H#8m1sW-~j?~Am_FN~6Inh7~)%R}~
z?2G!*I1v@hEaxmNMgH2FU0YA%@@`}=tMcVJPI8eE6xDe)N!_h<phZ?#uon0luF=ep
z<CsILRHZ3sWfb|;Ty*d(>@5U`9B3aR+{pAoh`>~%P3tR{URX5XfZ{=Rw`Z#06m?;H
zKNjel7^y>+*L<(aPTqpAvurwe2a;L7?c@DqMo^xJgOcKzj#oT+51GrOS_t+C+1gU%
zg)eRnId<#~?#R~Q%hz{NfQ0+S{6rOG0%%=~5~PVM$8iZAKfw)&7$uQPLyabP?Zu#w
zP707&Fko~p{AE#zv?2B?0qh5VU;L%4Rip02uq)n1-z3p^*$2pfPyvrgjdJJT51`Z#
zjGBXp7dZzoyPfF&`{U12Fx<SNxfg18?8&-{a_`#3{o0Deq(E<P!%NTMOpM26QA`Ia
zG!xlGR%K$N^CZB7$SOtZJei<+36lSBP>Q)i3Bgq3sQPlH%p#Mz<cOk#`WPVVvt;->
zrQnBV(NeE{;|n_|mq}VK7(IbkDFZ~|Q-s|mTVo5>%r#$)$aC{Q*<v07%soMidD=W`
zql)W*_+n0t_BgFz#S(7_lqAt)8R4kFaYa@#vm9Fc5+p7G=v*7DF+?}&Bctd?BxY2F
z=63p-87WkQHkVMT5Vys#plTY{njtq(uKb3V11F#WSg$qhgYNIB`HEm1MEY&9eQR#W
zTTeXh8(huSKaApA>1*>E8)_%6Iphp!SgQ&;`|usEQmGDVo^c^XZK<e<l_#9Xx&BbX
z3cznoA`_{-xyM&BjRx<g1V^<&OnG5*r|W9N^%#s?B^oYQ=eL!wTsT`aJ)Jg4KfM}3
zS4{A(>!On)$CKWQHyvo(uBaU$Yvu0E7DEYUZ7r_Ae_A&C5*o8rC?HQHAt90&RIBA4
zD$Cqy1F#VVo+Nsl6rPejsAh<<y~H4Jb@XZlzaPooWKy4fSFC5pQ8np46K@#oEya$U
z7V8HmcX?dj!%7`hn=0}B9iz=;WG`+;b|PPXKDkEiB-UyuYt6%A`0({$?3-NXh?Bhz
zvA}S#AE0)WwMJte@29Zk8*FEwm~f@C&_R^wmT}x5<mwJ(<hINa<y=f$D55)w?Vlj@
zr|0>$I~0(V`757aS}t>*JmJsU*Ob&s2EVlpuUiO3v>u+$M<~f&+pPQ$(f!N6FH`hz
zT3b>$7Wkv!;R!|lPb2h9rv1WIgb!Y9GJ;wbChD|jprTmk4>2h3LKvv65vbPf%$DT;
zOUS*AE75$G(mwJBJ*Rx#+8!ARMfP_5_zRm;4-A}yy+AC%d+i?Oq|drG`Cfwq7)|&I
zpD<0Zg$07V7eqYRw&n4r{uZ)}0FIBDk)jg|i+hP;@ZE3Gk=hsFKb;pa?3gUd%C~35
z2UkinmJg-(!+8jl-76lLDh9x&b_T;%lf(eN9WNm#`DAx9kVA3_Z$zi(P1$f%t((*_
z_HEvkq(jls$6{loXsFXMMKwu{2H;$Ee!8N9PeWyS{79{FbUuYVAzo7#yB(StV;kd9
zG8N+qU&rs|3spt~Sf3sDYHU(=(qR-E0ci>A(@`uS)_eqTyHu3AkxtZ$Y?*X}>O-j<
zhS-Omq8QuVHG1+Ax;Xh9%J>b|9<#nyU-G0@KND+Le=x5ZC#$$;s_9X9SyllOBvXvc
zTW|jP@Vg5+O=sFowSXaF8|QXhA{|&kx;b!x1?`WH_zh!41zEKQu3(~p0mVOwD+H;W
z^PmDen^+8pn2jJarT<kra^2PJg|mQ8h)Jfuqt=Cy6iE@38^h9EuK}vt<ufzv+z47F
zf)Of%n37BOZz<;9qBANoj3RuD4d&bI5!se70T`Av@eiOY5bH(8CYtv2eN4wIK0~Tb
z%*_kap+>$Djp-%$D2PBiaGUM=qEIeQ>=M7q{w%5Ayu-TV`BKS%#UX?;b$p=}3O90M
z^Y*%@D3ZS>`s@i8{*pqU@Jl2WVTo|e+|FM31{A8YUd1E<FQgDE6NBu#7=h<GENnd~
z3j{kYqjD_#7LjhE52igqkXI@TN2X?bVCssIv7l8&4n`w{Igzy}FG7*<<~O1p^&^ar
z(LtLUhmwB@4%ES?%6*zntQ8$Q@8UJ>-T?IM^amLwJU`1HLela|@s07@MOW9pcy0qw
zIUwsuCtH7h+Ileify5k;_pZ)2r49(zH0nxaRuRqT7u%b-P+O3RAoLZ1RuyjAx-~=e
zrm?|ryVc+#RgL6~mp94>Hs0SD<s1qU0_!^{re5zWwORX?4*?URk64(!(PXDT#Y<vS
z!5oKJU2WDvM~65_pW&ewIyz|0hi1qT5o1te@r-Y|zoSV0tX=y{0NFn?dwth<4~ikF
zpX^v7Ud@WDQz$~INr`6D40p{T?H3ttcN&6-E2a!i&$c^9IY*dLbn4PxH;gBPnhi%u
z1gN$(S!$0nQcH}yLNW5n!^o?B{YxD&@+uyl=sA<wSo%|9rdV;Q0mM%rZPeU=zL(q}
zE#7$pUDq_am&?2;NlP-Ed4-jj&q|!+>iTY*_J90-k1M*Wdlt1#X1d@M<L*vgt4Qg>
zQjDU8WTP}j$t%w4W;XkC$pEP+%8Y>WQ1Pm$7ZHf4O-pZp5ga^0!gJNacRcq?Ewq57
zwm{)Vj9j0uDCh=1VswDu;}SmS8n1xi2$kgLZno!s0V|~!S^iM^(c<7HmM+xk3eKl8
z4+-6mi%J<)x7s<HxK8BkU2Y@gh42Fv69LHx&jS3IWdL}^f~bT@vYK(C5+fT4#bM72
ze_E{}Yog;XTZE`WGoxH9Nd_V@j}nKi*XlPKH`BrFQL6nNNz(*l0(MOR-eOXLDuY#k
zB%qiJNXvZ352GGJpDA5X^KpUHxYwW3Th0U#pfl!Vp-94|RulTjmlK*AweykDN!*b1
z{ajUr)ApFflR!~&>{$q+gSnMy2A_mycj)`v<0v34)A8B2?!&N^3a$XCY24g@J&m1P
zX$Yf?NSQ)+dzHJ*^d^IwZwhw~b>7~4e^WlRH$2nF?ewju#W%m7>jE&yqFwz<uT$Zu
zatg>73I63Rtm6z4m#F0*UVT$zlU>TuBK=UY*W&1|^6o|%YgYB=Poissbjl#>+?UKy
zYfO<3^OYg^<gRK~yrpgZOG%g_?;%dQ&#YaS^RL1RGzg0qpF6P#tmqR6${N7KQ3_7l
z69)9+pr+HV?3`r9pjIEg-aY2%usqQn!U=8o+j&{ltc4$O<hABbY24YOk;iZxd9!ik
zy?Y-AcBAj1FKcj2C(*iAP%-EWPlLs^AW9_Gnan6}OPjD3$-emZy|;rX9|`1xbB32Y
zs;YV{s@9^<FJkV!kZob-Sbk`%yeO6FR??gXmSFS*Q!g$xV5J_2+Z!#c0^+rRV@N*F
ze|||k6nBBKAyJXO1IfLJgOW-nEnB!qy3I%wk<baQovJUS>9nF~Xc^)fCR{C`5l!4E
z`0+<;(79Qv`whE&+h8;WM`>xxtzbloqIAU|Bhq!`yF~avgA*%vZd1N=*)mZZfxs;_
z0dh&Y56_FDDF>1X5+v*r8jZu&aQ|NCW+hq=+0;O*Hy`pa39NT`Xwt2pHu+_5)oJ2_
z5wq&$Pt<~H@gMA!5j3Ub=Sjd_=QkFXkeno!dU(6tFM_@RcnJMRAQWA?%(<lwE~Ly(
z6`3@HW9tMQ8&M%$deJL!v7Pfr1T=&+0VMOMa9#PazS-*-V82ErWVH*8o{%&gpkd2T
zhkz7oGd$8P23$<<WTIyQ#?=DFz`>5=kYS1WKe1}f%6<f49nQ%iA_);u7SQ?0lk}q0
zvG;aSTo71V?LPNqiyyQypWK!wP-#iRqPv=2pukLN?_|1O90giOtg>t2onZD3^jmbT
zu&qW1k{Y-=&H?NfK!;Ye_-iGacrn1#Ap`^gG!j*HBr(oDF(kbKyN@V9C~ZU{MyXy(
zRCe}>hy$a5UMqbba$tMqC^Bs%Z<Hp~sJ+s1voYNDp?bFx;VtY;?CI;ANI*+t@6Nq@
zyYx|L6dEfJMBXT^RPv(+TDF-_p9)x;d2rQm3g{aVX%rYOjRHWQyLa#IkTF8prhT&|
zTkz{3HWPkDtU3b;X2`&SU!<W5Dj4rtEIIn@Jbdd*O({0sZMvn_k9z7<iJifOfG~0>
z&Zw`9XpLk69oM3+MAppfp~GBiRtc6yQA3_5MPm`wr-05k))v;;de~Jl(1YEZTUgkQ
z70fb3j{e20yq9x6G;7&XpNG_b)=jy?HaT4avWtq0Ip{%NtcX|qT?qeAkL<5p!FwGi
zDUR~TWTfN<%AZ6f-`0FiYAI=m^5=i?#fdcwik`1I+C{tX7V}Q4vh41Rs{2KKXH|!1
z8~fP}{l0U{u&-O@>Ns!HN!A&5>W@Xfbqa5+-nlQJ&Zd!#zyGmO$GWS|FEuRt^@8J+
zcLtq<bnAA=)(J4Y_`0m&tIuDbd7}UJ%)T8f&3o?Yzwg79qAQ-y_m}NA?cZN?Z%+T)
zAN&7*)J&rP&Es(L+r#fvzIQqqzRmtzc(d2<>mT>A&azAJ&a&;JHQ3L2?diUWH6J`o
z@2W$4xxTm(LDOnV)TbxctNNL*>kDkNht`_ltz8m(jk3I3vvlTOk*p7&k^FFAlHups
z%$D=r-TwLoao&jmf;#6w^1CoxW&P>y{{G{Xgi^FZD9%ma;8{PhsbZQ4G?W+_{C`vT
zWxjLRkw%d*1nQGr&n6)0@NP`*khe`V!>;P16t<QB^MCbm#_kid{o-4wREMUL9EvES
z<yeqpEp?J#%&YU#JIKiXA)%1IB~iq9T2;!!RTC7a^cqF*0z|XVqpUh%F$B#A%)D#^
z#$Q}YKkd2c^+93#*`G8Imw4jBGg&V3XPyS)_8&fm4GSGe0_MIQ&=BoYh1)<p&p|&S
z<k#<z8ny(V&3y{*q$i8R711tG)l=AURoou!eZ=M9@?juRwIhsF^*5O>&IMd1PSC-H
zsfbjAiBf;6z8mOUbs@`@<Lqb8vUlkuM1r0mat);sMr0zXc*pfg*=L*FJL{WR;u8Fq
zhr=Sj*?VV(4y>cHn6NA-xe2h3r!ISdF3d$J=lH>e`;p0mC%XM+PUq-Ic@zqsv<vIO
zc3XYpMVuSG9jVHuR?uOZNP#KCorm!9D$ia|QZV;q!O^2fWAOOvO=w%NLhkxrb645E
z>t7}K5#Qjhbc?Zy9E;m4?T!<`=mafG9Uw5X@bQM`9i@dErG~EAF*Y-TcJrL_;%q{R
z_?4f|h&luz3W2W`h|DIl>{X3^@Ogdbnq#9*0Bm5IpTAc0yWN+6h!e8<R>!TOeCwS1
zhgQ)N?aY@^`4->10bUj7Al;NgKOO4l-11b)a-C)#)o*vSP3^kHlhdVH{g^rF4*}Ao
zU-n#<9_8rBi|C;>7^)edp~txF7+BKT+@4uzG{%R>#qDBNep<*@Y5mhq|Hkhym@hF>
z7vWw+Ft9GHXUq8Aei3)Zdh+BOKEFFh)_*%6DR!kc4*Cdc)h!dP^w2Sb$(|HtGKK{J
zZ@%aY=er&JqW(ei%IO`pyv`M3l7&GrQXjtrj7)pPfNqNDIcgl#;UN7c?aVF>$)XBX
zV4O+xs5RMM6*_O_?}61$k@ds+5W1inF(QB!Z$Iz5*ESZ#pPqKA8grn`QYyDyRvjr@
zA7A4w;V7V&ApL?NZ@i)JZPPD&a<-F`d!0jPN_tPKSQGbk=u<nRAooBg5cL})a!eaH
zGIOjvW*+vz%%<ZbDUg)u)o-6Qu=#KP{)Xp`sNw|RnKmxJ-XGugbv3jcK!+qTzpllM
zoVk&P`KYcxcv1baFR}+?m}?VQua6RZd`KIkuX|PNzZ!8UtjDNvLB@`csSN+M^rD*b
z1E`>|$=@)uUUp73vuSzr!+U9NwZBdSDm>(Rb^fG;F45JH(E_S#reEK8?a)!u7W9$^
z94-klQjPy%#_xec*}ERw-n)BuJhBy%JAnDh>znXtbK9oFqu<{Rv`F6$RL64154U@b
zDhR7#(DIihR332O7j&dwYF)j^jqi30U1Vx(F2nZot(=ls0{n(mf!F}(SF4?i;_L!e
z)(mue(4xiGo6|0Tl>6%B?&*4*spCnKpB^)JQhLOcNq)|f%TQxTLwa0NWnN~~^&MyT
zbX^seTfN?7CKiND!F*6bzgcYfVE8io)Gf2li%fl*$zoc&VfJ2GiQ=w7y+Vj%4k_uO
zi;5Q7dzg1#u<>B8R$foT&3SP>ek_8xL-5|M<TBJBbz$Zt0HKemzDfWIoC!ghTV0au
zz8wFZ%swHRXJsxRsfn5Qq=y&=vqh1;j*YfhFAov+M)MF8Uhkka((%M>(I*r)nD}Ky
zkyNK{7I+=gKgpN|9P{`0ADek`@#`(eJUjIsI(pmZlZ{n}CvJ~BV@SFd<LzvCDQH0l
z-trt-D9oo)X-e}?@e8w8gCm;|t|ilSr%SM5<C}|%WE*aD=33%ZhAjVS%qg%g`5&Tg
zQWOu;x2KeJJ)v(a*&_5sY^BaAZZ-Zv+!<36WCP21dgIYWJtx(ju+Lfd+g`J_>5KRG
zjmL{vTES6XXH5GY3=%xIMYC2-4C9yNNdgo2DJX}1*|zM2v~_J~f;9^`MvJ;ZxRGk@
zoi%p@bGPr^+a>S$v&k(N8(6DbnLW$%-tSW9@R+A&HhOJSt)ywx{udT0l~utwy1#p7
zgZP3&T;0Y=CLo#GTY?Uo|C<_Jjt4YF4J<JNS|W?%N7FZUHoJH+1wrwZ?H(rcX6B!m
zZCj=y#hw=J&ONTap8Jn6S<RTRFSCK~@UF~!G4aRK9wj$g&(QUgq>Gb4GWUQb>ih`|
z?I;wgCxBYmG-!JJX6|*`@BX}}Ym{R`BpnNA*AagtO%Px;^Frz^0i`s<!&nIh`3&91
zKmgi+9`c#4kFRvKGzzjJ>0QB3TJr8$f9j!#Sw;@=`q!pk?WC2~>v!vAzz)LR(O%d~
z!Tgw~kIhH-cDE`1^kxUWUM8ek#(H@~+NI@-h3ZY|$RVS0oE$g7aXr8{Aqi9R#G9i=
z>O267qYCzPt>YQLLLDaM+e*V0-+d?b6VE|616ZDQZwGq?<a}BqeV+D>Hu~wzut61H
z`WI0J#4!RjI8Gaq`HI(gOam?5L*7*tfy)E;u~ORWem>B(CuuN0pM73v1XMauTiOof
zBNPO2%z_sVe$M6dg5thD9YPyGRKA#u<LE@8t#Y(MlW*UMk17)-{jAGqw(}&HHoUD>
zIjLs1OD_HS&p4vt-^t1!|IS<f$B#Sg501c$$jNfZPvif`i~r*VVdt<sNa56>x(o^Y
z@Vj5{dr`l8{MxvJ|8ZOwKR4ixUhsQ=$M#PxW;JQs&zelK<1qh59+}LY5AoVsZTp$L
zEc!xa@#CXL?h+FD9(m>(OA1}(>T>_kPqx^5j!fH`lKYiK>f75nB_36u^|p4Zf4TF?
z@RH7{<zF0WvUXQWZa_-zLqMGEY%6m=>8Vsp6+(sYh)&I}rX+o}QNHw!M%r?MEq#R<
zge@1xbS4o8^WqoZ=0^Q#E6cwLeU1y~_Nh7Mv+lR2w9$X<pt`E|r!OeO+a(Zm@}3hm
z`2wt)@R7ar_^!+gi>5U)ADTN|cUb5})mppBx(N;Hsls%&%;SEi1CL~D{UniS*uZ-m
z2>Ps#A29cE;J5P^P*At%YBuA?6|{;i1MVFhbjq=&A_LIKajA`qlq*NJQHj$TPU8Z2
zfa8JHw`6vbQCHEI$zXp^$x1+R$%}rLG$PV@7y;kW=Me@iu>a4M#oF!v$UGWe^5xq1
zm)0{tI4eegyJjyQOZXuM6@;>4X)6k}u?t>&{A9N3+Z2xbSycI?Hb8$&@Lk6&TSzlX
zJRy1;9fxtPG$!>1tq`L>!lyK-wi1gKP_Hv9Skb12A0@TXIcX<0^B&XV7Q5h3Fqm*4
zx5AW7o()F6eNBTV<B49I=$THIj>a?BWpPg%T}KSbdd41kW19`_F5KTaeMApLeC%g_
zB2+u|nM%%xaC6X;G1Y7xuZYYdGRl0IRrfD>mPO9qdty1Ylb&>)=f?3c`{=mL_}Hy~
z|0n(%H+E51>F)tpIMdbD^+w-EWdv2zVyIa=&iv%GVC$LWkZ|>?Ssn3Krq8`Sb<wuz
zP!ldbD%1&$es<b<@Seh7dMuzNb0jjeu+NL-ufGdc69>+w5On@gS%S}JtVqYtnF1>4
z71&xY?}G&Yi*Ie~aQys(<5sPE8XKqjgJ(3@1G;AHZ++@^HycW^YInfh^73-Gz=o<r
zXL7zG1W>*7r0Mj4Rb*vf^LyVHZ$|r6?6^|-ZlJ`d>>my(X!Lak>+#xCx_<l^9HrDY
zM!Y_tYF!t_%rmLK*mLd1*?aR}BqUIU)4Q<QVsoz4p>c!W`Q+ExLaj%9u-q`mEw0zR
zKTJ^ZIe+(~cx(Q&ra@r58BqkR-E@|cRkxn}yKt9|M?#g_)<9d&E<g3x_Q?&|gEbnQ
z<<CV`5RQxBUzwl*H*=z^?XMJjCkS>NkgCuRsUhd_nuu%N;_Sq^LpD&$hcFey1uv>-
zao$4%p;H7f#q9-{z&qp0$DSg#0(~;z7_~RIFWgq?gcmj+0jgH&WcolNe-*yI>+?&U
ziHy9SGdl#sVqWoJ>xf6%k0!$u&ZmSO>!AHjk0GM42;cdy5|1Ti@9Y`F+4LyB?e8}I
zw+gA|g#>%gxm+#w@O0$o8p7L*Xzk!})z*AP>6;!440=F>fY3cfCMjH3zU42sIMD_c
zxI@4({+szv&fJ(=#T{(52WmBLEDdlDxNci-9wi@v7BiXg%rLJm(dNAHn4k2bq>k#4
zZ6me8bzZr}P}93pdJP#mbkC$H2ldl?ysV8?1Iy_nj*iOx^`PnOKpV_nGJ1ew?QO7!
ziO}=x*|S!K7nz0)wRD|E^nejF<&}#%Ie1F1`qMnuR#30Kw?F!E6~^-Fr&o+w(SS;)
zRP8qR@m^Os#$!`G%U8^T%ezWEOkcqj@Y*u`K?xYrnN+AV&Ko&inHRW^(=~%%T^fDo
zZPvw%xqBF-y_Z+0&)muK+0?-2=FK*)4^*nqITuy~^-+~ShYUMD!P_gUV537-`J-{8
z^KP0~M`yS(E5@*q7+BI0^%)VqEY40KLHd^X5{@BVvnm`791IN2Aj^lm&ehe;PCCuo
zYRyr;hZdTFC244cb(Y2E8Fb!Rs~s&@{P|m_QWP2|gtnuxBaI{dlWYygpEpf1(UKO|
z<MfQ+mG6H`UFw1ZoCsYxWRN^s^W*{gLQa|+Oc$fG<+lTtgc6LsiaEfZE`Gp9G1}Gs
z+A3ALh9_8jS<dnJ$XWOMPrbfjYzj#j)RGbLWkm&LVjdChd0Nztgk|pUe-3S_5dKo9
z<S&5e@Oa*1M6t{J;U?#YhZ1DSv<wh(5JE|_>C6aVgz#bI<0j0~kdP@Uw2A_pbz4<v
z4K1iaOs>?X5AfWte7EWi%Kz~EUcCld#ZM-(QR`=AHoa0&dN#P~uFF8iRlIRB-2eN7
zO9v-DvN&wGB|O2BCtK3;cIroE+dMG2o9ot@wap&#coHRaG92b`c;y8hIQKu6xqP{n
zp}*4}HjemL6B>%pI6_BL@}HayA-g5-60*`8I%x*^U(7MwbpE?GZKh;bJI1-)ur<2W
zH_~|B)o}eRG)^D6t()W!bi!n(`^7A?%11|rHV!V-^~?TH9j86OOr;v%S}DFJmfi8p
zojx?`wBem8x)W`_nC3ZNho6)NC`fo5)VHO#AP$g(kU~TvulOY;$k=#dC+lhK@-c1w
z=iBt!Uv?w-O3SK|t309_euHXln8v`N{f2M*%$wD10m<o!*}E!FnND)KW=_^U=fRP6
z*sk}7(38gsP)M({$qx~dkGa!;ANBn8N0Fa?xA;NVP=R>N{B)%3?q+}g@Z!qA;ha|C
zxrjjen5FDJagx;-cq(h-KW0s{A(DvfniOH%>e|6Wra+?w@-QVTBg%0>w}x{J+aZ!P
zcRBCB)s>2vSzeCXhC7Z?SyBXPxoe1<<L^_(jN02PdROCiLncWBn3S$;)$^Mfaw6Ib
z%txjANksHn&Q<}^c4EV!UY4GT6p9p`f)CGmP~5#^$0XaPk0uW|@GG2$kpW_oBz@CX
zVeMMAnlUJ`kX}_f!TgAB1Hn-jJpLN(*r)Z>v8Kp@V=ZiK`7l5Il$aMp<(lekh3Mos
zXW&2QShML=YtHY=`SVZI3H1+PbW1t@aaopYtT-~kR}yb=NmG7)klv+;VHTV_XM-re
z+FxFg{#6SgNurUUE_(5z%Zs?hucxZ(_1-0h&oQRJ^<lV;fN-#lkZws)p9V;F-A|}i
z8_Nkngkm(&hYb0#8^C#K%Iw%NEq#zr^(CQNk_K3g1;No9XN`^X!@m;BWX!XODSgG8
zfaSjHm}ll+B&zwDu~Df`&y+oQfQI<LI-EmFkjxPoyru1A-e?JRNHUqO>qo?Xc!fIu
zB%9IR9y2!LxHmW>dk*dw^+X(oInfe)K$3~P7K&=()Dv)wMVNP^wWHxgW9w;80Fh0-
zS|rUe8ubz-n`qI~lblvT0WHaU__bPgI^wbdHARe*+|9<FreP$_v}?8r<jtHtm8)VW
zEiWb8Hm}MXZ6N8A0D)1Rn=yX}-#Zt2m2gJbGFF~fAIhmmsZcOyr1ps*_`((>R88dM
zXzV^IEXK^MVipJk!Ij!XIw$$bk(}YiDIu!8rzS?o7`6+o7LYk$*K$L#ITvIIl!_IH
z%0e3@nTX7=Kvj?LLfzxCz=VS0tLk-^)vcjvSZpXQ$QaF0T~Xjz69$7Nl!}bx<+Och
z8AsA4br4iJwFb%Dte3@(;9oP0y-OO)+02uE+hh6$=3xlnO^H?ot0>e&N^Xqj#N+2%
zghsdR?{-YGFu<!4LBb-GJ3v&dC@Yf2wX|8;q8-ewh}fl!3k@)oz;t3e$aqp$B+nnp
z7mu(wZ@(Ot7AyimE(icf3by2+vl4yl2w4;7!|+?TZjCKyIA<lt-W{#8N4Cxq;EGcl
z$c-_2^>+ma8?R4l{-<7@!@WI<o%klHO2V}H(tR1A@HUf_wVS>+E$`gB##BI_S}$wZ
zXe0lvjd>ZB8{*<e_M{qz{*l!E<D(`4;dZJ&T57(DkG`Y+)w%kA`D0bPW6iDB&Kjd_
QC(my=^w%MgKRd1dFF$38x&QzG

diff --git a/articles/rxode2-stiff.html b/articles/rxode2-stiff.html
index 468b1a612..3e8e41059 100644
--- a/articles/rxode2-stiff.html
+++ b/articles/rxode2-stiff.html
@@ -286,8 +286,8 @@ <h2 id="stiff-odes-with-jacobian-specification">Stiff ODEs with Jacobian Specifi
 <pre><code><span><span class="co">## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span></code></pre>
 <div class="sourceCode" id="cb22"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/summary.html" class="external-link">summary</a></span><span class="op">(</span><span class="va">Vtpol</span><span class="op">)</span></span></code></pre></div>
-<pre><code><span><span class="co">## <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_cc493e863ffdb1841288028f0470c2e6</span> model (<span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
-<span><span class="co">## DLL: /tmp/RtmpSPRppR/rxode2/rx_cc493e863ffdb1841288028f0470c2e6__.rxd/rx_cc493e863ffdb1841288028f0470c2e6_.so</span></span>
+<pre><code><span><span class="co">## <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_ebf53954066c14e2215d9e98d501de9f</span> model (<span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">## DLL: /tmp/RtmpPXkf2x/rxode2/rx_ebf53954066c14e2215d9e98d501de9f__.rxd/rx_ebf53954066c14e2215d9e98d501de9f_.so</span></span>
 <span><span class="co">## NULL</span></span>
 <span><span class="co">## <span style="font-weight: bold;">-- rxode2 Model Syntax --</span></span></span>
 <span><span class="co">## rxode2({</span></span>
diff --git a/articles/rxode2-syntax.html b/articles/rxode2-syntax.html
index 739b987f5..e654cf14c 100644
--- a/articles/rxode2-syntax.html
+++ b/articles/rxode2-syntax.html
@@ -281,7 +281,7 @@ <h2 id="creating-rxode2-models">Creating rxode2 models<a class="anchor" aria-lab
 <div class="sourceCode" id="cb6"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">mod</span><span class="op">$</span><span class="va">simulationModel</span></span>
 <span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span>
-<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_b794dde852cc7c8fe9aebc3eea58fddc</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_0577787cc435460f1d0396f10467e345</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$state</span>: depot, center</span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$stateExtra</span>: cp</span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$params</span>: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp</span></span>
@@ -291,7 +291,7 @@ <h2 id="creating-rxode2-models">Creating rxode2 models<a class="anchor" aria-lab
 <span><span class="co"># or </span></span>
 <span><span class="va">mod</span><span class="op">$</span><span class="va">simulationIniModel</span></span>
 <span><span class="co">#&gt; using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span></span>
-<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_a9629bdbda112e498cbaf24cafcccfe9</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
+<span><span class="co">#&gt; <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_b3707fe11efa354614e6788ba8c23ed4</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$state</span>: depot, center</span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$stateExtra</span>: cp</span></span>
 <span><span class="co">#&gt; <span style="color: #BBBB00;">x</span><span style="color: #0000BB; font-weight: bold;">$params</span>: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp</span></span>
diff --git a/pkgdown.yml b/pkgdown.yml
index 960399357..68486a319 100644
--- a/pkgdown.yml
+++ b/pkgdown.yml
@@ -27,7 +27,7 @@ articles:
   rxode2-transit-compartments: rxode2-transit-compartments.html
   rxode2-tutorials: rxode2-tutorials.html
   rxode2-wt: rxode2-wt.html
-last_built: 2024-05-28T03:30Z
+last_built: 2024-05-28T03:47Z
 urls:
   reference: https://nlmixr2.github.io/rxode2/reference
   article: https://nlmixr2.github.io/rxode2/articles
diff --git a/reference/dot-rxWithWd.html b/reference/dot-rxWithWd.html
index ee03ed7ec..f6da3282e 100644
--- a/reference/dot-rxWithWd.html
+++ b/reference/dot-rxWithWd.html
@@ -162,7 +162,7 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span><span class="fu">.rxWithWd</span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/r/base/tempfile.html" class="external-link">tempdir</a></span><span class="op">(</span><span class="op">)</span>, <span class="op">{</span></span></span>
 <span class="r-in"><span>  <span class="fu"><a href="https://rdrr.io/r/base/getwd.html" class="external-link">getwd</a></span><span class="op">(</span><span class="op">)</span></span></span>
 <span class="r-in"><span><span class="op">}</span><span class="op">)</span></span></span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> [1] "/tmp/RtmpSPRppR"</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> [1] "/tmp/RtmpPXkf2x"</span>
 <span class="r-in"><span></span></span>
 <span class="r-in"><span><span class="fu"><a href="https://rdrr.io/r/base/getwd.html" class="external-link">getwd</a></span><span class="op">(</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> [1] "/home/runner/work/rxode2/rxode2/docs/reference"</span>
diff --git a/reference/myapp/mod1.d/704a78003b911881e13e3865d8be8e09.md5 b/reference/myapp/mod1.d/704a78003b911881e13e3865d8be8e09.md5
new file mode 100644
index 000000000..8a5809ff2
--- /dev/null
+++ b/reference/myapp/mod1.d/704a78003b911881e13e3865d8be8e09.md5
@@ -0,0 +1 @@
+rxode2
diff --git a/reference/myapp/mod1.d/mod1_.c b/reference/myapp/mod1.d/mod1_.c
index d247a98e6..71afc86f5 100644
--- a/reference/myapp/mod1.d/mod1_.c
+++ b/reference/myapp/mod1.d/mod1_.c
@@ -1,119 +1,119 @@
-#define _getRxSolve_ _rxmod1_6b475db22b3a2d25441a4322fa9c2aef0
-#define _evalUdf _rxmod1_6b475db22b3a2d25441a4322fa9c2aef1
-#define _solveData _rxmod1_6b475db22b3a2d25441a4322fa9c2aef2
-#define _assign_ptr _rxmod1_6b475db22b3a2d25441a4322fa9c2aef3
-#define _rxRmModelLib _rxmod1_6b475db22b3a2d25441a4322fa9c2aef4
-#define _rxGetModelLib _rxmod1_6b475db22b3a2d25441a4322fa9c2aef5
-#define _old_c _rxmod1_6b475db22b3a2d25441a4322fa9c2aef6
-#define _ptrid _rxmod1_6b475db22b3a2d25441a4322fa9c2aef7
-#define _rxIsCurrentC _rxmod1_6b475db22b3a2d25441a4322fa9c2aef8
-#define _sumPS _rxmod1_6b475db22b3a2d25441a4322fa9c2aef9
-#define _prodPS _rxmod1_6b475db22b3a2d25441a4322fa9c2aef10
-#define _prodType _rxmod1_6b475db22b3a2d25441a4322fa9c2aef11
-#define _sumType _rxmod1_6b475db22b3a2d25441a4322fa9c2aef12
-#define _update_par_ptr _rxmod1_6b475db22b3a2d25441a4322fa9c2aef13
-#define _getParCov _rxmod1_6b475db22b3a2d25441a4322fa9c2aef14
-#define _rxode2_rxAssignPtr _rxmod1_6b475db22b3a2d25441a4322fa9c2aef15
-#define _rxQr _rxmod1_6b475db22b3a2d25441a4322fa9c2aef16
-#define _compareFactorVal _rxmod1_6b475db22b3a2d25441a4322fa9c2aef17
-#define _sum _rxmod1_6b475db22b3a2d25441a4322fa9c2aef18
-#define _udf _rxmod1_6b475db22b3a2d25441a4322fa9c2aef19
-#define _sign _rxmod1_6b475db22b3a2d25441a4322fa9c2aef20
-#define _prod _rxmod1_6b475db22b3a2d25441a4322fa9c2aef21
-#define _max _rxmod1_6b475db22b3a2d25441a4322fa9c2aef22
-#define _min _rxmod1_6b475db22b3a2d25441a4322fa9c2aef23
-#define _transit4P _rxmod1_6b475db22b3a2d25441a4322fa9c2aef24
-#define _transit3P _rxmod1_6b475db22b3a2d25441a4322fa9c2aef25
-#define _assignFuns0 _rxmod1_6b475db22b3a2d25441a4322fa9c2aef26
-#define _assignFuns _rxmod1_6b475db22b3a2d25441a4322fa9c2aef27
-#define _rxord _rxmod1_6b475db22b3a2d25441a4322fa9c2aef28
-#define __assignFuns2 _rxmod1_6b475db22b3a2d25441a4322fa9c2aef29
-#define _llikCauchyDscale _rxmod1_6b475db22b3a2d25441a4322fa9c2aef30
-#define _llikCauchyDlocation _rxmod1_6b475db22b3a2d25441a4322fa9c2aef31
-#define _llikCauchy _rxmod1_6b475db22b3a2d25441a4322fa9c2aef32
-#define _llikGammaDrate _rxmod1_6b475db22b3a2d25441a4322fa9c2aef33
-#define _llikGammaDshape _rxmod1_6b475db22b3a2d25441a4322fa9c2aef34
-#define _llikGamma _rxmod1_6b475db22b3a2d25441a4322fa9c2aef35
-#define _llikWeibullDscale _rxmod1_6b475db22b3a2d25441a4322fa9c2aef36
-#define _llikWeibullDshape _rxmod1_6b475db22b3a2d25441a4322fa9c2aef37
-#define _llikWeibull _rxmod1_6b475db22b3a2d25441a4322fa9c2aef38
-#define _llikUnifDbeta _rxmod1_6b475db22b3a2d25441a4322fa9c2aef39
-#define _llikUnifDalpha _rxmod1_6b475db22b3a2d25441a4322fa9c2aef40
-#define _llikUnif _rxmod1_6b475db22b3a2d25441a4322fa9c2aef41
-#define _llikGeomDp _rxmod1_6b475db22b3a2d25441a4322fa9c2aef42
-#define _llikGeom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef43
-#define _llikFDdf2 _rxmod1_6b475db22b3a2d25441a4322fa9c2aef44
-#define _llikFDdf1 _rxmod1_6b475db22b3a2d25441a4322fa9c2aef45
-#define _llikF _rxmod1_6b475db22b3a2d25441a4322fa9c2aef46
-#define _llikExpDrate _rxmod1_6b475db22b3a2d25441a4322fa9c2aef47
-#define _llikExp _rxmod1_6b475db22b3a2d25441a4322fa9c2aef48
-#define _llikChisqDdf _rxmod1_6b475db22b3a2d25441a4322fa9c2aef49
-#define _llikChisq _rxmod1_6b475db22b3a2d25441a4322fa9c2aef50
-#define _llikTDsd _rxmod1_6b475db22b3a2d25441a4322fa9c2aef51
-#define _llikTDmean _rxmod1_6b475db22b3a2d25441a4322fa9c2aef52
-#define _llikTDdf _rxmod1_6b475db22b3a2d25441a4322fa9c2aef53
-#define _llikT _rxmod1_6b475db22b3a2d25441a4322fa9c2aef54
-#define _llikBetaDshape2 _rxmod1_6b475db22b3a2d25441a4322fa9c2aef55
-#define _llikBetaDshape1 _rxmod1_6b475db22b3a2d25441a4322fa9c2aef56
-#define _llikBeta _rxmod1_6b475db22b3a2d25441a4322fa9c2aef57
-#define _llikNbinomMuDmu _rxmod1_6b475db22b3a2d25441a4322fa9c2aef58
-#define _llikNbinomMu _rxmod1_6b475db22b3a2d25441a4322fa9c2aef59
-#define _llikNbinomDprob _rxmod1_6b475db22b3a2d25441a4322fa9c2aef60
-#define _llikNbinom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef61
-#define _llikBinomDprob _rxmod1_6b475db22b3a2d25441a4322fa9c2aef62
-#define _llikBinom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef63
-#define _llikPoisDlambda _rxmod1_6b475db22b3a2d25441a4322fa9c2aef64
-#define _llikPois _rxmod1_6b475db22b3a2d25441a4322fa9c2aef65
-#define _llikNormDsd _rxmod1_6b475db22b3a2d25441a4322fa9c2aef66
-#define _llikNormDmean _rxmod1_6b475db22b3a2d25441a4322fa9c2aef67
-#define _llikNorm _rxmod1_6b475db22b3a2d25441a4322fa9c2aef68
-#define simeps _rxmod1_6b475db22b3a2d25441a4322fa9c2aef69
-#define simeta _rxmod1_6b475db22b3a2d25441a4322fa9c2aef70
-#define expit _rxmod1_6b475db22b3a2d25441a4322fa9c2aef71
-#define logit _rxmod1_6b475db22b3a2d25441a4322fa9c2aef72
-#define gammapDer _rxmod1_6b475db22b3a2d25441a4322fa9c2aef73
-#define lowergamma _rxmod1_6b475db22b3a2d25441a4322fa9c2aef74
-#define uppergamma _rxmod1_6b475db22b3a2d25441a4322fa9c2aef75
-#define gammaqInva _rxmod1_6b475db22b3a2d25441a4322fa9c2aef76
-#define gammaqInv _rxmod1_6b475db22b3a2d25441a4322fa9c2aef77
-#define gammapInva _rxmod1_6b475db22b3a2d25441a4322fa9c2aef78
-#define gammapInv _rxmod1_6b475db22b3a2d25441a4322fa9c2aef79
-#define gammaq _rxmod1_6b475db22b3a2d25441a4322fa9c2aef80
-#define gammap _rxmod1_6b475db22b3a2d25441a4322fa9c2aef81
-#define phi _rxmod1_6b475db22b3a2d25441a4322fa9c2aef82
-#define riweibull _rxmod1_6b475db22b3a2d25441a4322fa9c2aef83
-#define riunif _rxmod1_6b475db22b3a2d25441a4322fa9c2aef84
-#define rit_ _rxmod1_6b475db22b3a2d25441a4322fa9c2aef85
-#define ripois _rxmod1_6b475db22b3a2d25441a4322fa9c2aef86
-#define ribeta _rxmod1_6b475db22b3a2d25441a4322fa9c2aef87
-#define rigamma _rxmod1_6b475db22b3a2d25441a4322fa9c2aef88
-#define rigeom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef89
-#define rif _rxmod1_6b475db22b3a2d25441a4322fa9c2aef90
-#define riexp _rxmod1_6b475db22b3a2d25441a4322fa9c2aef91
-#define richisq _rxmod1_6b475db22b3a2d25441a4322fa9c2aef92
-#define ricauchy _rxmod1_6b475db22b3a2d25441a4322fa9c2aef93
-#define rinbinomMu _rxmod1_6b475db22b3a2d25441a4322fa9c2aef94
-#define rinbinom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef95
-#define ribinom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef96
-#define rinorm _rxmod1_6b475db22b3a2d25441a4322fa9c2aef97
-#define rxweibull _rxmod1_6b475db22b3a2d25441a4322fa9c2aef98
-#define rxunif _rxmod1_6b475db22b3a2d25441a4322fa9c2aef99
-#define rxt_ _rxmod1_6b475db22b3a2d25441a4322fa9c2aef100
-#define rxpois _rxmod1_6b475db22b3a2d25441a4322fa9c2aef101
-#define rxbeta _rxmod1_6b475db22b3a2d25441a4322fa9c2aef102
-#define rxgamma _rxmod1_6b475db22b3a2d25441a4322fa9c2aef103
-#define rxgeom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef104
-#define rxf _rxmod1_6b475db22b3a2d25441a4322fa9c2aef105
-#define rxexp _rxmod1_6b475db22b3a2d25441a4322fa9c2aef106
-#define rxchisq _rxmod1_6b475db22b3a2d25441a4322fa9c2aef107
-#define rxcauchy _rxmod1_6b475db22b3a2d25441a4322fa9c2aef108
-#define rxnbinomMu _rxmod1_6b475db22b3a2d25441a4322fa9c2aef109
-#define rxnbinom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef110
-#define rxbinom _rxmod1_6b475db22b3a2d25441a4322fa9c2aef111
-#define rxnorm _rxmod1_6b475db22b3a2d25441a4322fa9c2aef112
-#define linCmtC _rxmod1_6b475db22b3a2d25441a4322fa9c2aef113
-#define linCmtB _rxmod1_6b475db22b3a2d25441a4322fa9c2aef114
-#define linCmtA _rxmod1_6b475db22b3a2d25441a4322fa9c2aef115
+#define _getRxSolve_ _rxmod1_00d345ccfacba45976fe2f0b96dc54f90
+#define _evalUdf _rxmod1_00d345ccfacba45976fe2f0b96dc54f91
+#define _solveData _rxmod1_00d345ccfacba45976fe2f0b96dc54f92
+#define _assign_ptr _rxmod1_00d345ccfacba45976fe2f0b96dc54f93
+#define _rxRmModelLib _rxmod1_00d345ccfacba45976fe2f0b96dc54f94
+#define _rxGetModelLib _rxmod1_00d345ccfacba45976fe2f0b96dc54f95
+#define _old_c _rxmod1_00d345ccfacba45976fe2f0b96dc54f96
+#define _ptrid _rxmod1_00d345ccfacba45976fe2f0b96dc54f97
+#define _rxIsCurrentC _rxmod1_00d345ccfacba45976fe2f0b96dc54f98
+#define _sumPS _rxmod1_00d345ccfacba45976fe2f0b96dc54f99
+#define _prodPS _rxmod1_00d345ccfacba45976fe2f0b96dc54f910
+#define _prodType _rxmod1_00d345ccfacba45976fe2f0b96dc54f911
+#define _sumType _rxmod1_00d345ccfacba45976fe2f0b96dc54f912
+#define _update_par_ptr _rxmod1_00d345ccfacba45976fe2f0b96dc54f913
+#define _getParCov _rxmod1_00d345ccfacba45976fe2f0b96dc54f914
+#define _rxode2_rxAssignPtr _rxmod1_00d345ccfacba45976fe2f0b96dc54f915
+#define _rxQr _rxmod1_00d345ccfacba45976fe2f0b96dc54f916
+#define _compareFactorVal _rxmod1_00d345ccfacba45976fe2f0b96dc54f917
+#define _sum _rxmod1_00d345ccfacba45976fe2f0b96dc54f918
+#define _udf _rxmod1_00d345ccfacba45976fe2f0b96dc54f919
+#define _sign _rxmod1_00d345ccfacba45976fe2f0b96dc54f920
+#define _prod _rxmod1_00d345ccfacba45976fe2f0b96dc54f921
+#define _max _rxmod1_00d345ccfacba45976fe2f0b96dc54f922
+#define _min _rxmod1_00d345ccfacba45976fe2f0b96dc54f923
+#define _transit4P _rxmod1_00d345ccfacba45976fe2f0b96dc54f924
+#define _transit3P _rxmod1_00d345ccfacba45976fe2f0b96dc54f925
+#define _assignFuns0 _rxmod1_00d345ccfacba45976fe2f0b96dc54f926
+#define _assignFuns _rxmod1_00d345ccfacba45976fe2f0b96dc54f927
+#define _rxord _rxmod1_00d345ccfacba45976fe2f0b96dc54f928
+#define __assignFuns2 _rxmod1_00d345ccfacba45976fe2f0b96dc54f929
+#define _llikCauchyDscale _rxmod1_00d345ccfacba45976fe2f0b96dc54f930
+#define _llikCauchyDlocation _rxmod1_00d345ccfacba45976fe2f0b96dc54f931
+#define _llikCauchy _rxmod1_00d345ccfacba45976fe2f0b96dc54f932
+#define _llikGammaDrate _rxmod1_00d345ccfacba45976fe2f0b96dc54f933
+#define _llikGammaDshape _rxmod1_00d345ccfacba45976fe2f0b96dc54f934
+#define _llikGamma _rxmod1_00d345ccfacba45976fe2f0b96dc54f935
+#define _llikWeibullDscale _rxmod1_00d345ccfacba45976fe2f0b96dc54f936
+#define _llikWeibullDshape _rxmod1_00d345ccfacba45976fe2f0b96dc54f937
+#define _llikWeibull _rxmod1_00d345ccfacba45976fe2f0b96dc54f938
+#define _llikUnifDbeta _rxmod1_00d345ccfacba45976fe2f0b96dc54f939
+#define _llikUnifDalpha _rxmod1_00d345ccfacba45976fe2f0b96dc54f940
+#define _llikUnif _rxmod1_00d345ccfacba45976fe2f0b96dc54f941
+#define _llikGeomDp _rxmod1_00d345ccfacba45976fe2f0b96dc54f942
+#define _llikGeom _rxmod1_00d345ccfacba45976fe2f0b96dc54f943
+#define _llikFDdf2 _rxmod1_00d345ccfacba45976fe2f0b96dc54f944
+#define _llikFDdf1 _rxmod1_00d345ccfacba45976fe2f0b96dc54f945
+#define _llikF _rxmod1_00d345ccfacba45976fe2f0b96dc54f946
+#define _llikExpDrate _rxmod1_00d345ccfacba45976fe2f0b96dc54f947
+#define _llikExp _rxmod1_00d345ccfacba45976fe2f0b96dc54f948
+#define _llikChisqDdf _rxmod1_00d345ccfacba45976fe2f0b96dc54f949
+#define _llikChisq _rxmod1_00d345ccfacba45976fe2f0b96dc54f950
+#define _llikTDsd _rxmod1_00d345ccfacba45976fe2f0b96dc54f951
+#define _llikTDmean _rxmod1_00d345ccfacba45976fe2f0b96dc54f952
+#define _llikTDdf _rxmod1_00d345ccfacba45976fe2f0b96dc54f953
+#define _llikT _rxmod1_00d345ccfacba45976fe2f0b96dc54f954
+#define _llikBetaDshape2 _rxmod1_00d345ccfacba45976fe2f0b96dc54f955
+#define _llikBetaDshape1 _rxmod1_00d345ccfacba45976fe2f0b96dc54f956
+#define _llikBeta _rxmod1_00d345ccfacba45976fe2f0b96dc54f957
+#define _llikNbinomMuDmu _rxmod1_00d345ccfacba45976fe2f0b96dc54f958
+#define _llikNbinomMu _rxmod1_00d345ccfacba45976fe2f0b96dc54f959
+#define _llikNbinomDprob _rxmod1_00d345ccfacba45976fe2f0b96dc54f960
+#define _llikNbinom _rxmod1_00d345ccfacba45976fe2f0b96dc54f961
+#define _llikBinomDprob _rxmod1_00d345ccfacba45976fe2f0b96dc54f962
+#define _llikBinom _rxmod1_00d345ccfacba45976fe2f0b96dc54f963
+#define _llikPoisDlambda _rxmod1_00d345ccfacba45976fe2f0b96dc54f964
+#define _llikPois _rxmod1_00d345ccfacba45976fe2f0b96dc54f965
+#define _llikNormDsd _rxmod1_00d345ccfacba45976fe2f0b96dc54f966
+#define _llikNormDmean _rxmod1_00d345ccfacba45976fe2f0b96dc54f967
+#define _llikNorm _rxmod1_00d345ccfacba45976fe2f0b96dc54f968
+#define simeps _rxmod1_00d345ccfacba45976fe2f0b96dc54f969
+#define simeta _rxmod1_00d345ccfacba45976fe2f0b96dc54f970
+#define expit _rxmod1_00d345ccfacba45976fe2f0b96dc54f971
+#define logit _rxmod1_00d345ccfacba45976fe2f0b96dc54f972
+#define gammapDer _rxmod1_00d345ccfacba45976fe2f0b96dc54f973
+#define lowergamma _rxmod1_00d345ccfacba45976fe2f0b96dc54f974
+#define uppergamma _rxmod1_00d345ccfacba45976fe2f0b96dc54f975
+#define gammaqInva _rxmod1_00d345ccfacba45976fe2f0b96dc54f976
+#define gammaqInv _rxmod1_00d345ccfacba45976fe2f0b96dc54f977
+#define gammapInva _rxmod1_00d345ccfacba45976fe2f0b96dc54f978
+#define gammapInv _rxmod1_00d345ccfacba45976fe2f0b96dc54f979
+#define gammaq _rxmod1_00d345ccfacba45976fe2f0b96dc54f980
+#define gammap _rxmod1_00d345ccfacba45976fe2f0b96dc54f981
+#define phi _rxmod1_00d345ccfacba45976fe2f0b96dc54f982
+#define riweibull _rxmod1_00d345ccfacba45976fe2f0b96dc54f983
+#define riunif _rxmod1_00d345ccfacba45976fe2f0b96dc54f984
+#define rit_ _rxmod1_00d345ccfacba45976fe2f0b96dc54f985
+#define ripois _rxmod1_00d345ccfacba45976fe2f0b96dc54f986
+#define ribeta _rxmod1_00d345ccfacba45976fe2f0b96dc54f987
+#define rigamma _rxmod1_00d345ccfacba45976fe2f0b96dc54f988
+#define rigeom _rxmod1_00d345ccfacba45976fe2f0b96dc54f989
+#define rif _rxmod1_00d345ccfacba45976fe2f0b96dc54f990
+#define riexp _rxmod1_00d345ccfacba45976fe2f0b96dc54f991
+#define richisq _rxmod1_00d345ccfacba45976fe2f0b96dc54f992
+#define ricauchy _rxmod1_00d345ccfacba45976fe2f0b96dc54f993
+#define rinbinomMu _rxmod1_00d345ccfacba45976fe2f0b96dc54f994
+#define rinbinom _rxmod1_00d345ccfacba45976fe2f0b96dc54f995
+#define ribinom _rxmod1_00d345ccfacba45976fe2f0b96dc54f996
+#define rinorm _rxmod1_00d345ccfacba45976fe2f0b96dc54f997
+#define rxweibull _rxmod1_00d345ccfacba45976fe2f0b96dc54f998
+#define rxunif _rxmod1_00d345ccfacba45976fe2f0b96dc54f999
+#define rxt_ _rxmod1_00d345ccfacba45976fe2f0b96dc54f9100
+#define rxpois _rxmod1_00d345ccfacba45976fe2f0b96dc54f9101
+#define rxbeta _rxmod1_00d345ccfacba45976fe2f0b96dc54f9102
+#define rxgamma _rxmod1_00d345ccfacba45976fe2f0b96dc54f9103
+#define rxgeom _rxmod1_00d345ccfacba45976fe2f0b96dc54f9104
+#define rxf _rxmod1_00d345ccfacba45976fe2f0b96dc54f9105
+#define rxexp _rxmod1_00d345ccfacba45976fe2f0b96dc54f9106
+#define rxchisq _rxmod1_00d345ccfacba45976fe2f0b96dc54f9107
+#define rxcauchy _rxmod1_00d345ccfacba45976fe2f0b96dc54f9108
+#define rxnbinomMu _rxmod1_00d345ccfacba45976fe2f0b96dc54f9109
+#define rxnbinom _rxmod1_00d345ccfacba45976fe2f0b96dc54f9110
+#define rxbinom _rxmod1_00d345ccfacba45976fe2f0b96dc54f9111
+#define rxnorm _rxmod1_00d345ccfacba45976fe2f0b96dc54f9112
+#define linCmtC _rxmod1_00d345ccfacba45976fe2f0b96dc54f9113
+#define linCmtB _rxmod1_00d345ccfacba45976fe2f0b96dc54f9114
+#define linCmtA _rxmod1_00d345ccfacba45976fe2f0b96dc54f9115
 #include <rxode2_model_shared.h>
 #define __MAX_PROD__ 0
 #define _CMT CMT
@@ -684,7 +684,7 @@ extern SEXP mod1__model_vars(void){
   SEXP _mv = PROTECT(_rxGetModelLib("mod1__model_vars"));pro++;
   if (!_rxIsCurrentC(_mv)){
     SEXP hash    = PROTECT(allocVector(STRSXP, 1));pro++;
-#define __doBuf__  snprintf(buf, __doBufN__, "un]\"BAAA@QRtHACAAAAAAAjnzAAAv7#aT)^XQ$MAI<8a=W6xYB]a`D57AeqxLOA26CsgiJ^H207;*5%%*m+~,842+[*<<gQ0)>5ciWi@^i,T~)B07BY57jwL,1mJt,A_CCaj$91>uDm>/0s7IiP(U=]j+R9=mX|N3Z2t%%_PtW2[:{Z+0E6VoFy(${51UShkVW@}3r(Ir/&v4H+y_]p_15QnR^^>hV;!\"5rLJw@1z8\"O.Mqhk&M/&2z>n0fs*b:t\"Ew+gzU,)mTdMOQkl34Z5nrvpu\"S)k;W+U]a_)rufJ.2ZB),KLJ0v(v@C!uMe<o0ziED9SYMJY^Ql3B%%/82c&XmPdyjZL,0p=w>K4cGU!R01xp=zaNqh`,~Es*\"L:,hzBFuv\?MHEkt/M5g<!!\?,yn;CN4xRuuiL0IMxCdjmSK57F!:j+95uhvM_\?PD:0\"9MHjZUbIbo(@dVr!5u%%il`A}ly;#%%U<4TDf\"g%%n>ir7qC.4]n\"F*aEy1@s(oovyh;pg9.nn6D5+>790PU}Z7@DFL)&4/#=|d]G8\"]X3uE!+.3VY2x$k%%/l57CoSFZvg6:l&5D]|~+_2L|PGuB#rdO>3hs{^+EMFdydmsN,UX0nj@KXV9^6\?\?)gQ/_gSy7^>\"^<oh\"Vx=:z@b7WK~[d:)u=crD}l,Z\"N9r#B2gn{GFO9G<9|.a9~]#!4P\"^Ec(oD*hZ+gJa6d2nd~rN=F2v[,*JR>O4;9^`Oun62OrLp*frQWmg,%%P{E<P>)QM[_V&#R[0V&b@1^&nks~q*k%%IwfI2/06/oh7SHR`jxx\"IBb7Nc;MAm\"2OAe!tIA&wqH<M8D,Zs`MCMvQ`%%RY|6lz{R#m~_{NWlrnK)a]YB~J+#O`%%/kEdhhW7{^chc~9S=E|;<outw7)BPwk]i}gF7JZW(#xvCgIj)>pIH7pz<eZ4=(V*YcsE%%z!yh@(9v=bB%%3;@:3`e>B|rU3B5kI\"[]i}[AAn0.0=FhFEZT)oTfNu*)H{a.a=Jxke6&HPW|*Vf}4vW%%O9o464BRGMDHajjLlRm,B@f8l/<D");
+#define __doBuf__  snprintf(buf, __doBufN__, "un]\"BAAA@QRtHACAAAAAAAjnzAAAv7#aT)^XQ$MAI<8agtVwYBaiiq@CXx4gYer_m%%1)sPA(a!CM6RZX*jVDCF[EZ<i|>mtRsmE*U~GS|m<9ZNrg@!uLmCr;,hJt,A_CCaj$Cw!uDm>/0s7IS#i3n{:o3;mT5r1.6xo%%{Z\"sQ/4^anOJZr@KM\"${51UShkVW@}3rDIr/&vCJjZ|SVq//eI2pqn#3m<\"c)yauV2z8#P.Mqhk&M/&2z>n0fs*bh)l=CzL@e<WQTdMOQkl34ZUArv45\"S[a<qxp24\?DK0fvaM;Wp`COJYk_LUv8/(}w_waeHR2)jGBYcM~+S=ylQ{h#nQ@kmu;r(;z+R)>MYuD#DXh[\?1|c`L5gRRTVMoeOQwRR=T9RN!~;fz.j:%%K>;vtUhjR\?c\"H^nzZ@24Gxz6:a9t2%%SKqS{l:MNAF%%uGATl)n77O6OS|W,FSTHoT_P6(A=\"U\"P\?qhOlq~,:*9YGEs4i9>u8EsO;0F|FuQ;=O<NEe*GJ9w8<VS.1g1<)f7e>$XeOpM;QL%%;UC=N\?:x_Z9d^(DbjGuLCXE8kDXPonXa>Sed|#>ITV*anrC\"9G&k<~PI&y0W1WI+r72>Ip2e&x`@kF`/so>OvPNwTvfnm3ru3\?^\"L;\?U{N%%:OzCB}1#RiTHj)t2{cXl_3jt7eKs+X.lC$_&#G`~~88k2(qv8>.M[9N@CFO:Ob%%O\"V1^Cb*o[iyIo0/l={~:)Te.0VzU[8>=>1%%iw;\"/=r`m_`$!E{bz]w0`&qUJ1ZJTUhj6sNoh$i@jF6U9dJxI}R+YtEuN<z7lGATt2OAe!tIA&wqH<M8D,Zs`MCMvQ`%%RY|6lz{R#m~_{NWlrnK)a]YB~J+#O`%%/kEdhhW7{^chc~9S=E|;<outw7)BPwk]i}gF7JZW(#xvCgIj)>pIH7pz<eZ4=(V*YcsE%%z!yh@(9v=bB%%3;@:3`e>B|rU3B5kI\"[]i}[AAn0.0=FhFEZT)oTfNu*)H{a.a=Jxke6&HPW|*Vf}4vW%%O9o464BRGMDHajjLlRm,B5U\?QN(B");
     char buf[1060];
 #define __doBufN__ 1060
     __doBuf__
diff --git a/reference/myapp/mod1.d/mod1_.o b/reference/myapp/mod1.d/mod1_.o
index d1a119180fec3953e60b76d6649a00d170bef8f6..cda87e1a6f7f0890604d3d7989c9769349ad280d 100644
GIT binary patch
delta 21287
zcmcJV30zdyxxnv?;DRfnMp1D<A&L=ZhM9o@WDyWSU~pIzT*d(gm|+$ehDA_rW7_-@
z(<Z4CJt<i<xTGN_G1^E&(=;{BYeLi1XVSDW=}VK><TWJPByDPA-%ESW-0vcMfE@bL
z`ThKT-v8OZ<(zXr?uE~$UifV4>33F%b-wbpiVaq~y(uGWTf45V!sfNOqN6oynu{}6
zTO+dy4d%*iv5xZWtc}}qDpT!y(_B7tvnwaItSh%<uPfD9S>&zDh-@ivWqDI>sn7S_
zl4;4xvN~2}wMDjMyVBy-&53y>EA0vH&SXzQimO!b)x;UvJ<-vfl~K)G#CY8<tJ{=c
z>CM<&7`3WXuPH03Zz$iIu`j`!o#%~@Y0b)aRPHS)Z;Fq%<!Kw6+$$4YsZrHy_E=lW
zwH-+-i{lEL>KkGT<72cPX>N6Gb8J=p8t3|~j)DwZ>eel#HHD@v4bB8zWm%@TRi}?C
z*=|j3YH!r)(z7?KYg<KYVanFV-C3smiUvzTS!+`K>V0jNHmhxuS-ZxOyP=}Guxo9}
z`aRAqo%Ji-TQVA4Eh*mBDIIIgYtjlUTC!JWZ>!&)uGw3?yFg!OHkXzcdTgdGZD#ZO
zQlq&jvMZ&wBhHZ=9lbd%rPWtf?=YuVM{U~W?x@Vr>T=sV3}Uj?=+j0=muzk->2xL+
z7G-XXj&8}#EOoANCnlN<rjor|Z`oK<y2(>&DK@OmTpv{yRh-^bZ*n$u-I5$@tjNu2
z*4d3|ElD}erWEzYy&GG$q^vik7UkGtbMuXDNvY)}ahf_!>YCa_TiU8^HdA9^%9euY
z=#;Xo^+oB)V!Nf%ShH(&c9kQyr66ltPEtemwqlQan>DS(R<n0yT3%^+tkvz!iZbTL
zG^BPF=f&(%Z`EwgTkB2KrsXwiZ87b;Ga5GTj&Js)nOcjB+Dcn%-L3?0OG5gtsDhOA
zvd*ID=wwqz)`oqWm1cYCCdV3|wzGPh(~{uuCECQz&1<*CIkQ$-Vpitv?&?ZPYS6B3
zYE4RuE8K1@%E;PiN;XwRN1LL`HM_E^Vm;gKd-6Q$c<1`guH-ePwm5ZpN7?qIwDoD4
z=xBReN>u!M?{1f6_o~R8&Z>@`ZR*`IO=bC-%KTEQdqZ=Pr+$?^!?Dp&W~|9?&)a8S
zJG;V{(^8b$k?5p4_xb<(=8tCT%hI<Lt==HepZ`x!tIl3sy{0|clAY}<@mYPga;w*(
zu~-C$wc1kcX>t2(UcusR&$amqJvBCGfxW79Poqt!_Ox1>t*%C=&EL|KA<`NyPmRW+
zR@Wrx^wrh1*6J#&PH#*!)Y`PQ>MEn5rdqG7HEPsq)*EMC4R2kMcC6*COzS2HPP;qH
z<;%1*PNJ25+`8iEbsEO4Ji2b;B<d!ORd?|?$~O9+?^)shq$g%Ir_K}n^Ny|Kx2A^M
zLPm{-+dw;L<%F@8Irb_wx=Ga4`*V-Qi<aW-A!oM;0dKUpe0G;jsHHAnBUpVNC--!z
zKTTo+tg6W8|C5-(B-BN<C31R$W_*ty<;0Wl_e6uhbX{8SwOMQURcG{HK9<Psn_7OO
zX!*TZEAksf%Wo8|PO#UxJziT)wx`+FTr|E?M6{u5`Sq^l-)6L9?lLU=!bWZBTBBcc
zoQ4#qPbH`YYn8|A8-Iytg$Y#S&30F9i~B}R#1Y3#f=+PTTJ1Gj!R@JP<~Laazlst@
zswm<{r(S~5-*J2|w~BQ9uB+pBUEL&HSI6(VI_}+9H0t>G6P-b@wfnr*66z{8{!KzR
z(kdg25{ev4XY@}!vChBq#Nuhv39!+B@k9c*mGu0o*YkUzUgTH3o?G=sJ-6mXqn>|D
z(euZP{sxvAZkS+tOVk_vK_|D0W=p%%=C=5RlAKaYNvS!nDA$rxP<o?nV8kw9;EyB2
z7{(OGxx*Uxt!dylreTyZ#c^-Z25xH_4gAJ5AY(4S!9CuP<bUX7l5~%;*xmMKp}OAc
z753EI>*~4fqv3Q3>Jf=!no=C|LYO#)EyposBEKmU$1vr19+IdNoK{y=jg?y!^h(5g
zfF@oyQ{hCfTZxm<u+bm-XczbHmo%n9#<WKgzde#BaYiMLVUTgmoHQ!7#tSXcn8Y7j
zNe01XZO?Q%OX+R4ncIL#H!uo5qmqpN#~&@$_$=nk(j4v)w0eK%V+L+p8u@K$<hP}9
z61Ft*+tN4*TN?RoX*39po*ECgij6lgA{=KSa?-<S^gr|1MRB5FP2ksULV|QVVs8t~
z`SNPoC!D!no=yV4RuT;U4|;F&-}m?j3s}OZ(dQMatTlOVyU<ooQ#t;DQ_C`&J<is^
zB!otryT<OW3%tTiI30O5y2P{lo(NaO@@{{6wg2#wJAxW*Ui;b4o(xj>Ri|V8-qTNU
zcaW&}tDbK0|MclDzvE2S_M9v|zxyX$-+L#rrbeUopE?ui@BBf@jJA5a(`IY8`~30e
z7ITLs|FrlWhn{z0oy})%FJV@rS1Vq%b4T5TF(+Qn22SG8<DTe*aTzj{j5PWy&Z)Tl
z#+c}DKWADD$s(2uwX|Bj%|athn9{)T$NW%b^0rg1wK$#jhSH7ng;jdCYc2L#H$5wS
zUi-vxDhGx}t@FR~?AjUTT8pb8i{7sMkNhZlIt{;o>HNLVS%cj)i>dY3Kktty+-5Ro
z7YPATB3LZ7E+3>g%`~U$?|k7_vDsQ{tF(DNBX<sNy&8;y(^F^nZL$mX*5-O)kJHm8
zug_BJ)(SM`>+?{3Se;G_OYbz(S8yAhEPd7Y(TwrgwLdf6963?cXjw+u|D%t?R1?+o
zU-|f%i7UPM$&Sqv^`zl<L%-v)al(C^e|w$KSBG46Sc+=G@h6jxG;-U2Xpa#mPV)1w
z%P6>Pu4WtmKBQHP0*#1fnl9s>`uN$bQOPXh#0nAR0M%**v!%#twppG0Q{lhp(3~I*
zI66dLb5SN+v$M|Xa#=P0h)?HC=mT1QC!9nEap**v*a!H}i^(LJWhWFl{?aswS37R=
zY2f@9dH&SJuL7DA<voX`OCsMN5Lo7vH!I8Umo&G^{Rf0eT{Zb9n<(<v<==}ECXRhM
z^u>Quik8=4Vy*GtlA`6!(2DGDxim$~>s~tkd-IJlxI=kDQTl7Lj^ARFxa4yrc8AV@
z$z*o<Lu|st?yx&Fe`fb<Ozeuo|EABI*$pIixn1Ul$y;&cR|R8bZ^uameLczBQGf3r
zoxL3|!({wAlDFdMFDr2N_J+w@ek+U}Vd7Zdc8r~)9Wy2P|4q?~V}D&TR*rTYKcFdE
z-pwm<^j%6E@B5a2Ess0V8_K9Ot~mY|Hi=_fY%GKQ>p5I;#2?m%wq<EoPU1bqYK{MD
z4p+SKS4NY|;qv;yu;kB#_wO6!c(J<p4K|dh<+%S&CR59AFi6yjJl}fpXKIZTCXKG=
zjA)|P;uZP6AT#n`J2>*W^5u^vKKbvT8nyh+!tu(&{S6zv|FMtQcs-Y<k|t4i^maFD
zw-bjx=i|QvYSjGJ(5P>GZ69&}(5U$%h-PK^H8K9zZQ`iJ!6ZI!`x?G(+c5U|r<ugN
z^r0ffT@&Wdp7~Odev9&Ni%9Z1MXNY{3HxSL5Y|gV2%&N={Z|Cfn-M>>NCIZW@0c!u
z3dHYREP>sKzc4ERM4=89sG}s{Mm#870xgJNU^Aw_y^<JOfI%gJdl3KO90@!S5TY>D
zK($ONI4X%D{vBv=3h_?hFCqTM0tx&l;&*_i-yY8C_`)~mOXP#$1@!+rVa0z!JQXH-
z4RHdhp01DzWK%v41J6f%DdlW@;TBZzKQOQs@qU%mu@UjpV1iu4&%%TX5WgE1T&a-8
zA5wMJQfUA?YS0=i@uv{~5S;Nj#OH$#y@L40g;GD4hd5n7^LBXvg;J^@tM0N&dY*&$
zdeG2P#9x8>s}a8n3)3UsE8ZjxumKh1!}E5;3t$EHh|2@EAburM>VH4tUjP^HK?PAz
z@EGF#QoIW15O1F+4e%o3=fU<D5LdzT?<6S>+5TY|cn~${g@Lak{$1E`mQ9f+Ae$^@
zn)JK|@z<aq!w4QE2&Sc^SF4B<+rXwfM^q4mM<LD~h#y}r4b+DC8}R&1#4oRq>VFsU
zZ%Gpqg@;kWK~U}Eh?hb9{}Ayih=!LC?*da?L_9c5n$WueKJ)rccrSS`L2;9K0>=Lm
zwYvh&Je4Ys{bbG0f_a4@{x?vp3h_I@QtJ?RNd1dKGAeirI?6@75B8H%#A8%}`3Y9U
zyP+R9;++s&odHg_26<(l!VY=|YS0HhdkFDz;5~>x2mC3-_k(LZH=NV)1p=ymWq1MQ
z--eE^AbvZT>hFjj0aH$gjZ|LPN6`O5#Gj=V*!V&eD(HZyi$~l96Dmji0t`@%xDQm_
zh<I!i8&weY1*4<kKCsyV)Zm+JDfD+5@dqJlpF#XL;H1As{0ZRi4{>S}b~LnrZT>V=
zAo2MmG?4~LF*xyzsex$>xlAu8Y7yd*u)vjw8=xOG;=fl({jNv+pyZOGkdF%91x0N|
z{6pBTY7mcyeYFYkM_|9~Li~ORwz~qHI*IH{b#U*wA2m>esZSst0V_L$_<5M<FAzTj
z^?y5@)A5Cq;Pf91FQ9xj_}=G;m(eoz`v>CNV4_RF6!OBpf`w@je~O;nY<wXd6=)cw
zzXHVX0Mk?;{s}zSBK{oIZ$^9-Y)9XiCXi!BMYQ-D%<v4V*aSZN6U1vkGrvVV9-co#
zTmb_v2B(%a^N!@qq7aP=ehks4Mf_Lb{ppDJ$~Q&CUt$qRe@%!#57(v>0Zu(tp3%-_
z<YKZSPICl0d=)kP88mzg@gAZ{q0#(P(6}%|8dvVN1?D{)aV79@#P5OK<YvSVuny@@
zhxqH#bVVT(73=~_6(Rm@@cBB#<paQnxV+Kdiuk*rkGlhWSTyg4V0ZvEcnwaH?<4*k
zMC~((KcEUMSU8V(9}N7f;haU2d_8(+cmd_|b?YO<<?Gg85tpx9Q$XdiDdD;$%tJhz
znm!QC%TYlKY!k7F-wZC6i1<gavTVff1XFH7JQ@1g6@reu_rL@Vs6jb2Xh(cLoWb8h
z{4hA_LBwx{`bUO1T|W!BRd6vpHB=yR4_q+W?STE5VD<kF+rxh#9tF>@BmQkTT7HkX
zwix!SE2!Wwgp-n9QrV9cvw;`G%9bLY2=!N!%yb&ndQ|+3ingV{4Tv9rW+ueH0lUHu
z#7{vG+7bUb#6qjYufOBm3cJp2s6i`im-iz6XV`}yL;Tkegy#^iQU%r}yf}iBm<&a@
z_#u@2ZbX?N90gquB5nuoyNdX6P(~0KQMSeBz(WyVAvu#MM4*BfK^bciKL~D_gt&Z|
z<sdE}erCkK1I}Ry@RI9ZGXQSvL=7&ikhHf4@t**{9r4TLyP4FxvS6H}sQ5gr;|avS
z0mI90)UpDj;9~W2RR0Tzt~U_>gS4NB!uzP;D0K8G;wfO!uMnqq?Z9u=Y)Q%!8iawv
z5nn7vSAf$!LvBz71qRgMRfvad#P`5PW=32()dgWU;)`Hkau4TpeBnM=!QSBolwXDc
zP9lC6SoDvGSA&MaARgs~y$%}9Li}TDN;bYwj0*H%doSW)prP+0z6Az;1M%nC4CwC~
z;%~#kRC5FS;B^!HRYkPfin#nu;ABVP0wu4c5e}iVi2nsN@=L_Mk))UQC%y!}@D(aQ
z4hvlbJCZz(+@Bh8FVs&%{87-b3Gs{4Iz(YNDtH+d;zT?YwxK4(-v+1Ohj<zEb0^{%
zs(|8!?f|DomRHsfCVU<>m<}fVBjV@464OJa4&{FCgKNS%#E$^qI-D~n>xR?DHN1fG
zw;{R?B7PoT3i=R#5B96~5Z?ltn+c|nCt9WgC)cBb4PeU}#5;gDB7PBk=py2&l467_
zh(`dQ6NWBE55mb22m66M(<)e55#slNhV~==XE5a{#Gi#@_vIl@qk}pAtuVmVp#q7o
zC7<OgZVEpQJ{AnFAkXWQ1te+{4Wb5EksFm~z<6^0Np9B-3VR&YH^DUCM*LN9ovGk*
zaz9^6!7U15sNi?tt185EsL9!{4Dp*FSmaEjJU}lfSiT_2d=K!wfqsSsQ8@V8ooE0X
ztoShEJHVxmAueAOPa`fz`HzQlHklW|#}<LA<e6SxP6qQ7amwRSBs`zWD|1LblQ%_>
z4kj}rNH3G6N^*(GJrN{iGnL08NHUYZj35psuSSq=CL>pn0VX%EAlI0@G=;<#P#GLV
zDw$LVkpoO_4<h|c-X25-nXH^b!V9TXP9gbB#!Vp|Og<e%dYODDh+JZFb0`TZqB46f
zNoI1%T;gC-6-v69{9+y%V6uBIxyIx}OG&JW%EEc1l1a~8a)8OGd8D7ot8>X9ldsMt
z;l)(GsU-PK_9{sSlk1j~UM4?KkV{NHs~{m;sQiwCBr|DM5C@Zb1?gtet|9|W{zD+w
zn0#Fzv1TeC6-XtModP+)Bw0ZEnRG57gG}DHf`pe)nYw`FGa0&obTIj080ls5M`7d=
zllO*^kWwmbVI-N!<S^o3vSL2zW-@L*8DLVmoLpn_YAA^<qw>{IQpw~aq2vIQt)Zl!
z$<1@fAe9OgSyHZuSAH>@<d;)CFq?ERd1N-}WzsjBTw*eBHVN5EWyEZf%;cYE5eJi%
zvq(3S@w3PPlR>k{H74JhNn*EA`Q%Je$>eP_$pI#}%_RLyuA4~)nN-ds;T2RSFCqC%
zR)mlaCgVa#FOz@0iCkjx>I@Q6N#(0ENHUX;%peXXTW64NCO6L@15B!BkZVkSF`dM2
zr_wZ?R5H10Iyu1Pm(xf;m5Sx$qwRFrpD!fgJ1G9?G?LHcxrL;I$-AeKUMdxLk+*ix
zYCETqkew7y3ns}_itoco^8;z089FN8g6B67KL^`RjEZeC*N;|%QhMH?lB67*!4Ul=
zi2p&_Z?gn9D)<RJe+%)iVB2{F@x!x7>@Iq1nk(OOP`MSZTdQFr^29vQUpnGzmyjOX
zpK=<_M(t8uQT}ErId74kKes40Do=-#<lTys%G-nc-rub#4_3ZRK`kvfzpQV6t)fVw
zd{RX&*3lNfh$PB-#ofv&3(3KHTH)~!a)~`#1mdtOo>#uPoWwd5yR@@m#>Z9*fuB6X
z-@s9E3Gw`8q~D>4(?-I|BG-&;_e<dMh}Q>`@CG`M!DXbXVOaY^s9q0TSdSqdswC%G
zd)iNklP!1(F*}F#ql_InkCkx5g(CjKB64PUyG&Gm87yN${LNWpzKf3g0SyG#uyK1(
zJ-Pj9#A*9}+CDrKT=hLve%B%r?xvn3>os)k$o0tmu0vezFBS2ZL&!nepD6d|MCEdS
z9f-dPSCPYr%j29t{Auv~Q@|B;+H8sUO6w%)9!01!axpP`6s<~X!~qXg;w$sW{6@)`
zl}vUiNe`3s?^c6Mz7s;UO;i>yA{|VAZyq_{MBV;{CB)&S{lqTpd*7=tD3r75sq3Q^
z{*QW<Pq9n+#2j+oH%yBc){WGnyo&!qydK^IKSuoBNRr$_t(^p35gj*ne+h`6okF@>
z=)4*Q@;;T}a{qOx9#zbSR=S{d;iSKnMr~3!8D!5?QB`fUeftv9OXX1e>G31Sr|m=9
zsbZ*t&M-+8RJlh|(8WEVppZ^ULFP`XpoV4SOeY;^U^;o9%J4%fsl$iWBRdY7Mq>BU
z?ZU8-^zWq|o}5Co`>1??o+tZge=n~fpEKFLoFsQq`D+!~&*W!|$N-aTmy-D+ZU5v{
z(j(I6+otvnii)l<<)!(34!@#Hp)@JU;C-~@_w>c{5S4Wc`Z^9NrYe;AQ~UNGRy-M~
zp}k+K(ZS^T<$dSxXEjz#B?I51i-}i}*lxv}%D)5?<pa`l)dTeTt*IpWLApq)9mj)I
zE41GHhp2pQ3h94{?tKpjla7a_`j;N2&o9m)-A8EsJC~DzBedS9(~0*eU2MFnZ}6xh
zLZMtwH=Q0@K}#<UJ#;3t;{KjM1#+#2uJV;&a^RTMLH{wzX$RWlQhD!j#oghBuqpcT
zM{4<hBm2xJ6ep*M^-$rFf{_(ueiCsn+>d(^KLal~Pa*y&TxZWBULp0<*Y=pAV(vcy
DBBG%z

delta 21322
zcmchd3s_XwoxtxMUx@Jm_{0}rFua7BJ1+(m1OWwk2?#n0!@$hIz&vmm5Ky@_P1?pZ
zN$Vt@thUi;OcS(?(IliaaT}9v8nfA?CTXpW+a}wj)+Dyc?l!IMm+U#*|Ha`Sl*4{)
z`M%`u{LcITy62pG-}q$m8=p)*`OdBK>RMCxmb^N9L%z|R>)hU5Ql2ZB)^;?zi|f}*
zTJy>jV`5^8H?=P<(;%g8O<A?7d80Nt+t}o;b+4&+r&aeQ=V{G(Ta4z`9VKZxZ*ed!
z*S2=8z4ca$F1s_|o1LNUFc-Nik~XcbYfY`RtoCl+Ri?K}!Nf#cMX*+qw`T3MrfLgJ
zHU7e$q+53yb(^YM+G=Xo?A8a8iUJwSyA117du(;>Hl1;!ud=>AyQOAje1b7hQoLi0
zt~WEGJiT;#OWX3&jO9{qwomKoNNLOpCKT*QZSLKamh7!|l$LL7^XLs$qc70a-kDI9
z+_thdbyJ<)mK^AG70c<8H7_}BLrJl=E2(SE+W5{==PJLp*IkfrsoGLnyrq7lN1vCN
zn2{+pE#K^|aP+iVOVYIMt9N8?Ff^^(6kn;;b+@nXwU=#gX-MlU=txN@$x3VK%`5Ee
z$|+r6&=$-nH8!t`kIzZ#@vU52X6UZ$>~xoCi%PrKI$FI|@&=>1Fg~j_)ne>y=_<A5
zRODxA6D+0cTdkWKiyHk&YaBI!cw=W`qTADuyW7*Va#>>Hrd5XOd{;tqVxlkG9WZZS
zn{KG{B^0LE3wm^GI+FbUuAY{a?ah{SpU;$UOwBf1v=y6nTg_|o3rmuh8#0!y-sPxU
zmzhwvSub}M#HXg~Yqm(;x3<w`dW`0^{;EQ2SF<V0v)q_px2rGJCY9`}Sl5}iEZ`{5
z*0*%*tWPZ{Ep+rayd7m}n>K5Gt!p-HuCdgca&nTJH)hzHt9s1!*$MTl{4EJJ-K(>D
z*EE`IH}q|FWF>d5ZVKe@_M}-8$}E9pc~Z0N+qtf=+_G+!Wmmf0(pR@E-cXTHU(l4~
z&rfNwrZyMbO%BQD=nIssE|j{qrlk~DW|tdQTIw6hJAEdpJW$e};tFi6^){y^CRXe$
zSXGgoys>nvt!DYM%B|X(<&A4}^~-&29evwV=<<qoxQe6<ttrFbVy;?W+0fcOy*r^O
zuhCrAw6ZkUn-Kc1+kQW7jVsfWu_|Anf1!gP#-08wDY;t?bbI{{oz2u}FdH3>lGLcT
zOAg6sFzD<Cy(GEpmL|#WbXl~5jkl_fzp5k)t@b8clfTm!bO!j_8CcbI&S0tE;q;Wa
z8>@TTo&2p$*Q?f!wQjS}<nOX|*uCwZag1OISsz|A)#LW9^#*fo?V*gLOJ~?@75P;`
zd(c^957=y>(xbP99y*#DIi!A@+vn~GJ@QBfuRZj7L22j>y1o1p(^~`=)wM&g2mKyN
zU~Q{g0#3U_u(^XfTHKz{y^m&0cQ^a|0jDG1-{I^i3za^W5o>Ss2ZDmF!|io-`UIb|
z%k7W^pTDs~3O)Gvu2}_ZjatFk9Sqp3Xg#MRbmaJoWL|&C4E@LP1$Df&jV4%l;dLX9
zVMC)OboIDC^vL&X@_0v)EJCHN+vD`vf<je6wXLeUvZ!pGt)QejbpI2Xt9ZLHNkUVL
zJs|Y7xSLx9kKNnou!r_P*|!L6$Gq3pWe;=+UVC@0$5T!1(Gfawx<ck1fj=zyqf!!V
zGH0yoxs5Osu|vxqE7cfjpE6%o3Jc^he84OOZI!vz1)TGpKr>!w*;A&8?S6+p^qZ&8
z*N-*pOsX|brDd9wSYh$SEMDa34!H)tn#$W=r`3lp4}1`1v*qUphck+f=9Jf-oyTiA
z$t>7>&K+)=3j*EL4<(z&<8G_)yE_VLg!4|tq8B>+p01%~D2?iP9wyAhd=|E33hhpx
z!|iJx%0{Piei$o;?ik#9`nQ7<#1!6MOy<xBKe{c2b4;Dq9GdprlcC=|7v#5zP9L&9
zFLTexEQKySpC9Tz6UQCux^cuhf3WLhe#h1Ehr5nD-es4L-yL-P5wD{T&SP*b-QDOi
z`RcN&mJm5p&22P^-)IuQ(WG%0P2xA2G!COl{6>>Z0^3?Ro!xG_hv2aZ)w7f-<mcvB
zLvnK4Ts|S_atAtsG#>=qD+F7!Gg#SO#gY$wMmy||;P5VzZoBea9(yxwYa7Z=p-Vq5
z=eC%h-(q@xi|MJw+UdTpJ;=Yy)pMJ}r61E8_TjC(o<D~4oH0c0rOyoIzc8L;#;<0b
zmRrv>nP{c(2U_)bo@P2Nzs7ai(Nr#vetObrO{ybHHaB%;-YC#%$B~(|7Gbj29t^m<
zL#JOz<#u=jza<RZhHx3ik)jOzDaycYA=zaZ%|fFa$iN>H2F{pZ&lLl|^$eEK?_Nmc
zHkxi+y8&*q(fH9xLZjU=lvw%Ql>c#02QbpTc#GZV@HlPkAXy?*x@?`k_JBX=45go4
z$E^c;=H!hdBew=z#&I~Vk>7ER+>UE=8I6LEKE(JpW=2k*q%LX94E3H3aqHN`uVa%=
zDBWCMnP0|X6<N7eGV$xr#H~M<Y3#U{)q5?INttvN$E0D-{4s5^2sWF`8x)j_ph|l8
z!+hP_wsr!&6$quixQE*gW_~-EbwY>T<+M5j{xKIdtOql{CCvPmFpp&k`T7l`nLh=a
z`Q6=ob8E@rcbsO+$aA^i@i)7J>)k?&y`x3w@%VR)sZZ6dGC9p9(}>K}se1#u$>opv
zOn!@H-q2$5mo9oUje^VG6!g<ODc<Co$#0HKWRAr*xpL05oQ`>^`L?*RZeBtMJ~WJ7
z>FS41k6r1-kG5?XYq)V7K|`${Tja6&8!JSvM+nm0O;i_llpJ&X@oXe$KI1&T4L@(?
zF<&i=<M3?}yiRY2Gsry+Nh`CA+d*%%`P=a8*)q05hBK@r(JSRLd!^Iv;hzeBjMK9_
zO;>b?qK>j$wq{SW-RrgMLQDQQV@w|y-(jBD2ZnRt*gkN>1ofr-G@i>M)d}UX-*3@f
zFYmJWvz>YcC3D<~j$;Qs?i&i3>)u_@uLAn^iC0_f5<})p8e`vL{QoX7Wd7Om$3A_P
z#;XJQA9Td#HJD6KqWl))KVtKqux0iJ`5$n^<~?D<X~V#8vA^%pa`+iw<h2+*r|}zX
z%%f%4E}I@LdA*jtY2~hZ^PD(hmu=(^F(bdhZhB6YM}Esd530Niqc3o|tKQgPGX9$4
z?|Yb44;}O@JNjhhFMrQry1B)|k3}YagRukeD7ju9?TeD}ooGkvcGO-+GuRDoW&des
zUin5})%=sryz-5|>!D{}Ue)rRX_X^wc(3x!zAPHgi}M?Y7>=+JM_&2n-!{=BZ}@C)
zI@ueA*q0+L>{LAF{u>{QZ*sKD<Tu#xH(MP0UkiUuw2T&FqhG7j&jXldWv(yG_|E_q
zeuG(<!8pGrvGC_a`dI?6%jmT949Z`b-wy0e7hk`0|3|+W;jGN+@|$Ax*I>vSWcY}X
zy$0i(F`ERxHO6{NHe&ywqp!)hCo!JO5#0?<_vKqQ_LhZz65}x4sK>}WWDnIG<2Xw{
zP$u3z=Ka~ze-U%iG)E&z)(WvpE}YK3G!=x^iVz~HoJs$4iwf9)_-s~z{#7bsxPI#d
z1#Ch5<y#ccfO!0*As`FQsNgG072rcWdWHfz5r1l;0(K!jf1U#FL;PK3VBz2pmxbXK
zSZ66U52FVAp}`5n?*RT2#2=4Pz|Rmr2%7#?IH%(a_e@sE`{4!j-)*qsKOlYtCVCC=
zw?WkrU^>;5XJFuH#J5n+#uw61K^$`c`X?d&#(afa5YGk^tV8@0m{1Ag1Qu)+mGOs7
z@E8oNI-k0NOB!Y1Cs0S<24{Q@@ka2WR}fz~Rq5yBVNTc2##9FoVIXzj)P+ie8HjHH
z(=0^%tFcP`B*b-KG9%)f<Y`KSJXCNQX1EpccvwLT;sa2>6LD#Y(*Ny<zYko#4;4HM
zD}EgDh2ZVaApYOLUqoC0r+Wi&)r6pNdD!;fT&&FW3TiM72KWr|4oEnQK&mEN3cr&P
z&wzeRBRKWkr)H9VNlcgT1e?~4s2~U@p`n`L)PcSpqfB@Qs(%J1dI#d)n4{GHHsT*B
z(IX2Fp@QiUCdU!~5jfMch`$TI@e{<KgD^agcpW6FUk`EWi0aBFl6UpuJUI@={|joD
z1mjPJuveQE0*^tw4x%~^@m%P41>&zO^<^Oo6+}Zv>k$7jOtc#D2zUh85q}IU?L)i*
z6u5JU)6}4@>~nDDZ=eRpz-J#o{OTgb)JG8Shx$(-z8+lTxp2->^WR{_uY?y+-U}UF
zMLZdN@2`l*f`%f%m(_(W1Cz}`oN1ViFC?IXZD2Dsn$(WohqO_H>PNx=O^6qPs@oCo
zUCKrkgx!<Squ~hHY%gkX878DgsXC!!Q159}za>f;;BCZXfWJ4)=}egRSAuQ+FkGPU
zJtWFZgTw&iPX!ID<9`$QT*SA-0+%5k3-z^#AAl#yYQ&EzE-4Ges9-+0OfBN^unq^}
zufdaJJK`ExZy(|h!|%I?By|#X#rq(gJct@h0#hGDJbA8SiBpIlgS`4G;>Th4aUq=3
z@r5FA`uD>NC=Y-pK1IA8R{RCxD`27vrYI9r7xpGBOhVj4Qzjc<$Uz11FhB|7`@x5|
zAYKl(cOiZb>USXC2<hl6Qv`C<A|}d5VTPwr#S5UBe?|Ni(98wIV_+hmAf5vQ&WCYS
z&D;?`q$ycQL<Jv!^GJyAP#Oq1h+kHBiim#%{cJ~kCv2OJ4RP97ozaW4$oVWWU3UpO
z{CCvw1T_35;&X{^C5`6SLE{3rg4*p*AZn*0z8-ii;{OY2?>5A*f`Sc*N2$@Aiwa%@
zOO+v>0hVn>Tzvop5myua9>hNdecU_5sW+=Du7!#RQG+4~n#T}71X24m;!i@<o<%$e
z1HT^5Su`17W$%O+P_Aybevi1i-TD*a>UL`aTM_+J7Y5rcVHV<*)bwn8VKFM01U5@S
z+zBpbM!XVuKH^uwlp7K6hWho9=#lpftf&n&xE~sHBmNjXgYQJ#4MDdL@#mrbp<zzf
z&nCGGg5$(+fx@rCgno$lREU~?L;M}k+?$Ah47-!xApVjJw!exB{s!Ts0jFWbY@#ho
z>Hp~8Ld3U2{UpTOz<G^`zY0z&=8>ozng+^I`Mt2fZHSk{eBFrmLojq9-UJHyD&jJv
zxo;_a<*>8XK!e9og9=#4Gl)MjpG2>rXPLvWYq*5UCxYg$Af64D{S5I_u<(iCa;oBM
z!Iz^D-=TPpEG$6<$6z6;h<^n(vYCjhPp$&Q)hC@AWa@<UaYK$E*oNv?T@Uw8a3v4w
zs13G5J&0R@e+_XD`F1Y#j1YL-VN`q)*71GBzW~J!BEAO<s9uk$>zDuve+$)LqIk3{
zyoU-NhmQV;_$(OsbHslOE_(|+($#_g48LO$uT(>6h|>h2HuxzNm{5avW-FG+M|=P#
zQi=H2!7~~V_dyc!g>yQ-@K;#DuJ8iNQ{nfch;IaoUPe3<G!z3WR~MEB1FuDVKdr#V
z7b;M}tB^+nh?jzf)Z?Z)z;YP)EmXe*g7X^U3t-}LGl!DibrXDZA(8UM^x~JH!^04j
z>OgP9@6(9i2O4<|@rK2upY|ue2?KtP%J+ej&xM7l<EZ^<5pRI{*@&M6pDjmxH|vlo
zxd9bu;I{|yrJyM_->Va9hJ3mk)gOcf-Gz7pTofD_;$cp{0ZxTyP=ib0BbO03Dk~5o
zpq)CQ9LTFH5HAK^8_t=Ntp+Feh8Ivi7fiPgaRL(>K>XM6?0*+=8)$A?G}HKXm&&0}
z!XYO&qJj(H<POAJz{c%}{|0>MJmR&YGT>FjpH;9hBL?k)`XE1~!!uP?>8Yhki!#Im
z@cVYe?*I*)Kztqa|I=YkJ7#|JGZ^5r;R1#4C!een=fy^Wk4*xLsPlRuf+Vb`L9_z6
z50&2v<LyNJ`=GGnh#!J!{0i}>z;!0WF<Kq}ex;r)#Gr!B5Or~g9|z~(g!nE9mPW+y
z0e$!p-v)fw5SOp}R1)~wU8q4ftav}-#o$s$5kC+4@g(AZ21`62&RHL?z>4R>3RM-o
zGnrf|64N#FmXg?FDlaW1#Y|3~LVB6hP9gnFTBneUOx`htL~fw+2UAEElNY9tRwgwO
z<N%YiCz5ka>LbWCCVvn`Qc9@&WfZY8IWd~-Wilh03^HkrCRdnjSV&?^sqC6XikUPl
zB)v@D8BO|`JQ+<cGTAVdM3zw*7e%s|jEEzxOg<G)4lsFFJUPc?a~!$G<i0qPQch)d
zJh3vlDvs=BGA5o3GI=?UTw(J0I1*by<rlG}n8`O|NiUP%izWR`cF!jlnJk%4A~#aG
zd?v|aGG;z$W%76P$N?sQG>@EPvM7>VV^W?+QYxvenn$cm-ZqcyWim}9gG0Y>Ay=6E
zlR#ptsC-o*#Z3Nf8tG;79)a{T`SLV!k;xi?L{?LoDv&HD1%b3Oc`kt*VDeA`Imcu$
zfm~zq?x`eY6P39M#LDE{1hSXO59#`u{AdZeLZ$d~vY<xH&<rji#WfV)w}kXE=~zPg
znY1h+7nyu!5{ay(a>^2t#pFASNh_0gO(X}HT)LQ?WAc9%k!wuWOe86rseEw}u`>DK
zBC?lB?;<kDWZoijg~@q~NbD9WKUzqNncPg(&*TcKekL`s<RX*5SU@7JRGwHsvY6bn
zfV47Mvw$36GIaqt$E2`;Tx0U>`6Oj4mG@|fmB~wU$zCQ;%_W0WieD$c-%6LgZ7zx3
zMzL-#DQ2=pLwcF~d=BZSQhbyAavQDo*c=jBNAa(sNfwpz06b})g1n>#RS@#iTZro*
z2$#nx`C6_27CZ>@5I+bHnkvNMrg*L3Lj_eZ;GKv!LH>Fe@z1Dv>glejRNdvE@<Xsu
zO@c?KI^o@r#&QsMPbWucf0|r+^;9oj)kMuAXKl*wPi<nM=FAL|)gV5q`B*dXUV~UO
zNh8uQa?ygSsDayEVwtFUX+AmMOk4bLG10V$_i8RgkbNz*!pw=}BKzGdl2*5PMiVuM
zq_m3lQah;hXp%7Wr_NA@nO#Ku<Sa7SDyB=y;YbkTlSj6j2Aaq~{B;e9ZKLzp8ATe~
z!rDK8>K%s!d=&A+B00<2(|#g7RPND`N>5lnnv{|A_%pax6yiCN<WzXOTvT2OmM%x!
zJc&em>9{XXAXabKxJOVuwf#xNY5PIiKK3xU>bs~sGLppls3)m<jY=K49<|>Uh^zgr
zLVV9evXAyBtNnRUx!PYZ;%3nQe#F&rjv@XM*!u)<wmi1q{FbumBsqRDN@Kf)RQkm(
z4K?C9KUHG>WD?!37}Lt+ry@DRq;(p(!sL%95@|atpGta}oIRPG-A>&;X%1-((0;y%
z7<exrnnX<<ed-2jg<nh{jX|+qvqVGA2E(-Y#)^?zR9Eqzh`$b7<PQ;lZ86E}q}DEk
zcuh<ny+1wTXGC(Klg{gIk-SG`tlEDwsz()**F_h!at0afqETBogIr<1siGQpu=dkQ
zKb6DnBQi#gPuoX!Q^imPonn$I$l9YQ=zI?-C~~Kwpvs+8K|h;CPVJ-votQx0qcT=~
z5%Q3BWXG!}k(6CDT}+!w26xd8i!?;q%{rby`gha*n&yyCncN*ivihjJFP7ZS<aZ*;
zIVQg~lSIq3{hn#$h)jP!IeXxWEcV4{zBYTHH6-?lnvX?t<$hXnAeKlEP`NK+p!WfB
zvZ&cQd*Jr{;tz&uNDnGCdYL>HGjR4nRzo|RockBLm?*k8I3WI9Gif%_98`Wc9;ClN
zm`$?2Ll;T4)A}8%6<RO)yHq}_A%owgx$lV>()*B7|KdaR_v9FI;1I3vi6Q3>(R$}6
zkicQO*r<5}R}PCyM2#+nL>{3P_C}DbBXlOT;_XL<Dv)bO=qfM9kiAEh4hD}>PCJku
zQOf%t5$}z?8y*9};*na;TRKpAOnh{L{3uj-xMXAnl|PF3_uv51kNEv?Ab0}t<&jFi
RrxCwb>1SZa<KmW?{|A_xB%}ZU

diff --git a/reference/myapp/mod1.d/mod1_.so b/reference/myapp/mod1.d/mod1_.so
index c81ffec2c778278f6c7ca8514343f7dd0c9886c4..bc0e35ad164260317dc67a8b0d98a1ac3d213bea 100755
GIT binary patch
literal 90360
zcmeFad3+Q_`aj-1(=(YNkW4NJ_W%imE94{-j*yT9CXhe~34}8VIY=NEIUt-y<3Sdp
zu!}CPsPSBObzNL9bajp53f_3*s%yN~pr|a0#xuY7Q&rum8TIp>{eJfQ$M5x`^Qx|T
z-p{+9dg`g_>YnbI&PAEo^GvF$=$lzdQTioCNfip$@_sZ`aVh@LOi;!%uBKQ1nf&Z5
z4(hq$;4)dWP%)Ku*izZv$j6N3vYyNJSbOz?@0ITm_Vrw@XGN8>?Y!REYd!N*q{J3A
zfaiKW6GT$Q{EFi-zw^wyMOgIQ#?=b>Na-*c_?g_vP({&m?|LUhJ?}h1)YEfngSMp~
zHkJL~w3jRDwQS>dyz>OnKRwsmTY`EbWW3`4VK0V7$rJ7U(l|JTQ9T#UvsjW1nltkk
z6(|T#7x`iq?NT{Z$fWlhd2`3$ZEwfiUVh*A6J|VHclWZ;Y@)Uyg(HzI8frDIwJ+VI
z^qI#fKl&-D-v_oTy`vTN>>NdND%6VD%$Vt|1odwaSg&lY+>}C<y8ReIBZp0jQ-d*m
zYt3-7qC+vAebk|JU(~90^^q^M4dt0K4EYG8kw{e4i8KajEYdin2&71)|7jZ!Wfan6
zq-dl#q(mea(oCcjq%@=qq)a4vn+IV&k{f9O(n2H(+#;mKNZwm6fF($Ik|H1F0;J_g
zE0NYBl^~TNtw-93REx9;sS&9O>3`aq|0m@y=r#l2g0vNBJJL?1i;ylxx)kYhq}@p3
zutrHy_OShbZ?_kwI*_hH+K+Sq=^CViNVJR`Lb?H|6N$EqP93`Z<tI$r%0AxT(rmeB
zZaH#y!ufx!M^33qq>aca)re#YQJaP|jr;O<NE*sRk%uE4Lb5;}CdyWn|Il~Kp&KAm
zc{p+#@)5{yM54SC$qtP20mw%qzuBui5cn8iw;`Qx&OCJf@);{HgD;mOt=}`b<O<~N
zNV}1eKYRG2J;?8#HSE^)e|XG2{Dx0!dcHn%`9~qyEAML@_vO7v$*}V<(*2^Gnl@)8
zm1h8Z1nH<KKPt+zkW={~;9baPi*gEblE0f&@yf%{k$en^%BiAE`QJsdyB4_A@27RX
z3Vk84*O9Lj`6-lFp-lN2<g1b1M0qXpe;}W*<HYrEAuobVc?t4jq<2s*MP74J=Ao}q
z-iETCe*<|3<X+?xFU?v><?m45CFMUrZiP(wk4P7zOnDn}N_?jLbRTR{l;EGjN^RKB
z<DX*T{5k#xEO|eNkNO4sD%y1ZOuJ7&@pE>*{{@^@(x0<))i3z(_Y3|^`2{;S|AL*v
znxX{#jQ`eOhy$&Uv=)1BqhaUg`upB5^!M+-;E(?=^y}qcu(RbC{QUG6_=yY;P>N}h
z$7bi5QZI0}@B?!SybPTR!ddCy%^c9LQx-G-1C;I}j_Y>z34W&+|9-(gS;G10!v4d8
zf7pxPBk&_$cxVXs=cpH+D)25ZyjtKVyzqkpKk0=(A@EaP_*sE>d*Lk@FN%-tR?qls
zN1zCgy2%rd59GM(7Einu$0PC2df~SVJ275%UKF^?3tubjr+V3`M{{Jq#|!^n*h%%W
z(=O~3dht_D+)lR_UM$A<te6kis=R;q3Of$3xZNz`cFN1n1WY9I)9vNYwW8gVUi>`4
z&-LQJBiar3YImpL7kcprU;(0bonHJG&@aMIc(t1-_)%W`94qH{dGQa6c8k5*Ef#iS
zyzCq^aXUx7?7S@ORC?LjBkZ`m?U=cpBVKlX6n5&p?A$Nxq<Y!;9K#u`;CRpE;~LL1
zC12(E5wV`oH79K&LV3GAUUu99KkJ1r6n0v@>@*2}uNS{l;4NPLdRE}|Uic+qe(o#b
z{^|br3Vy2>e~$2{UigFQ$~L-gqd1&Q=Q5@l+lmCgTkwYn+$r!>FFQX%Pj;MiLcylT
zE!@oGcG%0Guf_4w>4i_khZvflEjMvHIGow`t{6wRSKP*m`FYUGPQI|y>Sbq=@IP1h
zk7>iUC$Oh>_j%d5NZ8r!WoIBxvc%8z^5<sZ&u%X}e-w81dD%%4b_%`hJS*Z;>1F3t
zVdtQiopr)av6r1N6Zf;&%T9;jS9<YrBZT#<)64!A(XW%8JYM>^OcwmZUi_oNe!Z9d
zaFx%K7_a$r(9H2-`f!FV5j@(K3HwL9?7uJUw|Lo4*X=Ci^}pnqa<Q;;)XPqQI4`$)
z*|}5jyS(_xg1_5~pU1AdFn_%80<=%ZMOQj^L_fZEn>fxsltPs&{QOe*-<iV!e-Ve*
zg+C{}{P|Sa$z8zt`mHb1Kt7Hqz4$xXbzOks5DlIe?cS~9OE{q8Q-z&VUUuHn`FWiG
zSK()&;CFlR+XUX@g^$HNXY0>WZs%5E=PtoN>&4Fz{8kDqwiFx>v|WmNZ2fU_c@)o-
zh%ga<aUQ4p5VXA^?Du-vpCI~mgl-UETPyt02Jv_*UgJAW;5ILut{2FkRN>FN+?=vf
zw{t58^yA1b__<>Jd`s}36UW`DPEY=2!cOlko_Md|cYE=V2)^5kKhBT$tH+D~rr_s#
z@fQjHSug&BaD)2WosI&wUJ*BI0Jk6RWoN6f)7#1DiXMkp!H@Cc-zxaZO`iPKf*<9@
ze?jnVUi>QrU-9C9EBFpCek2^Dbw1n+|BJ9=^RkmG_)ahW+k)@#;@iY^ozn|fii+0P
z)YTR>HkC9q6%{E(S@}6dW#tX!>#G`@${X@?(yQxg%kxXtRhM(!&ufZGTS|&5s%lHB
ztG1OZHFagNMMZUG<wcEk)tk#3cve=@RH76$v@mT<Oj&$FVrgkbN$I+hgv8{e=@sR1
z6*23Qr<au`CR8NH{nW_%@+Qd;Z=-B$S(DE;W8;O<(vs@Zq75abLaeH-YLqqAmAW`j
zz}Y41MJcbON$+1qbAteDnyPAaG$&JlS+!+)b@H;Py0NaTMCdqBwNA!#rv%aGKqV`A
zMf1y>($h<-tJyH+RTQCESGugcw5hH^fmq*A*8~A$-B?@SP*vMhQB+#Fku)_M(<@6F
zl%<*ZMN9MZvKGxR%FNF9J)qNlrp8li&1l{5q^YEIBe_vjQBqZ{2NH9px@dDrL!-}z
z6Mbg7d}hYQ`b>?D@tvCNJJoll@4ZS^YRYRG%Y8py;$mpf8mrdV`s_~Jf8dVq;fRaF
z9A8wDS6))>yC?C!QxlZxy7g!xt8q~pb-8+dNli^jY`s$6Qt!JJ-)CZ+@5hbr(<{Me
zPh)-eH1<C@65@PzAjxND?0-P@Jpl2(M>XE}0L1$qfOy|0XZ(NAE#JpA-uD?8@B0wM
z`+oGt`+oE%_&!<*zK>Rd@1vFA`)DQjK3WOBk5+>3qm|(MXeIbQS_!_7R-*5tmFW9u
zCHg*EiN23kqVJ=X==*3T`aW8TzK>R-@1vFI`)Ez~eYB?gK3da#AFb)WkJfbGM{Bz8
zqcz?4(VFi2XifKhw5Iz$T1mc-R+8_dmE`+qCHX#DNxqL(lJBFH<ojqP`94}nzK>Rt
z@1y1NeY9M@kCw~#(Q^4dS}xy5%jNrMxqKfjm+zzH@_n>izK>S2@1vFM`)J`lhOZyw
zlYJkpWZy?C+4s>(_I<RHeIKo4-$yGp#`iIZjq!aDVq<(Ch1eM1haon`_i>1g@qHj-
zV|*Wp*cji3A~xnf84KSJo!I~6CdhyCi8A&-7&qM9^YuuJjr|YC?LYXi9G4_+W0f@0
zO_O=ewT-x`fV+!LF-7clC8RuYV^z9c#~21dzrUJS*W6T9TTT|58_P4c)|S*%l`h>{
zv#ze1VYStDC1rZ8ilVB<Ma|W|-yNB*R9CI5iEgZmW~Dro2r1>r{Oqi`=|yqTamk8a
zjE_z*%KfFd=tSHmSMg^;qVIa@fBj|B=zki%hpLw^<az|?NduL?!80U0epZ}F?fb96
z6Be@)HW&gnimpG(q^GTEXcZnQTNJ0jLaVCqBt=uE2)TI}=}pRXQD6Vww*LI%NIJLE
zE56UI<Mje(U$g{8$0O@GzpInyI&R4HfFGNR)XYkJ=b$fZVnb)xcWaq|UATlt>mYKc
zgm04YC<#xK@E8d{AmJ_v7e%H`mGEH_-!0+d(InI6O1M+v7fSfm5?(Cf7fX1hgkLY=
z^%A~U!doO<|Bf@MTP1v@#NRF9bWG~oJ_#?;LF5M|yg<S`C7iy8tZ#=U{8u`N{D_1{
zNcd3+r|%N$TbG23N4iXPLc-VSWaK9$oaT_eos#f69Yo$O;d3RtN5V5D{H%m8lki>%
z7gL{W#o&=IDH7i%;S(j?A>nZn9xmbZ{c?SCO1OB`&QwtnK3OLtkCAY32r{ip!beN|
zR0-cN;cf}PPQr5~{MQm*DB=2dTS#3j;TK8#N(tXB;q?-}R>E5(yi~$lB|KlkcT4zq
z3EwB-<0Sl`gyRvtv2{u~eGgvW4okRr#LrYmB)nQDBR?wPcqD6VT@tRZ<)l0z;d=}=
z=D&p3NcbrUr{5;%TepPkOF8L!Bz%reMt)Yp_3w6(x>v&W-&(M0H?Tch&)vMp_BIKh
zCgBbVpDy9y5<W@7of1yp_t&>53GdWF<S`Q7B;hUzUo7FN624x--4Z@Z!gD2jk%SjY
z_*D{KEaBqR%CwadzE9%UOL(q?w@CPK32&9~T@t=q!iyz*pM<ZH@PiUQR>C_a{0a#_
zEa9sq{D_2SN%&C-pC#d465c4`CnUU5!cR(gm4u&?@MH<^mhcz}?~(9S2|p|0*%IC>
z;Zr1BxseOB{wGMdO~NxI+#%r;Bs^TguaR)4gbNp#HcG<VBz}y9&y#SMgqKNps)Ub`
zaJPhaNO-P<|5m~aC0rDlwphZil=zhrzC*(6CA?6=TO@pqgttogd<owz;gJ%)Pr`pA
z;RhxBG70aL@D&n%Si;*S{D_2)knp1tzCpsfBwTzIjcHFvc!R`0DdFW3eoDeCB)nU~
zr%HH_geOV(SqYyl;k^>RP{PAIx!{LPm2jton<P9+!p#yMBjK8ayCmEq;i(c%ZR?x6
zla~n7Ka&uW+ev#xA$$NU@m*mj&k4s@9gVHHlNSlU)Ib%bQo;vGc)f%NN_dNe2T6FV
zga=FbZV3;O@O=^<D&Yqu{1OT8l<+VKKP=(mqaf2Bk?<iB|EPq=O8AB1g$P`TKz{_j
z2psjTyX}n4y~pwxu7}*0bv2p#PP*HkupMKhzQp&66y;oB<on11M=Qi<*EQ!(zuVW>
zw~xuRfSx=3q9N15dG7RMhD;0Qxzl$WGA)$nPTyq67AEgEWV$Fhclu&Orp4^s=}m@A
z3*@=eC5B8N+|Hd|V#u^mo;#gp$h1J7I~{At^g-#|=`n^(3*x!cL555pbk3bt4Vf0k
zbEp6PquyWoz=Qr9GA)qkzai7Yi2fTgEr{s9A=5&L{_8TLy9U>^ful1N7GKDb7z<hc
zaE+(DX04}O;3emH$(dgAY%e*{OP=Z_M|#O4z2q=2d61WE@sfY=>d$#E`BN|1yZ>){
z>0kAdpYxLc?j=8>%kGZZkD=`DuslRra&J=@4uurllA=Sw-q#&Cx)r0P91}VC`x!s+
zSCmcs(cOO5{n%%7+>iB|-RhI>7k_LDg@M_^fUU2)f{mkIpN^N-**U1IG*2#Yx6PhG
zcfZ{2A2tni@0lG9yyu%A`ucjxF!`ReL_${ApsvwA)jz!jO5qR5XfKS8-FDY8c05A1
zki19Xa#}%Jeu2CF{8D%O%Q)7<-KF!j>^)Q8#hi0@gtlhwX-3`S?hY;2{aCj(tNjVL
z*?PQ{epx^V_MUxiWUyHRPx~BMdoq=Z%)rmTIv42Rr8i0JstEl213)>pY_k*jP7n`7
z>_&YEHR0Lq-N=K_Wdu6b*-%>PpyIZ0+HY{uenAxAh8Qa5xu|XoT7@(zj&>_G+RsJR
zN$UolOAB1_B<-G|UGKTT6=%)`24zIp-~;@KKDjNp!X3ElMR#DvpdE>d0-^TBbAfB0
zhW@3BjKJvkK*$c<^)3-s5g|LU_FXg&6Yf2^<j#t3uq#BR701YLRD6>vzI-ll^+{^<
zHSE@Yk5)q}(gLGiJr`I=zO8r>@ap3rJx-(_03W9UYy_o4Uzq2B?sbAr0$K{mF*h7y
z;bYt*=JHecJVqXov9(XT_oO;10;AxHW6%T8z`)aJ>3O8*P<{;g6J$QTO;NFIhp$?b
zwmOZb48Qsl^W!`|1Gx7%Xyn};16wiHZW?R1*~AB$R&2CHMvhffMC-LH^!aoj)i3QM
z6CK*4Nb^0a(*lns#rl<2;4VF$mY>!BC1%~LG{u&>J4Rq~y>F#c4W?cN%&}>=$Ldy3
zaTjQjHs;#-<;GmYyh8Y&qOo<G!`*w@#H0#GL~=1+Z8T9Z>Th9gxHXDxb_C6nR2!8u
zA{=}+oWfvRYUBBAIEC5p4C=hXbtg%eCv-G2ISBLG$5F_%(W%ggX)0Tv<p@!Hy=8J$
zKHj}$Z$6H#0(@rtAL2_Vxc>NpANfD?=>*u%&%zSb{+he}X#~Wfoc_Z%eSKJz`sYVh
zdoQ~LXn)<^elENH$(}2~i0xt*0et@8z~Dx*g^@kB({srTn;&V*)0WX1!w39_r7-4#
zk9X7hvgc5Q8>i3gJ$FXrvi#nNLY6-qQOxoK5tYbkvCYVC7hwCp-0h!dw+m`Juj+0W
zrrJf7wkKSCVA9s673u4&yCZSp1la4CJrh}WdwxVzc6)Y&GrK(_BHWFV!wt;Vqrv;^
zJ*3?Ot-IY~hPT;!hy=jlhRSw&Bi32zXb2pgs-)%rjTS|`zXk6&abJAKrQy9M?mb#$
z6aaVor4dm8V!N{1Kj<kz6IfS!hEh`<mOtY5%YzZnX19OjZvXp3Wae)q$9DBBBTD;6
zJ-d1l_^;4;cl+Z#k3m2fe_cR^KktL%$9tY5to=Bx(R(aTX!;zF(!tjg%E-L=GWTTe
zPji6#;wQ*MEWcP5j`jQ#ERPPc`3zG#^YBWTd_*m1JlUkC)`3gy;LNbc^2kzxFY9U!
zOj~`r7Z)!2{CSkR(R0n$<bOAsp!q^G;oG$KSF+l_Xnz_p>ICJ|uBPa~i5U^jz=_!r
zF@Y2FBT@q=uEw00ScZu(u{xqYaN@lYtzZUDd^lnsaw)19mDsNAj_ioD?v6_%jzD=l
zBSMQo?#8g%Xn-A;z0|a{V*$K4`8QJJbYwhjb9X$<*q@9u=6n0U+n#h}t50F^z;fAz
zg(4`mH#@njY4F8A{u%W)|I^+6jl1o*?UT3JdWOEQKx{JF-_36SXLkGRhz(7W{typ9
zh&xe{!g~?I;1OQFUSzeOLlh2!l-2%S`)mL2L-AEpZpShT#e*UgFZGAQ$*)J+pKf~$
zalom>Ve*Q@H+mfK!9b3~8@V)*UdnEN=O;n1AqdpJV?8^*@(2N6A21)jNo#*Lv;A>)
z8G@^ji_aVZ9g&EoNrCy0P4gk2=EG{54`nnT?u=*!iRQ&VmOmWPiQHpeMA4zvO|#-8
zRG1Z^G016F;F1HgEfdosJFO!R!8pN!k=3!|MFits#{XpaJ`o43^->(@GH~$4MjXz$
z+n$u>!ygvowMp&W?Qgi-Kh9`>ExY}NK1W*H_on6vm%X`l+}AJTV2#Dq5nYhDOW(xl
z?`<1=>RI<?Umsc_6f-EQe+CUi@F*Izm>uhxdEO&{bUYQbV@<vI%u#UuD-rwIaJdmi
zir8hJHdVsSK?|{5#lHDTnz+tvf4r^B;Wp<-I6(@fXrYHV(e_N{-9izkxtDM?asubo
zr?;7!rnVhZlizL{kIsh9W>e~~w5o=?JLcNh0DbZxwe$XD`XqxBg-1+i;iCTH%IC3F
zYmW6qq7~!%Vrlyq|Gf!>1N>Ajrc5af&Y%K#gAv0O&aCC`$3C*+8rp4k?YZJvEOS`>
z#(26Jc-fQiQ4Ak*^%%vr3m}gUo%vZ`_2{Ct1(-hiM8%T&CYnX?nri9f<HWfqG`1@i
z>wZ&2I0f{Rc#q>1^RzIytxLVEuX%7q;IxyQ9qo^U_7q!&&$$C9J)L~4X}FlW&){^<
zrtZ`IS9F*41zwhf(^^NmO7*D9iPPf&^gMnJ6OUi-@x|(YzfMswXl7@3WJH|xSh5j^
z89B6^;Q9;!P%i6hT0?|X3LUOeQe)cs)TRXwl1xV{Svg9lnoq`f$HOrI@i-S)gn1YB
zSn~0vFwy(B_1?cl=WQA|Y1Qsuu%7w%PmVXSSpUE7fA)X1|7-ql`+xQ){ZGrs)aavC
z2M1c)nN$qceHbjhq+zry({QoFCOEF1oS2dL5E8UHB8Cp25l~?ux-mKN!EEh&(3C|u
z8^e(2M>u*O`J63+D1ro!*~duBN4Qi&g>#KzKrzL!usC{t&8aB1{}Uf{PPj|)Q9`>P
zX&(ESYStn+)4j)CjL$B#Y7NB2yJOaiG!vN4h5e3*@Sgj1TP|E@YHZ@uN{o*XXv)>|
zTS?s<3=`G#31W3MBd{_)Q^cSeeZmb(=H9;BIEdQm`S>#gFffW*#@bGiLqyO)$b$kn
zwO8~%yyp>+p};bTDknfh_aIkCM6v1!CO)yCZTf6ih?Dbb8r8ouH*Gx%uZ-0Km*%Of
z){r`FH92uQ5-MYTOj`yoKINbKOF!p>7=uq;XBUcLqcIz_0ato3jOKE10{~XuAWP(R
zA=9k=2Zp2vjMWk0#`yj4>CcVd1XTFR_`Srb{o{wh%kGGnm$k>5_1Fhi)+fh;JsHko
zw5d@;I0N(yL`SgD4hFWG0)d!Dp`ZA`20v_c^8rgT)vA<hy~f)LyCIS-(?fYZiURea
z(cY5@0=m~fNT2o(QWYvO8lR+5J>hyMqfq69F-Vxl%r0h)7^G*VL81@g7@3|`XIO;z
zAjRl|l(r^qErs{=4ydI|M1Fl#klitL?=V{F5^qM9mY<D}KOK<|pj_bYm>M%2@Asrn
zXg!Ni+r1}j@qAi3J_3C=)oM5OBr?Uw`whh<OkwUB!E2m%>owX$4Kr7qF=}iPijO~G
zZET#U*QgeX7rA03Q+$rb*9*l_t~hP9Q7jZUas|E9hd$GEB<2gn<y=t&Q(Un^C~COk
zUxs3-P~>yPhnaef#iZz%dQ|8Wxc+*>hjgJB&K1uZ{hTipCa#!iDCP*o>5o|e8VtpB
zp*YDE$%b=LLh%4sI1T3}3&jmw@wE{Ey~br+(cjMzqDD1W+-H~y6^f-?achPifI&hr
zgDcvMHt_ZZHYdh##Wus;?}l=PjVrQ@Hog*y&rh?SJB>Cz5sKHi;&a2)dqVLDS3GGb
z{vi~%az%eXo)wBcTydw-Mwd`Da7A0X-od{J#R{&-GupUKC{nqizaIyMVggsVj5e+i
ziU6+o+Ay_KD9-n=p1)><q*f^2;)?qWQ)NPNj4SRm++8jdf8>gEqm3+~*v}PlMjNw*
zqJ=Al8f{DyiXyIX8f}acig{cy#AsuPP)y~D&kR!zp$O%QM+{RYq4@42*7IlQ>S6yb
zoX?W?xZ=AsUGZ<Bc#1358KyoKiX&W6Vwn1;Q2drFdJV5$6N+72@wH*<DWNFmioY4A
z9v6yiuDH$c>RzFU=ZYtcHvS+K!?@xeqm52dbWFWN=)e6J3&1L)sXapRE>~P)G<CU9
zoZt#On&PW)lTiGbD-<-v6(vG(4OhGXQ(Un~D0Xti?YPe6iaA12#ud$msaT;{$Q5O9
zm)9646meX!%4j1@D28&yHlq!zP$*on#Au@zPtfUX^3jJZBsQas(?an)SG+z)Z~S$k
zxQ{DtH-h+gp*X}9*Be1RA{1?0u@_$PQ@~9^v4JZV8sl}9P%Pn!u|^x)g~G)ZcB73d
zp%~2-;YJ(Fg~G}e6VMOd_&lNb^aIxO&kR!uLh%Y$)EG5J3&mf#Vv6C_K%uyqE5;e7
zzQgwxD7x)j@vUL%6QQW*ie$so+d{FNE1ZU@=Y%4KEBuTOJ}eZGTrt>a<BviykSm@u
zOkF1wU%bzHey3roT`2y+6(K2lteS=5F|OEPm?{;D!(4INXncuKT*Vb#hN)Dc*vu7=
z&eq$A7K$}oagE{CP@%}+idT&`tU@uFEABPgI3LEt9?TVgHQM-4D86}*^_-3mJ{evY
zif*o0VJMCZ#c{6a?+3lzl!o>$t~ic1a0!PCo?C_DYOc5(rnuraLa~D@%Fz$5pcfUA
zsZy@Ell8;Y7lO;a80H9W*BMVFjOR096o)pK(&TrHe?C6_GU|)x^{1D!a;t)k-A{t8
zJE=m))M#q0BQcCM`gzY?@3K%78NmyMqURt-_ZzVeB=i!FE;LY$&}xp>7*YNP54(Go
zadgWp>Kn!79HFTkeQc(VenRLZj_xsh_!ps}9NlTy{gBWfyP3bgH_#fK)p~k3`i5cQ
z72>|c(Pxa-o+R{Pjy4<ULxkSW(bo-s|4isUj;cm$w-CCGqooFVEum!`ea=Al61s?^
zeKS~(*B}tBfKMj@mY(P0Cg~47OR2-{pJ(E3&{OV?uzVM8$9>gPLk8&nV$W|-T66kS
z?4^$r#(ek_TA+^(bMaLM@cA8m=0EQ2A4Da5#=-+<%cs2Gc#5E6a;!Q1+&g`J{FwlK
zCj=Ai7)_cB+l2^Rh`@yiT!_Gh2waH3g$P`Tz=a50h`@yiT!_H`?<0VBjo@9iarJmr
zZFy1u`#$0n3IExb+V<Bf-&|5%P*$Pfy~cShOX)?)MWVLIgx#Ww+M@b8ytTI+uf=U>
z$*aksx8!D5t%HCU{~D5T30WF@|443A1HBTZzrlu<B@L*P)tKJg&`@66l&<79l$TaD
zmM?9tX)J1Nt|@Ae5xGpS5H9mna|q6R@W={byxpQw-jdXXqN*~bw5|q$E1y?FFDPC{
zFLiCM$J?RHi+~o@liM(nThfqTw;6`AtE)Dqmo%4FZp~;cEx~)Jxoy1d#6yQS9+xy#
z)zx~SLYPnQbIoX=cQ^Ao9B8a8srNvHc6oW#I=oHAXpTHEFsY8fM6H6_s)~$t<#@?8
zZ-dDt)%BG!ptoOMSCdgM38FYJqpX7F5Vc6<SiLOr%$9n&8xVyey|Sut6Dr7}DCcK1
zmKnvG@)A8l`Tg{Kp`Q!yGgwq~C;AcH6+#4k(YmVIx|*EkjGAUI#6!%$<e|w*o#Iu5
zFqf*!S_TMZZe3MlMs-QeI=nTT*Cj+ZQ`b<_->pR?Nj)Q3X~a9Z>xoYXKtoAwSslmV
zJzncu)kN?7#h$(Dw?3o10rHme1|}-a_4Rs@0h_XFH<xfK6-A|fu-*VS37t}3S*0{o
zZQ(NsrDn_yWKB56t8g^Ku_``Ss(6zPRqHV)fD((|JKRuJN@GenoBSx$ig+7E6c|JR
z^|_(NXt0Gh*g_4q@CI9WgDs-L7T#bB%^B8Ui)gS#Z?L7mK~V%%G}yu#tgfm}ufamX
z^SL5V`^kHd#OvIDf(t-;13P=8JAb)54))Tb{+F>cvHz{@KY6LTuJ?G;ItKrzFQMn=
z^|!PCS6=R}zstS3cFowiX=!PxOY)lBY3V4@wxM<#_M4N&mE=!cTeu`5CoOB1t7LuC
zvMnp-mQ+=3N=+|pNm#!EZ_2NUjEtSwn41<=GA2D|dfv*yNgI}Bq|aNGxpG$3j+C0F
zyvCZ$Nd>#+FWpfyD|zLjhV`jqnzL)t8)i<aT+lQnttdUcWJ5&ymN7UcQ>Mi<CS@%h
zR~28oeMVjU%$oefhS=!oEp?HR+gFZlTrw?TZAtCo1uGj;7v_wO*q#_$u(WdHvgN5e
z;~O%v8m5ikoW5Yg${kCWZJIW%JS(n#Q|-9;nptB@CvGiiUKY2_H7++gXH(_I@j27R
z$8Afgjj3pyw61bu_3ZR**{S8TmM_UK%UQf+V|9GO%7V0p%?XKPm#!+AwW*~(E@4i_
z!kCK9B{?&f*B7TRUa(?gQFg&**R-geTZ*=nl)LldCT^HNcSUK=uE|ShZ>?Uky>eXb
zlGKeg%`+RKW^S9DH!&q=MRUfujKa!Qb7FUt7H21}&&$hSmQz>0c*&N$yxIB5d5gyE
znpv?edc%y!$b~60H#ZejZpfQcI@Z0mcH7F-xP<vF+osPbNp6aZj9j{K)6(tLGjbNC
z&5Mj|o}ZRqJ+U?^Y4P;MOLr`vGH+?VyDq;dcY0LX?6C!7bLVWTTwJ|r*OVEPl2^>n
zY)q(1PHA>!HZGnSGjGSd<|Q*{FP^n1vwYJ01<6}nvz9H5j$I!+Yhp!Gc}hfK`QrMV
znM<-GBWD(*&t5cVMoZDk<g&F<8S6I8Z_Z9H%yezcD9o*^Ei6e{T3)teTuN5{vPmVi
z4e4W(=a1hwYgca8_^mO^W0z-5Zb*tt$*PYlAKy}(x^Z6dw8pxW#hY^%ZOPwUQCkz=
z&>TN!?bz&@a|*UEij17Gcw741ow4Kcs`A|%CN{-wFD<Mtir>(bRKBora$$6JdPLFq
zar29J?Q*#`#zk$~>`IBwS(UsfHGSUV8H?9NMlK$^EOu@Bx=D4bs<vj;#Z0T7y?xh=
ziTUNxG0V0Uta7EyPKk|-tlBbj?6lbp#Wh965o0p9uiLg}OHA?jO$7^LS1!m$;HNh(
zs;i8sO5HGTdO>p8f|jhEd6QQ(Wi~IGwJoU{0dI<1oK`-jDJyOKmQC(iIj)Q;D;tY*
z(sMR1DUOWHTd`w$^|tMK<7##lZC|v!x}kRA#F8~D=I&ZBdE8>?r)|tEtE^m}w0&)9
zW$CWurL!}4%$$Wcr)^1^I5&68#x+$t*Uw8@uyT3SxR%Z7>$5gYOsUUuC)ID8RlYJ|
zcGR*-D@q$PBO|ws*;<(zmArZOy19{&@iS9r#21&R%-zvY5I;9@W7gPJFtaKxtu`h)
zX7;?wd6_HoC)VXxENY%K(Y?JSx@7i(mW}1p$GdZv@0hf#VrRnU<%nEzUBdK)xq0(*
zGTbE_He^@l)l8Y2STJYFq9{D2#KxB(X?4OCkM!Ry#P#tJJV>GK?401mlk2Xr%<WQ#
zn`hGh6pjRX`0ssvxhUI?_w}_PcOl=0-2GHv-(lp1$R9>tiTotx&-C>TG@%}HC-PIs
z6ObQ$wy&=m`3dB$gd;zIyaze`ALd@<4<old2Rq1J$Ui_{iM$v2ZsY^au!}qb`3dAM
z<VT<H>$?>B?ic#{jvzmU{3P<8lYM>k|A%^B?CW!AcrVz?csPUH^-5n~6Y_(v_Vpd4
z{I$Nmdntb%^^h07(buP0;K!TbBX{Fz^rgrvksn0fg8WY8;rQml3FL*yKR`|!9v87q
z-Ik}QEe>_~KtCJBDgx*Tq}(U^`idbd0gibA;R^!&x7b>hIm2d5ijNpe6{&s((r&~&
zh5)e<z8a|$IE{tAIZ$XqI*2g|6d+-{kq$^09#1KUkoEy%|I21xfTP1SHz0hkIW53>
zxt0+S)n=I=kZOKg4Tu7g77#u+z%e(#hB{>L6jDCM7eB&c8)0XVT7ls^*=$RoJ$yR^
z>@6y@HZrM=3~IwoZDa?;nEv9~1le+-UdGdXeIksk&1_()UKqu07BKq76>U_v37E?Z
z+YKzitL`CSF~D?xsqPVA*<Kj^Z`4X)_@V&YDE6m-H34%HvM|7rkH}E{v%q%%A4NDe
ze7{Z6Xs!Z#mjJ}zDvI-!=D7jR4sC8g)Lu(QK+NUVi~v`gUq(Q^sbWBYD=i=<Eg%XE
zFnQeF{_H(=xv(__$Is(r3!1&Ag#qE0o973dw3%)m5D=ch1}V*-=GG2yPlB6E+#KS<
zv@OSfF?@9bIt0CfV~qVzE%>$9loxPBHN87P*a(NZ3!F3He2KEMh}7M%k>*ch`xe>9
zF_(xwQ@{3_7LYF(+cqsL;G}8}PzSJ?lO_y;7KIc_$0UwX_P@U}100u|(gVWVs1K^?
zEcyT_joNU6n+op3IDXmxse+=-loimWn)brSOyZD_W#IJU*glUk#sWTmz<eZKD|Bz+
zc=y)DKz9hbu9y1y*#Guo{$p-lZcZov=LHm-uTgdH=dpR8iaNB8WMjQx|Hlh;t~BAW
z?Vz#Tt6{`0r?G6aQtizu8$%pBU>L`a3pPTrrX0mu!~VAyY;>4%>6ln(922l%9x%Z3
zpg=vEU%9CJFdq7jM42|i%7C2!rXK^;?-pQB0Av5xlf~G)TSYHOcK|v!9#^wtm+7il
z|44TPI{HpPI?75mjZFj7k$7Rd5P=I3xDbI05x5Y63lX>wfeR7%e;)z;ecAf^vGw<1
z``R*Hn$t$__@+&l&G`DQEYmx_X&cBhTzK;Wy=sZJ&~-e=_cht3zt?-Xkm++FZTKw<
z+w}K))8`P{bp81s`|3zOi_Z?U>7VaTi1w`<R*v}bg8mM@xkB$R^mMIC8!pHB*2yzr
z*7*Od^jVEIVNTIM!}a0|mok1Nn#;PMZV?Finwl7Rl>+lWSMd9Uthb9_XEObOsoWl3
z6v1R%R$tgIMBqXME=1se6@i%F{d{%$tpRPRsVq}+MP4lOdXcw^e4ofWMSeu&T_Qgz
z@@|oz6}gh8u+1TI`i%>1F(OYDd9KKdMP4uRu1-c*=BKC6a7GoZYp!i-cE-j=Cq&0g
zjcsOP?5?=D=$M3w0#>G=hKbeUi*pydH-Vqn>i5{VMPq+&Hl0>fMo#$^&Ko0V3Ej`t
z47F(w;-oc4v*4Q!ihU8N<`;*!k%ij)uEgmr)MgouvkkJrl&Q8oI5mWt{H_>-O!NB<
zQB-RO11-N9jUll4?M3ytOUQ<5Jpj#M0${HlWP2gL<F7>xzXMR(Je2Q6VcZkgQHrOQ
zacA#JoVda*$~P!awBn?@k<f{m!MvnRd>&XUAlt;-V3SG?#a;w!)`ajIA%$w%L{J7t
zqnv&%93L1HOyvwejZFEC3!po!359gNP6%?+g@OGFoNWEZO&bYCsKe|+CHrDDXufLL
zVz5FT+Pz@={RZVYx*eBUiIca>=I3}1LU5yEv)QT3C@xZLcE3avZNYVlZ2&6{46apd
z{;cQ-uFY0#16euL+JX_X4H`g3!mU+fAOzSxg5b2$Pf=`vOo$4uLjw+0j#2!m9ah~=
zKG=fHm8dYe(=z0@5Um+bU_+T+vEGF!*oHAdvpPvLoC!8-F#=~B!GwX<S-3E;jT}HB
za#-hLy4glCAw1X}qu55XveVi}9yys1Wpxs3467ew%_PUiG9kg5O~N=PxU5^L%}6H9
z3a%~Rq}awYA=P>rW`=D76EdtRxRYm#VuIT`AF;Phv{7%ewVjyBlUUDPnsy%6Ci@?T
zB309xVPFaq+?uuxw5hf~L6fUF@CA-(%*{fS<CGXyF4nZExDJYCjZ|vdwWuD)glbLe
z!q~<$p<dHk$6{76p-Izv5Wz$yv}jrbrr305tyR;kxC@!Y+TX2dPvTO{#e{vDwgAUL
zG7}DJS}xj|!GumtD?~qMvfdolw7+2pW--%8G>!h{%xq@*sHV}4-4v$j(zHi03+FK5
z2~FFLu%|NNgr;qTwKOI?t7*$&Z7vf|A~q1xned9HRiSekOgM!=AqJUDcuUjXgk~NS
zy5Tj1`Aqmg(>9?|Hxqg^tquLjV!|2t2HFDV=vhspUt}$0nqEz#Um<5RjbhOPFbi^+
z#%9s}gmGKMgn<?<1e(Q6a9Ff=u#Dw0A=IMT;PnzVh~XCPebmWg)|?hc={#jAQ$$%D
zmGhK*hR0a6T?l&tQ@AWzIl{4wX;Lj(9&9aV8n;E;hhw{tX>u)EHKxi6rYW>&x1*Jn
zOjAs)#3`$orjnY8Q&uxgy+uogwKYuBV$mj`(`%Wg)uLUD-V`y-Zi{9?Y>Js?pG7N1
zqb00scuY2KuYyN)n4VK;iTlW`_|d@@s+!hgO4(EKLBY1p+^b@GcbUpRMa|%K`HHQA
z*;1^7Nxz=W8=Lhdgv?gS1cw#Fq}ZyM5N@4|v9)brbIobp3dd|4nI_8GLP9ljCdL{?
zv>ImGWleHIsAWQ`m41J1t7C%Onv2fZ>RFv!YdR*oZ4=WJS|`x4-@w`_wpL?)+Zvfr
zX<Y$-Y)wq4w~nFpqM6lcvA$2jW~OPiZpVbOZDGQ0>uc0b3lsJQS5+ytt*m^|`U!P(
z8xuOM4`HNj+nI3Kx{IppV8Rh=0fk^E6OLN_F*decOy~-xZ?V`eV!{dQQF5TwPK)cw
z;D%Dgb}@_1DQg-newQ#-x0T+$ZEIti9_u$0j@>K*XRZIFsI@cUOY2l};8#rOwcbqY
z)gC5%uhDA@_A+CN>Zoc_I+!h+>L}Z+T*+{Us_lj^S1}=6)k-ld_A!=I)n0|+{Y;2b
z9r#+^uUR=p)yAOt->^1Zuv?}aVC7k=M!)mEnhB|@17E*Vu3<t3v09YhGQq8C@mSOj
zva!fjwKhz;Yx!WR+IVPw$An^4n+9vwF`-h`_5|WF5fkcF$JPzX^{m{YI<{0QH!{Pm
zs+NoJbu!^n)lpj0q};@W-NedLZf3$>5@M8Fn6OXPp2XpFD-#Z&6|{dF6Ar3c1lqry
z35Qf|8BG7437x9;J50L6Ot@8bls7aee_+C4Rf|K+{>X$oRqa{K^gCE+kEq%T2zN5!
zsH*j%(LXU-m#S6Z5W9;BPpDcU8vQd9PN>=ouy%wA&#D?dNdF5na1x!Hr`*fRr&PA^
z+|Pt=m8~^Lnb4zZo8a_=EKp}vEfLf4Atv;y+ErMe{>o^INgIU&>S5N5&7_^f@$d)}
z2AZ@i82B3#944(D4m`^0gqzqI=P@QYP3(;GI1{2w?2L1a2{9&i#(9DXE|Yd2=HruW
z5~Z57^|--soC$7|_8{6h!Gv6sM!!yfiV1}#?OTN4X(kk#v}d7th6$A>?Ew1o924qI
zS`ReOGoi(#Rbd@@feEc9?Nh}4B=camN!ta@i%hf6r2P{M#7j(b(4>vUIqzi_olcXs
z6^_2bx_#KB{fIjcuQK6IleQcy%<F84Jz~-#aI~Ca29BB(dmqNs_J+9&;AEHm&0pb$
zjkOMq+uj;ZWhJ<-Ua`GB<SLYH)=nb-lSkHiHwMG@4iiGH1L2~rn+f68FNpRo6GmA7
zO-sOgOmJFnpf%`yCPY{xXhr;h2~pNYoKtKcGGU4}4M&&lUrdOxen`v4M@&cvu5D0k
zJ*@1qUPBI?X2LA%qqHD?%!E{HHVz2eCrrq&&L-E+Fu`rzfmv(&ltnDpI)awE&zMkX
z4a1PxK4(!bww97`mer}WzCla!znP}qdMo*Pj%iw~r^)meOxR(iRmk=w6I!jbKH1JQ
z;Zo}d6wj}iu-m$tu2;Wi!d~lC^17D^`>Z2K_=X7wtQpjWZwH)FvFKQTM+UxQ!bxo?
zgzx<h(12i3g0&x5l~XqDeY}|bM<%>w(;5-?KIU|{Ei4BkqNr@pdu(C!Yiw0z^Wm&5
z>@!Fvl`(s5VRh(*S!FR%>|vEKqN&WJ%^r3!tXQ}-&>r>-BrBI3_OOW<Og}D#+QT+b
z{J70<d)SqTnVm~cd)P^Ya{y;X*~2bJd;VOSVh?M;Ix>)Z7GqcZ=rrzkFBszz;s0&u
zGCTQyp(_48bZQ%<n!f>MUK8FdWE-mnKMIE56W9%;YlDwaFi}OmBgn*dbQOXBN=TzW
zP7VH&Dt)6@3dYp)s-*aL&<u!BgGWQ@7l9qKPd~QyqB8r6|4tg?NY3})>x3{tb*!e!
zCD;wyWK`sy?4aXiqUzX13_7z6yV+nwdw6Xkh2l35v%;T#e{Y+lI_?GJ_aJt|UdHa{
z`+7A^`zdO0H*r72j&<N0gU=3i|6)YnHdPI@;;iN8kKI6u-e?qf+&C+T`V*}NP9avT
z&dMhi3pVwM^O?viZK3HrO${!gN)=Qo8m2EbDiMoUCrt6951IDw&_nx#;H#<5@35QN
zO6a?oTlR^AiIM0`_z1Xl3630lVi5b1M{q@Rtzw@Z)DDHs`jiuzq#*LdVJ(0YcGqaC
z60Z2&h|Hdj32sj|-+}VrGX#7`6=t&X;2!`ahNsNNI>VcoBR7#MH<8AhQ0yg`D)zbN
z0Mr>w742o13-*kmR35yDfHrD!{!sepG5r!M-;ULJfP2V7466M}bZ)=~);at2=+%G+
zh8_UX?^<Gy!$qv$Ehr7U3p>+p>=ZvbD*TS3IF&&A<2aiAUkQ8)WxwaKV?+263Ou?L
zA`^>~-3`b5Uk(0(n7zcL5%I$k!Y8qPJWfvjuLp+!@e9X}>8I#=L5d~PsbCk8X6vM7
z9wfp1HI{z=Hv`uZb3JxbX-w=cEF1oB1-B5o6Fb((eiTe>MvzO2eFydF?ZE4abTg5t
z>iu3M7Y5h=pTUn4={S+7Znp<Xc=9-c?*D%98DjohXAZ!)@ixi0VEzRo;QvA3VCejY
zW5*g!_&MgaPB1?VP9-MYWnrzC{*);Uzew~?95*2B_FQu3jAJ`hZN+Zbi&TxSvQ6wz
zVml^!Vf|V!N<WWEihUM#{+~JCrWW3#7CxmG_$<IOs`!8I_=a>pVmFXx)u0eVe;L+p
z|Fc0uf%uKWj_Ie8o=rddgJYoocR(`fQ%FzY%jwtu8*|qGT+j;AuO&SV%f+N;PP1V5
zmq=9XtEtWxK^;`**Ls~hs1BRHMjgd|2d)kL&j&t0%!jcXcAQw;$|>^f>!6oOcM3b!
z@>itq4;(Ya>5~yE|8Ij$m}`E1*fIJx6in=3qveJT9zTwZiX5X({jW=k|8}ed{$@4!
z3UY$pAi|uWh5rBG1as%V;sl?d{<ouX|3K9dgtq<YOgSuuykJeRP_Up-vx@%)I`V?l
z;Mv5Tiyia2nE0MlJ}~|#=$H>t!|uUM^N*tu3{%-L<{wUn!eBM<3bM49EG43WKa!>X
z69hrJmo&uTYVZSK_&tmrYxy6<XJhOc5ygKktyV+T;E#xZM(10w5PI<y|7CRO596l&
z={WTtE`*y&GeQlfe_8Aoiydn;|EEU!CxqfpFK+Z7#as3tNC6lvgg0rjJB95%6o4^8
z^FB_;{<x~4-ud6-#12<0On4BdbAMcqFd?0s!^H*@w&4)=#|Pi1$hqgS8y1E>o9HBA
zq6;b0U#P^cjM$Y9yM|dD0E}If{GLkeGJ;(juv0&s*%iwI_`EHU2Gwr1-3o-j?-2j(
zb~-ZcW<>|-dSr(Hdv6iUCRE?)PshF8d{zf<0JzJaowm$9I+$%zl#4=W>9m`>b?`MX
zTSI7Jw3|=qpn|pQ;!s-4?B<g?I0@h-L+coLLI?8!whf}wlHJ^;gS!A;8cIu<-F#FB
z4+FexAgy0^^AR0<)~qO(52B^TZa%DoUjw{C*zDB7u^PUrE5L&~m;<mqgpOjnd7lo_
z7c74jKnIoGyjutJ%=n)e{<KEf&8<54EdsJvfGs*W9?TAZTAu9YdL6t4Hm@8oh*huD
z!RG;9CBR}G9Eect^QX%KySY#Y2VkK052Q1^-JGj~?_e~4EkL&pzKe>#8A#U(c5|u@
z-VgAA09`sr|McN%0mkTH0Iqkh5nz-KhM@Xy1?belL{vX0z;GSx0eGzd9Xj|3z~2eb
zrh~r*c%1+h9ef_~JT!<dAMEDdo5lRR8Q}FnHEcM}>fkjvG;a`Kj}Cr_^Tdq;?AE~t
z0d@+Tr*yCb;7tRoSwkmvums@E0z9FE^bP1+MD;Emd=l%-t%K-V2=h}1`{3_wf_X#-
zUj}%)01xZn?*aZ^RPWTm?Envp)(`6763n7Mi0b=va6G_23UId${u@Wf9RYL!g!!q1
zG+plu-o`@LqJz%>{F4Cdb+8lQT>`As!DfJe7B-7@Z~?$0gXroF^HT>$1H4;+xjOhI
zocW8W?$*I!0PhK)OFGO?9sC^D?iEaz4sOH{-zUHr9Xtx={i1r54t4;1K(y}ELAshf
zDyoO;U=qLw1?bSh!8nc{3Z+vS)=wS$1dH)s1*qs?7r=)F*n1NO44dU@fRBXInGNfw
z4sHPWHv#tO;2eODhS50<>!%Ke0_+mtDIGkGtBuD5cv1)d2JmqKp3uQafX4*brGwx4
z<B_8PkLutH0RJ8u%vwL9gSP^FQh<kba4W#$0_@bmMF39(jb_yk>fktlPYH0J4))>{
z__P3b>)>+$pAlfI4j#p6^w}^vKikbMI+zINa{{c_!2tlD4=G{HN*&yd^XCf!EY`u7
zad+)xP%UE?>fm93FA6YM2X_K|Nq}x0%mesx=yp~;RR<>kd_{mR9sC}b->-)5X3Q8J
zd=cPl0*unZG=Q%Q(5Zui0iGIiHLD)3gP-7%>kR=qbnqmm);|Pj)4|&Sz8Q8itFGwa
zJ*fVc0DC*x^i(WWV7@KDvpU!c=06AjnN{!6!4xpx5n#6tE(Wt(fTwiu8=Se`9q|Y|
zNu3m6#38Ip1D^GFsW`&1K=Mx^5z|q$7Yrcuc|zHaB*m`?^&@tWkz7k6fFBIK$qD8_
z+6|_i6FYno#wlYkas0ASXRseYKRcCzsI-zw!%><aZGy#WYZ}V@QwjTMLN^_hq}2cm
zv^-K#i%Qa;ar}%@YY3oEO{RL<vCmJD)bxN6CM!vo5`ldzi>#yj0K>Hb1a}flJz({e
z$P!o^kgu^S59l_Ph*e|ZQH`Bkcu6<Jt-0uD(nZmVqWp^#R8S&2a99l5tJ+EW!GI&a
zz;qutn?;HqGjg&Ezd2Vbi8;Z*BD-;|I%tK-MsSJ&htPX?drj7pC4jM|^Gs0>{R5mq
z^*HtSnnu(3phN8Ig0`ZX&C0hX8+A1T7`mYh?Ge=vqLQi>m^Ld)&{=_XIwAB5fo_rv
zwXH)PY)|&HCxr+s&B-jNv5FFsiq=0iS@(&$@l*$!dgw|t%6$9SG=hAK13Cq1<QR;@
zrzYnO(Bl&Kcf>~H*9!x;05;zvgqKmT6aaTuRrA#tRbA3wOI7m-DO6(=^CO}TO<{Ui
zF4a*+9TQ3yc`=TrAvMpw*`X?f$<>e+T*|*<vYsts{wUX~JXa58jfY3U#axVZ;sv0x
zb-Q_VvtW{EsctSP=BQ#`q8gFoB!BGYB)LYnXizbyNHvyw)^N);ob$M4#hfqIX!oqK
zRIbsjPZ{$Dsm5PDYpj)P<Qi?XN;N+9tbunuu#JsLcP1AU^EFb9(U^Y5NZ8FgJZj9>
zYuqW-SmIfuL$2X8qWid1qs_C%&2kO>`$dZRlvLwB&l;)QxFf6|CyiKrCe`@Jv&JKG
z8!<*dG%?L0oj8IbJ$n9kxkjo{W0+K9p=XU3<Qlm~jTouMPR|<ukZTkhHRee*{_I)f
z1Gz@MQDe1K<6X}hXXP67Zwc6DZjfq(;DC|F{2RH(5xvHwcBux~c5|%3%eR{~tWLD?
z!(AVCB_tV(#|x%G`ay&9JKNMl&1iJgX>lwKq{BR9GA=})G7Zuv4KB0=*EIpFJ$>)z
zDbobHowpeHLZksrKoaOS$9hQQ-+(kAa|pQ{QY#@f)qv(vwOb(G$e?CGiwSxV;(Y`y
zNJ|H{ifX<9`5B}Di<A5*C!T4)GATv$>f@;_T;^NU0<3VzmY~zfRw&Alv14$RXkxc)
z*zZ<h21eMX9y(3kC@N|KV*;HqK16yKX~0$>tQ&ohzCjxF2u|_Gct=K|`EaB`kJ$)Y
z*b4n@$Vo^x3d-M2?7`J)$jgvM3>AKMpbZQZcd*2Ai}|*oXznolCWrG8a~gn0F+O6L
zapq&2dT2`&93%6On#NHCwjwnnA?A;oVkl1gA?+a~4aiJF=$F5LL>d%7lB~`n>^aC!
zAx)zH2tixFTb|*|!*_)?jE2j9G!3Hjpn3uxjK=$Clsio9>Dy>JN#4m%l3B^oSk&(_
zv5SMz>zXT+Bc=hYLUm)4azDp$=Gcn{-+&$V%xs*dgxrJE-z_GqKCLG+m(bXuoQW9m
zd(i(|Oks&Y)q-%0)GelQG*W6wIE~varf4Ru3&_J6{ubV8HQ%`s{k+v=rE0b&<#rQ$
zbVxya2T2K8ig|sV$*LcG{cXOgw|Si@q`%GUOk<=ruQN>%ZPo;knPdv8nm;JWL{bht
z45J?@w;(Y;4)I{Yj~loj*FbY6k`i*S8Fvy)R(;uk9k!{59Fs)f51IyZKLQ?P+(_P-
zG9(Wdum|{cj=oOyAIh8b>IaPKYjM~gFpXu#@(107tF;5BY1EI&usZ>X#q)q^7Vn!f
z<VhRu#_$Gm#YE|Epj&TXpJ|Bjj~du#V$Vbo3~FGXDTa)t!}uJ%fqkagynz4;*FMt%
z)}JNHkQDT1FK;eYG}j-FnHV8974~}cXRqEL=KEgWAGL&;-fNnro4ygBpmuZ9F+VYV
zt8RL?DOBpoZj+OTDu8;j+ccGV%_rS%(+sH(yG`>&Lm~7pGg3@e{mlOsA5Bc<Td+V4
zqN$u>8cJqYAsow**xHlA*Pcy~YmnGHO5v-|9>|vwmdq??Q$v4*d@B;0TXRfyvWK~q
z#^=^?XpSKTZbo=BOoPbLbCAyxM*d`){0XD~5IP<y@b?JLJdW*yd?ivc{f=fn$870w
zikq8$8LH=~3VnFV;@H=a&m-BE#VHFoWQLD@1f|C-3pq3rP#98R3c8Wav89lgAla71
zD>)o0g1CxkZ)2@p#IY@q8;Q0sO<Bw#3c{@frDrR-9C{4mBLpo`mT=@BkX|FCyrM$M
z<JcFFKPPNqRjsm=W8<y3wnqvq8-sU0aBL3bnS@nfnJC~`A>@3d8Q0IFDB^%ujRl+*
z^a32sEde+On5_E6W&f%|=O;FOlk}C0P0=J%1e-a0xpMJkLJ64;Po|oz`qg0n=K)%$
z1O9*zPBjgqwTu=-tbtSc{1NkhDxWzx1aHHTPc^YQvk9KpAPuUA=Tl8p8<AZBxebZU
zqiEiEAkEPjKBumO?ph?wtyms|+acYGq=ei6Cq7oK`bBXsEvnemLrY@N2)X#NI+$F1
z81$n^Y@&axvZeb4$j>0LiT<(5&RqY3+)db5SjInA-4qv-AHGroj5&CQe;b22_!%#c
zfIb`v4t~zR8!{Qv1VS)r&#CMQ$TEnzNNm!6!D%g!n}~)<`z41CK-@<VChd6+-39S>
zBsOWk;@EqT*%t;eX}{*skAS`<8co_>j?oQ{U?iHf-*9LW#PI}S(tgXKxe!wbTB3Z%
zkrj}Z6GD^rdybVtE+LF2?GGHQfxHokP1+whwgqxCVKiy`IMxRFBBU9y8DfC+NsI6L
zzJl|v`o0=WPdfA~cD6#Wfeg6=^}5vn{p!D8b3KMQfWE`ottL=iO#P6rFig*>)>Jod
z=3Sm)KG0=kctZ=6BiYlQ{Yp5#y7jy|fgbyq_a6-&P@%k_hST$(ThQQdkp@(ktXEE|
z!{~7k{nJvquQcd1g8QO6oSqJa;4I{iG@z`xL3v3XMh}Z-Lrz2*L|-R;Ssg>qj@Cmi
zB3@0o@`^f?o+Ryu*iN)9SoB_1htq?l2O!@=SR}4=UQ<WWbEfAZKTX*8h~?|*czW#g
zA>?-no8rXXD|H+_h58<HFJWz1FyBx|(nF~bm<~i5c?jpjH`USf&rs}U$E`dh3!irL
z!u$B-&o2eVHCMhe;(MWc04=cV`JwfkjB7-EViy9M=tQ_rsMapJRK$i$JGQBZ=)?J_
zw*|~Tsb@EXtFBS4tyrOGWAEoct{!rxVF>;rU}|#*YNK7#PVCX(biG0IGZN`0LE6Dd
z3k}i-66r}#(prP`gGBnolZ3lLd@~PT#tY<=1Ir)zB$~U-AWf4<iJqk27$mnuS}91P
zcfU7CB@*c(PfHIPq-_%ERzcd#L-3?Q`i(?-%9HegLAp~SedbB}+8{kEk%lF(5a1c3
zP_xk)xJeJ!hZ1)V=j!2d8l)d3Qkf@dwm}-Qyg%^mo}@(vDMljQ<w;s=kg_Dw%Yr1r
zRb`OYNu*v+(q#r|yF?n9$R~rabe%yuAd%*Jk`5cBKS`wZo}>p2(vuQtuP5mlgY=F>
zy4#cVx<TrdNUwR4J~2o^h5Zr#&Xd$<kS0ka=X8${3~JD4LApfB^dyZmNUJ5%22WD7
zL28yrS9y|>4bqho>0VFLJcIOmiS&ji2`|s(o4HFO{pd+rW02mENaK<`LQrLp&PgP<
zCux&G@?X&(@fuH3t3is8NWb<Z{mLNCmPikHlCCvK%OuiUo}}9hQk_InT^_wVVvu%A
zr174lM-0+U5^13)>4ZUgNFvpHl1>_=mnG8Go}~8-(iw^LkSFO&gJfCRAMtlQNqq)s
zltj{!Jwg!Fs85w7i8RrZ<TOaR5^0eqDb65ONu)+k(j0?yi9|Z+Ny;@y*GZ&DJV~n!
z(!CPtJx@}tL3&Oi`OWYM!B&IxFNrk8lXQhavaafncp)d@#<5sW*xBQ^26wc?-RWuY
zW`pFCNH=?u?l(wFB+_v~5~uA)4AKUPbjFi(%pkQ%q@bBR62j6`2I-JQn(j$@*&y8~
zkyd$<{%Me&mq@Liq|*lJBZ+jIC+SOr#NJ+r*gxRJ1*9kGM}stUb^qvm?n$yYag}0@
zlSrYnJVFp+kQPd$WKYsCgH$Gw)(Vn1fsQdqJ0((^Cn?e(T_ceWdy=LZq(4if=R8T%
z4blmTbk37B(;&Sok%r9nI3DI2q;DnCEKgFFK?+&ZAMp}T(h`F-MIv49Nh&l*nG)#^
zPtt0Gv{oXW^dzk_NLwV*c~8;?gS1Z~4NLI|0dA`D&HM+6G{=*)%OE{2k;*(tR~V!>
zB~rU5=_-Ttr9`^RlXQ(i8nm`Q;xBuWIt|iziPY;!`n^G#BaueV@d&}42B}aY&GjVR
zV~{pUr1hSphYZrMB+_0_(&Gl{7KwDXC+R7J^sq#F&6D(;L3&jpedkGf*&ux;k({X>
zA$Zdu`4#m?Jkyi(fkARgqz#^=&ka(tM7qk8)N7FPB+|W}B&Av5n|Y%|dP9)J$1gvF
zbg4x8(UTOYlSW)Gk))5PLv(^N<UlZfSFEz<{wJ66_Sx$!QN&aJxwsENKVr;N+3o5n
z_<%kh36D<l_)V#WkTMCuZ!q)Jc)CGW3$c=*S|BOWK=weQ_X!NBX`+g5!ft|m9g?lK
zOj*iryF3o@Zv<_`e_zPw*lUnqLJFLZ`?3Wb`vmewge5uYK@tCXlRs`N02|#@Ql=Dg
zEO1OHe%j2-q0l>!%<=hX92<QfA_?8Y{m7vGskhsrZ;i$c6WY$s8LI@Xm1yPzLioXn
zTO+)VWxpGB6!X)PYMZC(6UoHYo~rNy))I5#ZmzQKKYS6hSRi!&LSxVB)yP3*RFhD>
z1eG#m7Vc2Y=OgFH<M!CgOHou0{TX+z*vQT2BR2_sk3ho6&F3SR1u2~njNE)aa@7zk
z2&x6bMy?&wWrWbk&F3R`BjjrdTUS*FET2ZR3-ZHAG@5Q6@mC<8B#1^ci(@^IKR{xm
zxqxHeL+&MvMsp#@g7EBM5K<uh7&@C{lOT^pVk5kmW9CNDaT-kvy39tCM#ghAZ;)u_
zkA*;^IhEJ3>@Q@ow~HfyR9ReQ-G7vi%RHf@rZ#Y$KdZJs7j;J|=0=Hdxn5Ze>Jg!%
zLA^<Dlg*$}&j{5^P$@%#FsLpus7nA~JI^x=>XJ-0HZFD8MqoJzMi$cWk$Xqu;e+b`
z4M6k$le=s+waQE$pi=0IkZesg^~x-cH9>Ab8h@>5V6kW*9p-IS${ckFnWvYS&`T+l
zAd3jtp=?M){)Ib`6NQJlyk@_LUvLj6au4rD^*<qT4=2eUwkT7$H^-qrhQvIa%CS?B
zUqc#yt!O~^@I&rl4EONgpwQPI;i2dW_mFOq2XPONEaD!r?_r^+9{OP(rWW0<4^r8a
z$}rq64MM`44^k&k$EQJ>fHd4zh1ndevY*Z_1+W;2#d@&RlLq{snQ-oQCG_i%Z1u`e
zj_iSS1tHDKFpkjO+*^<WyKu*KIKLr(59B`~MbAd-C*U6jVT2k+zPtwbX(VN6tBAqp
z&|{YS8xx$wb=LhScF_zD3J?+Fk=Tc^(s(2~b6^FVzshs<(4P6M*P1$<S_%z7yGVSj
zEP4;ScG(&gKkoY?sOhe8bR76QQ5$~RxSziP+<_F7xrm#mua+r6cL<4jGp!wj^U=VI
zvpFfz@Z~Ao!X5L;2<XG;5ystM-hni>zPd~~&z|<Y1o3HtYLxR+6H)(L2<MQ-97j{<
z69|gLedLiyV{4W3DFiHpkcl)rh7OUh_yN+jYXS}0_w2k;32hnD_;-X)VkRkz;n9RD
zrI!uE#b9hjQlckd*TGdk22c}60pE`_qE-a>HW;E)E&Hs%zBOiTRN^DX;p{SWzTo^7
z7QmtBU-(#wtPtO0U^VhgjnEY0Mx+vSwP5q;O+#GHM5D#xh1g;Hl4qFHtyu`enPBLL
z(nNm}?sOrsfS#E|0sRo-J4lGknM871bKuv-0C;3R)t8R(uL$Gf+@&moX@uc(7KTVv
z8i}L~O%l;+fZFH?|0YIIzR+>E|BBY%C%1DKaDUkg#!*xcEds)Pe19<cSPaYbmCUjA
z%KPkb{Y4PABC!#Ee+s$X3GpCtYL)k=6YwO2E+pQM4~%|1C0bj~8&yU@MnC?=`awUH
zcn3)t8YjA&3boM>8jyT^!H*5X62V}-$$1A0hbcp)mdzKV&dj2sMkV74h$z)68M9DY
zv=kE(q*4sUaAg*K{UioqqU~{>q3;6_q^NeNoI#j7NQ<wAkb4t$13~&8tw!Apfxdno
zi?rk}2<H7qwnROI!g3%vNO=^(t-`pOz0Xkni^)=e+OdsL(T^~<BgIxh*oVTEBoslo
z5rylJM*K?Dz6%b~v}N}T%mX4c%t$pBQQ+l5yE%Nb@Ln+&iAt9`6=f3|Lq#<eGa28@
z3|cRAozR(oErc2f*Ki_5)qJDC^MKzYaB;vWL+=pE<Ay>Bx<|4)^l>5d^jjS9tiaM>
zk=%*L0d!7aZ0y)TDCjzExje%_kd6(+Ul5JEkfLZHo<rdn($K-8!b%<<yV-c>W)Mv&
z+Nh&LrwU=a5k_T{OJFonMl29Q8rbAtEZ$l6EoKv?+1K0f6C9-3&teID97$2L6jS)F
za7-kNk`;o8xz(NOMIkDR+91UoFk_5urhl@3R)`BnjwOL1Do5rFHKCrBvHS(eA76J4
z#E;QzWXlG+ZN?aNriBcchCDDAKG+A74Z+8W!SXX4ycd*lSUpVCFm=MvEauZ7(BYG(
z$pB9fP6g_Y$}KzKDcq6z6XaH?vv?THR8l8GqSZM=jg}@1#SazzQp2|mrMQGfDE|H_
z{$`vLa8f6djrWj3G<ZA<;}Ea<Ss}88aL-P|;h_|EGCqv@N(FlTn=*or23AAQ4na4H
ze7{5;O74&L2sDm$5B3Nw6UrDpMr<&}Sd7*Ekil`P%3~cN`oO|M12Bs@7HPl;Za+NU
zU!^XR>l1W$WV2BTN~}HyCi1XZW{yFyT$nSHsGp)UlS5R03m>L2b}~1GSaepJ8m3Tt
zI6L~Q1w%+Rtv}5E$zyD!mE#<v#{o*3fe2fyrQeOXBsNy@%s#wGh{9-&Kp_z>_|F()
z$E4;iOb>x()=(^ydV>Qn<YTN8hVq!W7>vmkKvmdm<%(q812L6nP)D(d^T3-Jn|d@W
zT8;H^9pA#51LL&;G`}cNcrqNMMh1lVn@0~HO`UY$V45SQA`a<P9CmO$jSd~MI5!0U
z3In~s=uxE7nImF`lVyDSNK;e%{Sg;_G!iSbzfv-*HV@`(&<pqH8)J{5)jRJtKrtbl
ztZEir0WP3!pbN}`JUqB&5@Sy7W_z08W4MD&w;V&on#>Au`X?~OA~vH%k8ob2`Gp17
zAA#hPV+osa81uZ5!1C0ib1el!3jJ4DmNWmB_D2}O{UhpH@WS9WlvP#zE?^L4=z)TS
z!e~Df*+0cuk9;4}^+<n1I)`LZN1Flz>7=Z%-5{zo7(3Hw+C^cf#A9ckiJg{-on-};
z%CNI;BJm0muOjic*r}Eoiru7tx5NC7S+%Mbvvr{*0J(ajWx92m#k|lmfJxQ_t_%sW
zsMbJ>-+9ZZg_fu$%cunwzbw?6ZHcm~3sAT*#4-s~i}gl}IYiz<Wg#RaG$hn|0xj4;
z9%Z&Ku-FSNqgGf(uCR<)VX?a{HuL5}v;Qn}bD_o0Jn$08O@$WgEasq<`Lf=^^A#dr
zk4C8j>n(mFC);N03f6|%T4|X8dDJCnWW8m?dW*dPniUp5G&i5B!6(Z=>V!qX>zHg!
z7Bl6Dhd;|<isiPyTBZwN3gs1+RLU0O6Fah%Tv=d=Cuso(b9k9S1(r!%Vy!I4ui?O3
zVYHP=BHU!$LJJcMEwg!*bjk`X3%KOwk}${Y6bd^lczxJO_Jzbezrn)H+h$qDP@NTA
z%;LRUz+HIAvVynm;+fnw2A@DwjAJ?G5arDELMk~e&DLkwh%5}TSVAy8ETbS!y}=T-
zfM(90`e)8KiogPkI@=O&JteAyS^`lT+GGiwWtoI|U~9AFSt|iWRax?qG4==|W@jE|
zXH<?QZ@wjJoF(rf%cvYn)SYa2e_>wIwDg*r9O|E+1I+;oETiUGMlQrOY+hio-T;FO
z_36jvbts!$W@|I1&|u<XI+HXK4%%3eO|B+BxlS14?KQWWDNrGvyHLvj0w@eDU%_oy
z&0AHjS;8aDhK5I)c+qTUS!P{{nkKkwwuV?1S|1jY3TcHbEwH3oFYQ+@uuNg1ikhpb
z5_(KM#h5wRfsav0!LZpDTGH4suqJ+ip<4gH+P(!kj^a$WXQq2JmStHF%aUc;9$P$2
z5{*U=%T};i%ktZ@kR=&pl5-}bnI37vXl9s)Es>nWm=HoBiyYv_gpDNvCWadmFJXBu
z;Cl}~D+mE6KmrT9*JdG*z)g&kWz8kP{r-B)qa|}|a?WiZtGd7X>#x85`s=Ui>Z%$?
zb%^tT`zQ|?jx`5y1NS4B<5y*4(eRv?I9IPmg_4jFXPU)z;qJ24sQ63Z2+pUpa;2H$
z@4rE=%W!s8B4Vzw)b)49`Bxb(`|>I>3+KMy?AvG9z7_HqMvxh>?HcIDmSTTcbF?{K
z`*;latRsBDnB#OjZ?m`BjCsyxW6lrvft9OSx5FPgZvUE12kv-VUrhj>r^NG{Mk{=j
zF>fAlo8$XmgW9UdgFZF@0>kO}BPKAm8B|!qwZ^Er7&Zy=N@vR1IDyhuyQVyDqu%L=
zxIFN^mCO~_CyY|B@kJ+ECYlU<+CjtV0r&E)D3J3wrJDyA;dC7Oy4z`URoL4rpus;^
zx_)c8;O}hC)V?XB?1-`WQKNL5ah~&XV9bOZyNoiF)#0@5`wddetK`1kwbZrG4SudQ
z;O4P6Te-Qm8fDiib9>9!;Iv;<>9pB<IV0z4r)$5EY<%CqKBH`n3bQh11TOCMOaB=q
z|L;bz&6sb4jw)AA!K9p3WG}9LOoM+$hU<mok3(yu2o!INQFW~`>xfb0ddvVXbX^AQ
zoBE0Kpb2aSFcwpYI@xB-Lh%n8#bw}p1NC}qT#va23@=%poQXrRj0SsWrLhu9DY?X0
zxp$w@u+>;d-P9W9A8d}QgRVF6=!m+hI<GSvcQ~IxGont`ag|XDz}c7;aKfm%*(ho-
ziW5&7MPzf^j0MhlhU<&y)OcS!3mx1XB%X`latOJLi#~@-j+$j;klf*{Wg^UV?t_KF
z9lNTGWz?ALYq(;d57dMs2^Q4JQv+uSq)H6uZ%M^&n^Em_xsBP+K!(SS3()?nE<v!D
zVv5i`!hWU<TMbmYl$2ApgN7*hDDnR@CoIKAPGT;XO2?!hfNbSW#=s_{{7z%wDx-4P
z7}#r+hmFdMje*T{Q+$n44h7MWWx^<TQWvG~qdx<@+dGK+5)}3Vgr#Vz<a$dRj3x@U
zadb5Y-)HW)6t%%<7^~84M28Bb;0hk`50Z|sN*)Dyo%A?h)Ep!gI-HNOK?^$fVvL)h
zZDt@K(J>BbMChCn^6{q-pb|I)a`L5b8p|mNG0%m7Jw3*S-D+Gyaa#?%f>V(D6ygF7
zu<oODhzO8B1xq-TDSlU`7dE=Mio2lVw~a2Q;x49Q=(KVxQ*KtpLj|aqX?z=V$EEx7
zG>yt4p|YA5&lCB|{Hak)nsg0R8VjM|i?$lY)DA&~r_v*a^BO#{(Ji>?XN=DyaE=>`
zC^*QUq*^~m`Mt>+7WBz<7UEBj8OtbcE5^<&paGiYNmZ>TTDGf;hS&?7<wk9zvEXI;
zyKIfIAYs%xX1R?8jy)-(c9&5yU@XJH;zy8w4VtjcC>t>9(?%HwXt0EeJ3noda^QSU
zabbpn#=Ye*@U&6RDV745KVVd*jrkm)7KC96lDIe^lGJ9B)PiF6fUz)b%;prNV*sEK
z08v2F1!==`w^8C4qkbG^;bhervT6Vr95ttnS)5Tb0CZp%rHy$UxClVyfUzWPRB~Vm
z0F=<P+bBhAa$aJ108|YaO=+WwQ#1jn88DWojT#Ot2QYWQ_<Y)!%Yn}W0MjpIn0^5O
zvb`Xa?F9gki#MZ3F92u>fY6cvK9>RLO$T)W(M4U5=oJG7C@OfiM1gG2%Vc{VfH^GN
zIh=yBg`}Vmk_rmcxS*gKhhP9i1q)zt2B2|tF{{2}&Qw$%l873oB&ZxnD$Ix^g-xLN
zf<hD@Mj~<{BY~taG$98W8YBgUkW^4W4nZO0fYyZ^q;)76fT&vmptJ%IrA5669-?0I
z%>j~vLP#noV61`%m3F4eM&X17iNZk+QMhx-0j5F@VmedZLiG6)Mn!^Iw~&5*Ce?hT
zLQMb=8W2ED2B-lA>>q%ze*x5GfI3h>@ryD{=_LXN06<88E&x0P#kqJ)qd;>N1*Vn?
z(E$k21yGd%Xxv(a*Yvbe#LbNA8<N5)2p;4Vz$3r3a4et@jzv%a5V;5df(d{`pXO<%
zF@+qERLDV+LZ@IVbSj(^G$x!AX$<xcV;1%=yk@?+gTkGmz+ietIlxp%N=#9FxIj^S
za)FQ(&Q4H}vqSMgA&M_300;^J<m+RaU!BGj#fOqb@r4C}sVF`%1rO*{6rXg8!sYiO
zFe#K)7z~vb6mYsiLF9C2k|W<fKmkDHB7FTcz~P)UX#8G^*Om#R#`z3&der4OQ-&>!
zL2kmRcU))R$NIu24M!Jx4Gg=NIbDytc766gjLR`#T7r?Mi~0bEvkSLnZoDNrcU8M~
zv6mymu^6vCE^4b3Pb2U1@J_XihRDyb!9T`bH0X8wGuUAoVF|ud*la%E`t_;4ls{!&
z=1=&Vd_L`~{(vtK!w<Wd3C))n-DIY=<M+lRUE#t0vAC%PV!M4w|42MyQf%)?USu+S
zwTXRt$yj8!+2zN=J^p*;!LeY<7fHr~e$D6W?!EN#p7qEG$yv158*FJ^5eS6*fkA)k
ziuRRlA+tH;9c*t423E9&+MB%I0%)5GplvRIwndACqn#tE4j;~PQI>o~0kmxe(5@_i
zc2xni?Wmxuug=(mPa7Fcrp&mOil8?8v|eAYITQwx=<M`IBK|>GL2v)1U0Q^e45_N6
z1)1bpofpP{g&SSr=yrc<!)W}{M3CFwM%o~!%C)JffU0jQpz50-uWE%iwH83TLJRbw
z;c0#<W+W26vd_QU)Doj;Zq2?>)En0DYY3qTRN(7bpIaxfAR6jODvbSpA+*$>=1;{U
z1(c|O_EdmQ3lN|HjTWHL0vcNZZEdASitZmym{_N%YzSyb;nbetFd9g}#LkD^ux>M{
zjrd3VhQlGOsq}@T;gr|MJfN0NntiEYPi&8J2F>bq*sNZM&FXd7tX_xBty*{}8cUeL
z^|7Rx>{c(NdNtt-q-xzXt2@Nzb}bx)%u#Lg_KpGVN?KmV+h-WL4Tht!5p7E}lum|Y
zQPz}IWoYqg{=rxxr7Uy{YKFW44p&#sqg9KVyTid|EgBn4s@H6bx++`LRoT+61^VIW
zAPouCt_5ShIEEY;Pf&?`{^U?pi|zI&;Oz#};YcbR^^KS#Nv!*dZZpmJMlqIvbM&$C
zMS=%Gf66bS3aWut^#;_c-hf)w)!eGyfLhg^AXe3>+DB`fW{##3{yx~DsqWfZ)$6>q
zU5j?1Uz0<p;BJ4k&rGG`)ZK@)Bn><J6XB7aTJK;i7BT%%ZFoKQx~F_YW~!gMH$Hyc
zXc;gxW56J~N$NH?`A0_l8w+R|E7U7=g?fczZ?>xHSEyGg*8VATRd<&w)Z6$9_4;1X
zt|gNl7>j5$T-=iEgV*T|`EZU>$_%8mKsrGkK%bTj_|Y%&@$@>3r<18Lj)6!=ljtVQ
zpgS-OFAz@pBC$Y7^Z98rgKsoqM)79Rx1rzH*WbIPdy{WNSHH4m+SD6+o4ReZsoO@I
zx^1+n+XmKCs@g`Ix^1+n+eX_e*0<!i?A?A}1TP3chN9_!FY~w#?OXGW_*2?2#z4)Y
zVPg!G)W(LxL&M4iX;L@GmFl&=QoYt!s@M8T^;%!qiZ(McNN>oB?;0!BYjmZ0jjk*}
zz3}eTd*^uKPJcQuJVvgVzt7Rw7cYjxtY?m{4=3Xh|5#5VHk2@vNwn&Oc$@5vrP1GG
ze2gK^mS|Y>2ZL;2ki&YNnPNU3PGUq^N2}B=V3oQB;J6XhN?N6E0jt#8{VMf#zp9Pi
zIXWT{>Q0mDa;uxcs&*|D_r*dX8mh%pi6DGAJasA*PVgxMnkm?BriNp|^=1gezAfxk
zUz&|SNnRf&gT5l=>+R^@pe$tbio9t9Ha#`$Pi|miURb7=k8h~Yq5m2N&)t3uDpV`4
zUEO-x)vc#p-Fn*9tq02^m5ZfrJ?(8;JQj?pvsO2o_V&!1R5%pHU|=Lny@u#8d?`$c
zVCWD|cBdogIMQ*9^G)``029NIf!-msv0yAE?I;i%!Rv&%(H}^~5|{ZSm_X5jW*h`m
zf<7%lGcEMOAto;)%v?5a-_i}UgGp}8&RUWoc6DsZj!(O|2ScwlIvfp0XduJi;#F1N
zqHZ27Em^CjHoFDCfvKF0dd;@9Y3sXu8+UehTE^BbnaQ}G;c!1XZT5~p?Wfz^GXOuW
zp^0%VV>*M)M0KIAgr|nmIYz_yWO{@tx!1R06jO##e~uf#;Ln7b*YE7rFxIC7t~Pb=
zbaytb6h9m8ZWu|Wqvo0+GioNn0X+H>f#EfytJ)gcTHOsp?uJWS@W0vJ5Q<ZD$e$dG
z28I){Xe^y<NJsae18PXo1Z5JDc+&`6*&xQDu|z{M1w!x;5djQWhh;Qg<q;7G<veFg
z*POj=(K=}6ljg%=lG>m@NIln{VK!YIHAw(TOfM-tv2c<)M)vN?-fiM$G#HK!k!sZ)
zb1S;?Kwo+=Z$<^~Jr?NJav2IJ8I33K!|0)$T12<ICE4A9YG+#U1%|Kmh5WE;2#iT8
z)cSTULfT<}=vUSEkh1#vs7i4Vc{tUDc`9md5p&cXjHPK37?n=0wFiozL0(7;_yfb{
z_6?|Fh(@n@WXjQ?If^%g=72vMFe4DLhpPaSEWM^bLJdALL@#qqaY)(iXJ9nQ6x8cW
zM<X$RP|TRAHi;G9+)CIP4Tq=_^5>s@+EAtg6YnS@QMlQc2=i&45HgqFGw}Kn@X>Gc
z#$lU~v@_eiiFS}POA(7;EJH6l+x!^&j)t{ZBDf`pB9Ed~;3#C@`bY$GNU>ZwH<_^!
zl9RV3`h24K=seu0pIJ`sY+6=Qw7uAFGXX*O`22|>{5SoiQtDrM{p0h67DX-eVF(rV
zp<DK)*r<SBx|>n|p}syg+s|Iix3DQYRHW>^lUjH*h#?w%+zATNvb{H(+U!I3j@h#G
zh_<n#s}DUn?L?1uQdb;J(%cTj0l!$_GQ1DQ;30@t?yNVpl+NLB@+vAl*Cl**(!}&o
zc9_5nQy5+V(~3#F77q0E@9NorGIXI@*qog<5{~xuF*{0ycbh$!X9&|xM`DP1rQFui
z&F7|+9qmdnb&uCuHjgWeax9K^6`*#4fidLjy-d{DmS{Yky4;_jaWQH-oa`KiXHEIg
zKwuc%q@`?=%e)6u+lwaA)-g;*x3wEupep4?foj|+nrt_;j*?=W$0wXQL5>N+!{$tu
zMa_6pynI8&ECO!G7mKIh`jRM5P)kO3)65+I2uRB9-OwrWWwuY>2>JKC=?nxNisnlP
zF%qPYh8sH34pp1%D)qOOrY5k>Oc%nV33o?MEw?>wBy$Q=C6E+-qF*B8M`_`c672g)
z&f5mQLa`)C^6BWfB5>@O%Qr_cNLk;pzF)RF8fEa2u5$6z6`?+8z>pGk+&kJw?FoZ#
zRaLhtQ>u?F3V1D5&b@$Pr55z#W3#kfw61j_`n)FWw1C0?N)4VY$y^0=OSYn601*@;
z8ho)?KNj_mgadtJBZILB{9>|~X1;)BzTuHKTSp%Wvc4Q`(~Mx0n?UE`$3PYxCI%zn
zKne|bx1UX&B!@8}>!!v}Eic!M`?FpVJrfO^(bp#M8||v~-GqHy*0D4WTa6}Sd*VS%
z%wsCOzjx;ba?>O5!*JNV$6@{p1HWi6V$$ms{~oMGw==b1=)(*T4vmxeXms91T+(Qe
zeDW_djO^fpl_>R}8IUveG3UZRZ1Tp4Vuqs4+OYU?fa=VQz(~25o8dvcACOwm>A<rm
zrMh^3!Su@#!{^d$bFWEnshD<3CE>a^LeU9-G~0Vek%2~kdNr&@Ko8X|$GCh6LwJyT
zJ?>$pE7Z0RT-v+7n+JVxNFnm)Y_$NJ_}E6H<_M-eFw>bH{3^P>q!wN0XJ5luchRiz
zI?wa`LI9XkrchMm?pT7E6RLbmvNN4XV1}iWds&`k2hFmRMIb|Ntx12#yu!p`_8E2e
zlQpXvO<Ox3_H7Q+(0^DP!vG`4|1}F$p-iR;L%>HZ^7*9SO=c>)PJldAk3WII0AAnN
z3_o?0!5F$p*cCpc!OJ3&^^%Rlu@SQ|k&a^U-?%51xU!L(gA9VPK(aAmhD?+`U^b47
z`Q!0MsuYhHn0NwOV>+2=3`YZzbkJ<XdxNj7wE-GRk2YXtvT=|sfRRMPgN;Lhz;u$P
zCYd~sm)3}$vaz?Z)zj+nHd5+dN!QrxgNn83()&{8NW2kCXcGR!m?#9Af6|mGKZ7a!
zLogg`_6*Za&<yb#4DfKXntds22uCmBRDfur6(3S?mot3oOC{2Q6tLdl7#l^?Ej$c+
zDNAN(X&3|VltjZfgFrhH;nXmZ!p%@f0gVSK#s9`$Yue@~FMWLVJ*#ay^?O#oxLU_4
z2@{CWf4=(R)i$|vTR(B}>lcIK^@~5b*mjxz$;H=Z2tj(%^|1?)AG<zr$>hWu_@7+1
ze*I4rJs_j|i9210KI5|O(9ckGzl`3&qY?R`%hs!Z$a(f=(D#VddmJ6epy04ea@Zx%
zW3E?RHoyLg>!i!ce#C;~$oQD+d6M^eA{bx<d14=Py+i`NMA18BbXM$Ru3r)UD~jGJ
zqlMUD{`_Y|>yxgBDA7Z%hx0Wx>AH`yypN&>WOO#mN!Jm|@(4xu%V?=7#<N$V&=k%(
zVVC}wqK}H|)0)eDp38kf@iF`2%uj_noy+RaI&g-`p$gBk3H_s@Pt*uMDZ0+JD6M@K
zuj^cQxT-kGRa#y)T&QD3zbYd26}?*ot)mbbai1>wg_;z<D0*Iv@cE*bi~b+tz6?1&
zOWet#!$q*7!$tQN<yw(GUUY!)14Xx6aN~6YHT=5qo-=Dy`Z%F}hURjCxr`f68mQAJ
zjo&+MU)F!`WXZBMKW^M(KwI}1Z#&WM-saJ==ASlRGiJwmJ(g2FzJ_edQyPvyk=08u
zHt*^$VuM6O((Oc*Cy&~lx9W#i9QJ%&qqCHf-?2IC%jm8y8?|^sv-4Cw{QS$IO3y+@
z=4J`b$p9JDyFjvx5!c`j_64?-10`io%P5z6or{2c96X=4JSyTuGA)~yI}@_rl}Ua9
zH>>BN3q6kW#+edu@D@`7JwFbm*#<$S|G@b;zr+1U#=}lqP=6SD(J5!`LdJ(0FgQ7h
zQ@xAvD#`K!@}l!~r%MPmp}$VN{lxn&;Oq|fdz~*jZ2_ipCo2_qwfgnW`<!#*TCVaZ
z^!p$|c~<efPN>cpqwAgc<vJvM#mPBgKU_AVCMTkXwIgIrDcm^`ZqDj1mUoHBo2q6p
z&%2UUM}CB+XERxz!)rOZjf_~{oCvSx%toxriC6^~DJJ#%b(rz}`inYm^a6cIhmU`V
zqs(*Zck92<?P-?(|IlC1bIqHxa&JZR+IoQau#O7c-Ep+7_yOJ0wA?r@P2Y5jj)rrK
z&JtwXmAez-bbepA_37W|ak5cu?Y8yk-EsP@3|#s^ELQ|Q7N_?0_5Aj=n3E31xB<rE
zWPU_L=6A$?$^r9Z_4z&f5xZ>=&^3<Y-PBMH)u9@A>icc?+F>;JB2JH!K`vo9L>P=o
zbBx8w7zJk-<FodU?XE*Lo<oZ~3n%oC?LTtZhPi=!4>7i3K$H_Tq(cibIX!55)@~a?
z;%gjE%83q3!li%DzTXZ-?zbP%?fW-#>`~hv>}V!`$k5V3KA?Td_Nv`J?$%$;MBs=4
zO+RispxdI<q`wC)wh&>zAndDjcY^Nd<N)0N#`f=a+m-siXVP;^VUZNSch+;WP|?rL
z+FOkJ+gp4?F`jNH{%$d=cm2D?2MIV>d<xMr5es?Bex%%243D~xQufP~{S!#}0|I_P
zDZiXa$y4n>Dwe|zqI<j;Hu5;=9QzM^(?Z!RDSHKFub`Z?QVvMU0YNz+C=XjHcS*`!
zf^wIjJjy8{<C$5G{jV&7Y$Vhxi)4ZS0QjVYPiEjZ0KX&QcQWuxfKN;KbO!!EfX_<!
zEQe7M<L)zVX!?x%&U#eDo%IjaXImbNKH-MZoN)hfsRM^9>VI7NV7;XsG7kOVFj<f0
zHpw}NrU>>=y3r(_bU#>c-@jI9nP(b6l8@auhUuI7$A~7bl^)?4Z=Ie@{~bvhM4|`l
z@uM9=XHPACaVhw}2(gG0*Vr=z`T_oNX~xn&U3$YZU^guL*)oZtcqi%}ScK9)u;{N#
zQ2M_vxzU|h`V(~@)*;b{b#E<ZrGIPjjc!Zn8Havw_-F}A{}JaPOMjy7)<ux+)<rkE
zt(gXp<cUSB^iLpKmYxd0r3!+8lJ#)k^3rrx`aO#vy(#F{^0HsMcziLk9EYS?g&bOZ
z?_#)%dlyqVpgQtabm=U@kBFxi+xHLXPcNRL#8XI&M^190_USsP=;^xSby7DNI{eGR
zR~JGvuP#KSAN&BESy+59`_e)E7&b9zAUU(p-ZY_~S$KUNT<`UDH`h7h-<;5IuDcu0
zB9EG<YblS~<F&bYEO>PR<*@*fJdXuu7R;OnB3K>^uCJv$YHzOPdDPyGXP(FDs@LWt
zkJsiuUnBEi5t@Fw>QD3GGX6CG=4w>o&DG=8Ab=|wuYSJ9*2h!7H6JuP0X<cXu7m4O
z|5^1L)fstC|DFCu_4PGS-~O6M5O7DGM*@YvQH{6&c<RThpQ&NRf|}^*jcU7h0`WK1
zARoqJCm%c8nJRNh{;L{={IBZ$HLQ^PYi`4hZ9>1T=AIfv+=J}tnU_xd)lXKQng_<G
z<~>mbOMjy3com+GSG`3~Z&m%Sis?lEUDdk)GWJ1<^<T|9H7~c6r{;aD3Z?v3)h$&p
zrf=xCRDtCnW0^~Kyy_2tc@a1hDj&<gj}hZh3O8B)PZbd7pDMmxi5I<ZSH4_nsiMj9
zCn``yPgMM}67v4C@&pl}icWx3b`a2$K3RUa0$%-a#VwVn%3CTQtkjd5dqRJZvlN2>
zR;<bLqZJsP9Ibev68*>voF&{nV<~$MFubk;uWZ+mWLHSZ_=~AszrONFB@FZkvLQ{d
z8YkW96BSQYP-&m4I8lKEEXTCEmh-{6;^m5B_OyyU|7*p&*;wjS4}y^}I53%<d$bHn
zIa>BXne9vZ2W7XG=PCu!Kc5Sg{CqB=uaME%5zyq^JIbJmJ1BL(j7Ir+y9E1N%N{5r
z86PN{Duawya~Y?~epv<?@x*2Pb=iMpV`X)X&pA{I$qtpiUy7{XFWo;kPqOhj2TGCW
z0gC>TjLt}gtjFivIR~=cIp@Z?Q2mW_Z=&?*lW&^4A3!#j@lsUKII-Cw*<cizDZve;
zKPjaW{G{|)X|@E%N{^T3mf+da*Rrv6%8{l&I{U-f_T~xw!`Tm%*jqAI_~`88v(bKz
z&%Un&Ms}p+0R%!AW-%Ym-dlqBy(QO_p!C<2{Imqp|Fq-?fsFdgREdwyes*?o=6pZ)
z{zH(CQvmtPFeskE*<SiHKv7yI(bFZTN~kEOO2+5Fk>%!ks^neD;N6n<DA)H&{!F=^
zExCCPf}7`z&w+o#8Mv7VAN@@6UyJo0%)Y4vFGRC{I9vb8?1yGEp)|CIgO$w4s6D<s
zPyaVK1M({`SL^RoAEgoAXIyNAtG)lJm81lKUBfwXo%u0981A*;O9dXc;46XKv|4Ry
z5;wTGkL7G_+C1&-?HreXX~;?W9<}hHgB5LBrFPmve;aUp2Kj%)=xa6k>sYc2dsy(d
z9yj>{qp#BBPhNZROxL;e^KUG!LL1#nRQcCT^1RDP%C*s}^Jzc9qa8n=IBzf?*O9SJ
zL*-58<FyQ5kkO}%y+Gi{EcA4A7k)%r^Zrm$@^Ce9yH>1eYxD7Sz{^oztM1~s)Z2)_
z*WQ_r9{^rKb|TJh<lVBS-7DyiS?FI9_^Kc0^Z8WZHI{ro2fsk_)L8iJ1dh{Or(J%A
zmrgqY_9nsSz>o9uedHW`o;wHrzZhPso!rFv;3PYCod!;FKJsJ!EcI6B;P~0ee0-C@
zPg}~B5V*^thaU<26$_s!;4|srWrm|(HuHSlJkWkGaO>}`{zdRVaTh0)@-N0qJ(X+N
zf^QM{qy<k2e6vLl6Trz%<QX^ewCQIVE`-YoUKM<_yYuDIF_%H|ykg<+75E8@KIun4
zDc`lCzVOH{Q_vr<(9?0yZb6e9ObGf%EcDL;C;nyP&=`5V{C^7EdZx4+Q!XU`QH$QT
z0w+G!pXS^p=tnL2eoNqcEqeHwz`YjyUj=^JlCNGw`A)wofs>qN7Cx5<e83`4Sm0+Z
z^tTCom4*J8z%`3J?*b=1obYp*I)!NTJD7x9&&56;hB}k{jlh|nE%JOx;60Xnzb)`G
zi=6)`@Y9oA5InN$b=<hceQt18(7$8hQ$LH#<Fd%pE%5JH<QWtAS1j~*3;eK!{srJO
z>EVrY;O{fMRI~ojV;#B#(wocD?lubighieN@R{^*lc1ln=;2|3k6Z9JfX^iV2ZH{n
zg?=dx3}x+391HcIq7u4*<3riBYbT?h25Uj!%cjHHHwAy|4=(>b<6o<dTJ-R!z?WO_
zHw7NI$UheYEYkBMmU?U#_$wBiex-of|0XUhs*hbGg8sB6-`fQ~VBzz$z<Vt6{7K+m
z3tl^C=6YEMoaA3_;lEwrZVNuj_`u#ex!gEcon5yG-1<A^6N1l)Nlqu*9i8+~`5v~E
z_XC0NwdA`DT_DkKw&2?Z?y{7ZerTEK-?8L-7jWj6Eb{+PhI1Kn|NinE_-_TDlRi!@
z>+v1nl<!H4Je5$?Ozol;IMEMS^gu`X6FzP!Z=CTd)vUkOdWWE&vhevAfom50jKIHR
zk(186XYEl80A+jJEbzU;Za*)Sw1eTf8sqnL!u?F;{Tgs8*D*`E4hekRBIgN#AGXN(
zk-*~?`U~JRh`-x{?-2M&OTKhuJkbwZ_}?S&xP|`L0$*#vKNk3A3!mjs0`XsKk*81K
z%PsWZ5V+UE|6YN+E%Yw|r}}bl=6dH}L-1oE0zYc0ufGU<+|TJ-MAgv|$Ry`iEct#x
z;KwZT4+#9Mh5lNBYZf{07kHV4{#AjWoaBN?|Dey~^1Nc<(<tx*7Cu)9yvHKvbpn6K
zg8w6MsAAgnoS;8xp+C#$F`nt;N$$4su)d1x|70hBmcMiE5%_80A8r@)tAUfA&sxeg
zB=Ef!{QH8B^(Q4THNfg!o^pChl;j=3XVN17y!o8oE9?QoT6Wb5-1>|37Xl|ar!0Ej
zA@J{5_<vpC!xsD}0-v;$>(>H*$HM=w0<W>~xezaC%swrAdIdge;iCl-sbnf03V8z9
z%_6tU`BEdk02T~mk(JMfee^?-*dTVi1hHT(>GP*Y#p-<Q>h{q3*cm8%A*_MJa^i%4
z%tu@F#<WlZE6set^vK8<tzgSO@deJgq-^yY1j<=pl)FCL(<GJv^VLMO*3pB-PASMj
zi--7XyCjorV{gay4ZaQC>^LAFc7XUcZ1!<*^ZH(mNvZiZbzQoyqsw>c#*MW7x4&au
z*9M#(<m=zwnMoQ<#>Cn~tf$}A-LZWO_O;M5Kjl?!1#A_>&V8l=v0RER=2Trwhb_{o
zq^s0dy%w<16h}5-U1nyRg{m-3>T=Ttn3)a=;-@*R-PE;ZU8k?v)9hIx)R=obl;Y-B
zwvbhkwncrFa{(Kw3)oYO!#?ts!gAx8X+<epE~;83>Pxbl3LsTqbY8&rNF2Fhj>fZ9
zuWn3Oe5$ICCa-#f!7hJZ3$%AlH3K@e!8bG#i_*@iL<$?_lvAophQo4HxfXB+P`fB@
zX34T@9tG^%Xjk92QOFjLY3qm;n|%RWuW=HJsy><uI616<!(gzP|8LyjhVu}ztM!#j
z;Z@&Ryi$F?av>)NL4Ym&+kM$V4Z3q0()6=K<x~q&$YB}kZLok-+zQyDjlD`UoOpm8
zm4Az0Ea1GW0yZ3Z(S)g1<UsRps}uDMabyn6e%e+V)!HxM6dN2@mf2BiJw(PA81>Uo
z6WedDHnWwH8J{qxN!wLa<-)dkRZ^VKph}9<lvGL8O}$lg473+Rm3tv4Ef#QAMj=O6
zHIdU}8(UR{#+FTGQrTPPY@y`4UbAP7@l8sq9Gld;#wM@2zildDU{}BiH#i$WwO}ps
zB%d5rVV_UVZd25g+yaJMah?Modu8|ZD(8oDl;rkx#bH1J$88jF&}Xx{ikiIYv;`c!
zRKUSxIMtCW9D8z62Y3h1Y{0f20+8{3>Z4F~)5eh>s&=(PU1#m;4ZDDIbJWL21)PW1
zRKUQZfb)n7I35s(^{7@s0S6@F@G4cVI0lm2qI?q(`^Y)SHnrz%saNIOq;4=xUUiEp
z;0Upn0=TO8&jp+u(^f#op-x)B$wquPe(oOpj5}%ZsGCTWSKUZ)1QfR%c1A}2p}MN3
zQ^3)91soh%$f>*P<!$oHi3@oiDkeGA2d3(cq+Q)a3pjeJfWs`+Ewq4x)9CaZ#c@Lu
zMgy8BIW~ergb=0@JRFvHQEY4XU?;H$qZ=FplN!@J8Hje}1{<(Vo)NH3(;jTR1v8C>
z*!0h0a9RtU9Gv&yLtMg)_=!;5#Um-r!zQymxb+O-Oe=PiGy|F^MLUi?G@TD>oUo5A
z{2p^yOyCa(!A(AK9(>ZDr;zuQiZTTQCLkSa0HI<j$_5ByI*1edJUI4e1m6pQnS*IM
zV#XgG!dD3K$PPD=QG<gCbGMAa#%ojFLxb{`U`(cEpJM?!0wKdN5gp>s;xo_R94ti2
zBK)@t4u{2!4BY~kf&6|apFOzAO+O{l%kVr&C_*e3%ey-8^uec}#$^7qK94RL9uo;<
zI&1#hK#Qdt8TpTrFt}t`gYV#U(f2SL{>$`oxh;L}bKt+se+u8H=#t?&@yt=qL*7&P
zCH((1NB-gl9xFpyr%9KTU*;pjVZ?oslfH)(ic5xHVNYCsN+abLu(k&`#9Q(^D$>hv
zGuDgHCG(f*W%;iWaqAeBrk(vFzsZC-@>3dX`Wt}}36`z#Yi)i=_LX}{IVARN+~lTr
zujMf^{EEn5(qmZwyJWZ@Pr2#$cJO!^dii7KuS_RHT5g-0{-{VV!ybyvxnw#S(r3xs
z^qF<oojg(HlAOf2mFXmUFK%+v%k{-F#B!Ou%c{ShiS)AmwYVS`;V9N+<zBM>B=#_F
zD1Vt=u6veYtEK$b{Eu7GAN?BVEW_t439R|Q0CFs;5gS#|?!~(C(>HKFR(km?!`DDY
zO`O^=U2>iJS;h2kBMlW%rkCr=weJ!)P4n^~^VglZ<2QdnK(;T_%k|q{k^ZF6pZp##
z<(77E8`9?HU-l22Z`n;eJS`dX$4vW^d_K0M*CqwwI1i7BM;XX;GQ<xT=H-8El2aZN
z=_Q>ty?ib~dZLkHPEGQdDUrTMMu-Gfe&r&6uZWj{OfT0VNt|>=mkjwjdyeJ){j<p5
z%2%e7v=@o|<+BXy$+6&iq>G30uQmOObEJQbHXh(|$%^NX8M(!y)_IQf|M3Q=pmjC5
zmo@)OMS5#F_kX}s{kLVgsk3^Me6Bc0`U@9uin}DCU?U%8_*LA<``nzKrHaAzKGsd;
zURK(x&yoJ|FL27uxruRaO;69#cjc108aU;0k^lLK#8Q$>`;q-0t?RRxC}ZQjgs1-o
YCCa(1`4jt@(%-+6Q-0Ty!1Ad5Uyshhh5!Hn

literal 90360
zcmeFad3aPs+BSYpZ|P2xq_Y4CYk&j-gwWYY!kV29BoIOZVNC)FNgx|pARu5AQ6fac
zj5;`?#${$0$5C{A6=zVRsDNwSanv#HL2*RJC9e71PgR|h>Q=w6{oeQe<9A&gt~yoE
z{oM7`Q%}{Yvvm#^XXR!aR8`S7qcT?+loTaZD3X^CqNz%<VuxmuGLdmjTIHX~&&}qb
zmMab}lQlCHQ)!1SmF@L>+;}c)xm=I6*C6;F`3_-U%jJ3&R5{nq>z%vaJwHuKY*D^E
z*XkJ{lFE4{$DO}-&wE8!wA{+o3i(LsFd6unJe;A5qUE0TPKtV-d4#B^<<tgkOFw8X
z|9@#OU({>e&h2>S38H^muC=!W^+d>c#sAA*42x1A+WVz(a0sJXE}CbtBpWnm<}WHz
z5T0)G#VFdPa)^*gf8xkM<+--ieKi~JefPDS?~cCcyS)pEx*I7JiEPnO|FBqTOE;)(
z$~Xv0YMatVIkid_#UP+$1lDE>SFAt!Kw?`sx1w0vxGAOW`v67h2anXKZfFZ<$k;aJ
z+&o3;y|_#19#E9Ff!#=O{KqSlaPoz=FrFzRL_SjFza|?6@@S;7NYtGPNRdcUNK=ra
zkz$aXNZJ;w6+|8<@^}eP5amP((~2`tNJ2^$z)a+`kmewv&q^9nCQ>$%3uysT9@1hY
zZOhjRzmP8hz7(ktsR(Hq(sHC?q!majkyatmwi;;-QVG&Jqza@eq)kZmNKHtsNIQ@&
zM!E#)QY4xemm&Sqb~zMRAniq>C4`oPeMmpI?FZO_bT!feq~9W4i*yj__eh73ZbIrp
zlDCUbAG-Xd#|_(C%zsPIN%-=ga~{b#*o~y+^H83PbQ~#F${$Dh&jYs|N=N<#66Gh5
zo<zz*%0xOP(w8k+hrYag=E}-_Q_E8Ree9!qW{<eN^3RXDM&9&kZQs|2F8?SvcjX#z
z*CM?m^7oLxi&Tb`nl^9c`=Y!K`3Fdpe}wcQQUy{UQaRFhQ#QQ30r?rEPmpZM3zC)Z
zr+00->5^M0)rwS)oYH2Invhfa?%|w+-y&U%y5A$;h1}T9%Le3^K>iW9Nh)6o%nVsU
zPGt+q?S8JiNv86ZDDOepfMgZ&WhhVDaq>nD+Yinaz<i<C%2xriL%tgM0g--#y!PU(
zL;K)she%f=r*r^mBy5aAx*4epX*ALpq+3LO8}eI`!jUFlmb3D9<l~UWBmF_-e?tC8
zq)4PcBgykcE(e(vB}fKDxo+ss<DX*v`8obZEO|eNkNE}rv{HuuOuLUm@pE>*{{@`Z
z<Dav0^)LAE^9%ma8uxQ{Zv6#2hp|ovqK@Zg`Gq*pIyT;u40$Z<{9J$E{e}KM@eBUg
zf1zJ5{eqpXzu@PmzrasqxW7_5pLfD0`qCh9wnzgD7x-f4r@zum3o15^f4{(c$~d0E
zGv!&FzXNd=JKe<rtHAg8alHRFcf4Qldp-C^1mESspMcXZ+3)k<FB14U5Bv>bC*Q-)
zHo@=r;9LB;{ZJ47?Sik|qOeWtSFGSWJ@};}KIcT-u19=ndtTVFde~Vl>_mCk`Bv~9
z9{ei=U-97ID{!j^9tnpjZlNA_@&z95fxji}I6UlFafw5A!aeZM(6=B3^CXLPQ_-#;
z3RR9Dna}a5JX015yv+l@RK%^b%-zmB!EX?J942g|^C`94lfh-U$F>r|?-l&v0uL8>
zs)wB)p(i`xbbw>i{JcoSC+ZfCYvcY$fjd3$coCmukNCVG_{kppwSu4O!EYCMu?POW
zu#@UxXNt=G&lmmDj??>u9mlQi{!apd#?j?rr$FHO9{AhBPN;{Soq}KN!S}_2&gO{+
zKT+UO9{6Elr_{qvso*<3_%8{5l?Ojm<#BL&j<4W1c<}cL`^g^m?-zKg2VR6Q(md(*
zm~VRx9A_7};N<ggE9E#TP#g|=*y$8_mj`}K;D<f%K7k+cz(az${i7aus=&KF@EU=i
z^uP}ayvGB7T;QiY@N)w1^}s`o+@Hf9^Y&{TuoQ<b4}3Ci!U%8m7?<(lyl~KiUnux(
z9{hL2c)2|Mc^u<F_B%c748cjC@O%&c#lrqx5Bs+YfA)IVNfLI7J?z{e>~wnAc}B#c
z%EQhpf`8D1zfRaM^{_wG!2K`vu!9FPJkB0=I)t4n4?C^GAM5RW!fNZ=R?**{E_eK>
zu+!jSr_RXRJ?G)iRN>EI4}T7VPsdSdF1Mek^87tvztzKj1JpELeIE9g3HwJp>|X|Z
zY&~~zyJL8!WN7U!<iHmK_ZP=kn@78;2HtM3N4uAZc8_|rn<DJ&^{{iRi1TR=J9i2@
z-5z$X7wz_V@C(@WD(16?{pWGsr1+fl;OB{P?DoKq3;RbsaU-{L!~?g9c;;vGfW0Z=
zHVO}2*tmH3;}ZBe54>66{T_Iiz*{}qeMaC79(cdN+dS|ujC|a?^0<R`_)XiV=r7HO
z`~_Ur9(led;(yX3{)QpEUp*f9PIjH|uQ<d4|0Or4+@s-3IH2Lvgq_nKcHY$Z1)Tp6
z;b*bn_j>Sm3%t(*ACGxO{pwlD?c6TxJP48S-VBc8u@u|x5_Zmc*vS)i+KM=DF4iO3
zA~3JWPPe#@9nE3obz!I9!_FkpuOo{%ajj_AB(B31k8z(aaH|JS*UQvysyKe%5$&$j
z?A*=)ZM{-TN-AsX>q?rM%Nm<YN|chE!n~65ipGk{>Zaz3#=^Xen)<qm!m@QW6<qi8
znv(UcWhEP`>&j}Xw^t~&_2sc8CH3VMB~A4;TPhlPR$kUzrj#_cGVP3Y2}z0N>*C_p
z#h1mE$0a5t#Fiz*$Hi?ZOIaURR<R-Or$#C(nk7TLjq+{f&0gD#jTc7Om({E<*;uw-
zh}Ct~O|qt@N)xjMoLg2YN(E)jTK_Uz8U<L}TwSZ7d07I?sVmp2lb0nmP4(qvLdSvX
zbuy+oC5T=JDn%(MnP1VIkx^Dt!-lC~LkWuY>z7rmZ?11tAT~7CH$%W!H`O&XR@XIe
zC|O^%i8QsFGOEfNm8DsQB})qnau&@m$;vJCKA<zarp8liEofc$q`7SUCUT=>Ls@l=
z7D&vQnvyMLjZIz~PV}0Y>@_nk)@y34(|c-)_f+qh-uEg+sjaAOs__1JiF4ARHC0#E
zdF@WzuW-lvaKy!7jxQ=Js3@!P-jjImsR>F=eI=U6X<C#<U9PDttF0}IZBQy&8@#vT
z{Y;GW{<!gedL?-6X{`62#{P;UA<k<DlDuZd{tBx10f_fLs`1_jAl~}`#CtzE<9|iB
zydT$i?`L4V_d^ix{m~!q{n4M`{b(h4KUxXik5+>Bqm|(OXeD?*S_$5dR)Y7VmEiqo
zC3rttiQbP^qW7bf=>2FVdOuo;-j7zI_oJ2Q{b(h6KU#_2k5;1hqcy|((VF4?XwC3`
zv}SleS~I*Ktr^~r)(r1QYliouHN*SSn&JIuC3!zuN#2iElJ}#P<o#$Rc|Tf7-j7z2
z_oJ2M{b(h5KUzuNk5;nxqm}IaXeE0;TFKszR<ie_mF)d!C3`<w$=;7vviGBv?EPq^
zct2Vx-j7y__oJ2K{b;3lKUyi?k5-EJqm|<QXr*{RTCq;=#~{|}{UF3Ty&r{Gr}x7U
z>-2scVx8U(M6A>Mk%)DAKNPXfUo#foA3CwW<{`+hxucBz72}47d)_`Wi;evi<Mu0V
zmgADdW2~|!dT5f}QrCos3V6EM>?~oAD<Kt#2dmQKI>yim+Vj<d`j+PEx(c$;(o~VT
zt*)%Ldi~OEwd?9@7*<zPUskTw+E7y6w5X-V`==u_l$z>wwJ}ZgF|1U85+Un3GCwyb
zJ)<NpCN4$Mit#ZCdU>!E7n6v`$twOCkm$Fb+P}dvX|#VP{0>zsU&yrx&=Ywoe}i`h
zcpa;RBefs64sW20%FtmDuu*jFD3hMH=CEqKnKvup0t>0G#uI*%5-sGG5u`ULGemvu
zciY-?|0CM_4e@(!4Q~)Q`@tpJ(eOxmk%+Cki)A{l%k+a1Z1mbfyvkv+gMPG$4V_`X
zUu6Pzp%Nagfyl!pJX6A>Bz&HPJ0+Zc+p2BJ624mlk*7*H{r*<lToUfmK;-!nzDdH1
zC0xAXVcJp&*M6H!iYf_Dlk7A|_znqgmGBw~Z<Fv^3EwN>r4rsL;dK&zP{Magc$b8i
zN%&z2*M2ih>LU`qOX44u@MH<^mT>XvjcHFxIL#Gp>yhx88i@R~gkLM+y%K(vg!f7K
zJ_$c3;a5m_zl6t1xN;*G$j3+tw@NtuMp@e&5+0+0$U`N3w1kIC_+=6vCE*uKxKqNj
zB|KTeCrWs#gin`nmxLEfc)o<Mlkj2*$3mlTr4pX5qnQ5^E?y-wZG(ibm-wv`UN7No
z5<W`8_e!`}s+hJ@!pkN8K?&b1;aw7rSGD?fSi&216!Txg+a&y`gcnM9w}jL0{k83+
zg!^kC@*WA-)_zi-mhe3izgNPqmhe6a-z(wgBs@yO`z74%$J1hR)BNv}_*MxYE8z|a
z50&sx3D1%6a0#C*;ZYLqlyIkn&zA6H3D1}CR0*eR7j1J%IQ?yew&hEBg$5!omhdbI
zFO~2a5?&?YizK{3!pBH>tAuZm@HPn_A>n%^yjj9KCA?C?4@!88gm+2!770Hr;d3PX
zh=dnO_)!VJS;D&|yj{XiO89Riyhp<0CH%C6-yq?=5<W%3`y_mvgrAe}@e<xI;fp0)
z5uKs=f2o99C48!cJ0yIegojGFcF9HRa0%Zk@uMVsu7o=!JWj%sC0zWZlW9{Wyh`G`
zBs^Tg^Cf(ZgcnQrHVH44@M;OKlJJcZ-XP&?CA?L_w@Y}NgfEluy%PRA3GbBfl@fkX
z!o^RCn6^v8FOv9&B|KHak4X465`I*|FPHFc3BOXpPfB>Mg!f2ziiDq*aIsV~ZLfq+
zllXlSK0(6IN%(3B@0aih2@mb!f)6sf*4MW1E?y$spn=Gvy0}2Nkr&zC*~JCIO}xnV
z$z5C^+{}w?pW4L*!hg?;Z13vg0^yXjEx(JG2(xM+^5QNo5bn#1Y+u^N1;XvT$o5rT
zTp-+!7umj{iwlGY@FLr{c5#95Kwf0~wk|FZ9>j}mzqgAEgyW~d`qtUSi-hARd-`^;
zix&wWs-rj;OSrh@W7@+KK3w7-k?=_pepJG*m+%Y43lX>wfx!s;Ct%FCuH9#?u6^bQ
zv2MEdbT=CYdR)67w;pGtfy8%96y^Lt<a@{h#wx^T=iBpV-WeDe=wvc2py$uLpv$yy
zo<H-bF4KZ}{>(kPObg}tGq>onnaKxqnJ!AspSeVrX^}X8X0tBSLU{g6nJ(LyyhN92
zK|FsZO_ymQJbxxum+6M+{F!mOObg=qGl9BHH!kPTsJcuG<M}h6|ETqsZeY-VU8V&R
z{nuq$7}0-SrUeoG*JWA=(SJ=wbZhY}B4BK$!r}`#5o00CZ>@2c*WemM))#rmc^-0>
zhdjqaPV|tcdB~9-@@NlvsE6$5A)7tqA3XZ=rHA~fhx~zu>=}<&JoL|c$WM644{5Ti
zW6q-}yE@E&r!1wvc_<ErxwyolL%}xC8!)yFqoo`dIrs+{Kk-V+CjaPaKj(V%-}79L
z_8VR5ldczjYz~2eEMdSp(7S<+lUAROm$o^1sH?P0EpqLivw*JVT<sq;4{`0A69c^O
zn;!-S`pPl+o-{{7R@b1e-ageovlU9=56NinS!%C6`8YcsAwNX&K84F^MQMdauJ$jN
zy4qjDu^#GLKi`zQZ`$?5=?H1d+1G-)CtMw-eAlDBmYnv-T}I1^Hu@zw9oYLiUC3Z_
z9PTFv969^4lnq${pM7;cz`;vzkl4K;;IsDu<ymu$;mCJ_cumA!)Q3<Tn%mxsJm`F8
zfMcB%rIijUZV#pX#&Fs%hyvW`q+&rb)pep(NK@iyx6(xWbX1+PZpitxfE7>D?rGZf
zpAT4Z_IyBKW`q?!z>k=o?Liw{0lQvs1!Vf|NK_OEbuXL`So;+8FK);Th<O);+<;y0
z5MdP&as%q#LGv)-+Lur6toR1IVpLjjocu<`H>l!E=L1&vP@}J6xAuFq8oVJbAnKL#
z0mbCoiWdN{J^|8WMEU{n2`a!wU<UNX1rF$5Bd7<^Qb>;UP>98ka*vqHrxLhFWNhtI
zu6?PF4FOSb#o_k=G%)ZKT6zxYS(G0|{y3Qr-L0rtwnJB~Nn4#pQ|2<np06huyT9UD
zm}{SdM&8viqzz;3qOo=v4Sb+!#YRhH<oJdSF<R{n+I*^|`s)YCM2G1Sr1@^uX@N(R
zV*N@ha;-m+R+!WN1!mnVG{u&>I!0k~y=S3Q4W`~0m}Ap!pT(t~<}T18t<Sap!ab3g
zYnWFE|0x<<moe0}Z?~9Kp@>L6#%niC6pZ?tm>Vt=#WpvB=1HoR%9#-kJ{wMBu+4S2
z^`_Zy8nfYP)OnffdPr9wbTl$~2=m&<P{^{<sZft;DqElB2+{H9DU+-6@$M;m@-aSs
z%m0t@r4!s>e8GPU@%cZD2b};1`8imk+Fy0GKZSrelrw++W?%q|(%}5aY42y30PU~2
z+Rx{<KiPK$7_r^#B7m<iI54=8Tw!FN^-MaMVe=zxdD=2sWB7pouoT9U;p06xr@8hW
zig4ldnY-`Kh<ujc7g5aeV-ck+zb2vzIW4xCx$OdM|IpR`S#G<aw)3j4c44YrRN4J_
zG9Q?<wP_{VI_v64oQ(SbEYq`)<+c|_MCG>UMug|KXGVm&P;$6{S^GlZeeOQe?t|9V
zZZ^W(+<im>;BY}@J+lextaLO4j7?S23LmCL@e#TNC~{qLHrb>o<*t3E$S44=_RAuo
z0K|6Zw7=iC1WjOF?F*x(I?Q)OE6Rfr(B`&(<Z6H7?_}m-l4HC3vWe3EQQuYl2>e&*
zysQ1OzDFS-jCbI+9>@S3KhgIrVeKbqjoxQIT0{rQBXscfg)lO2zT7pH`_tm@y5w>4
z5X&!?h2wqCf#udAHlJZ?^iq7Pr4UgIoJcmQsdeB|J2*37e!__X$DZz%fV9<T`f=f+
z&7Vi88-3S(P5$?y37Ri76TVGre>tc9KkZK;MqQv>*4-QvFgY_KJYaHegfn1rVMJ=c
z<kgrHlglv?Cf7tX1Wdj!q7BS|$;To(kxNm<sKj>XcH~B!b9G!6aRkZ}nGq%@au<fx
zN(1cJ^J4STjs@_d=V4Ohb!0web#)wL>`x}>^S%A^-A_7l)l*nJuv~Uyp$JUv&rRuW
z9(Ku(ccb2xf4SPfaqT`~{p2mSo}urj3h@b>_IGmI|CQVR8e&6}WH7|O2yquGQh4hT
z2Dk9?^&+SJJfd(Iq@4Ee+F$+O55-r_`5ntB6c36}yf_$&aDF|~{?zVA5eJ+~90rd#
ze51tyHwJPXs+Z72dNH^C?Vkj}iXc$`j`!7k<rV_IK43n4lh*!BR{LY@G6Yv4mz<?h
zL?mKqQeZyh(tIeS`LLSiLpjZdJ0sdaqIuED@?#NQ$lc~e6dh{4G%I?b!mJ3P9@DJA
zB?o3(7N$jRT1Nqbagqfir(?$p2*!Pk|H;Ts5eKaGQXJ?qaM&e!9L~FTKPk<JGm8}^
ztG&1VbyxexneDIUwm(1MNZb9rp=HvZH?~dq`XwB!vA8;-3li7*H*orU%L<?R)_pNB
zfK~{_42l|@K|>HciUuua$NQ#!=@vjbo{HMBre1RPC^-Kk5&PM2xe!K**q%?DtKg>J
zLM&IYZ+wy_t~1*o+uiMO84DxAK?<R0p@%rp_D$v8LJ_CA7jZRm66e*Ywi}wK?LMxi
zywyAroei18rqn-ZRSk7@q+8hleexi+^WIdtlfj9?Ehe;ZQGape^XRHI$NPq&75(~R
zY5RZv=Oz#i@J<XUPNw8d5xl{O;R<K=a@V6DS#S;QGA8f4;u$P+SpCMiyBV<ON%$y+
zkGXoBV%rUnM~BY*tgl*h(b`PBHqj<3mee=UEP~fuM<<_foO?oIyJNBLH%Ek0KtG9h
zJ6<tQi-UG|t9u4ohHVI#-m}Hg{upSd*fM<H6)@$gl;h1K#ngQor*k%SpBlWP+cOZb
zCkdyujtrIRQI(TtCIaZY_dF&Zzux1E)qj7TqF~U>&h5yIIOn!xBMuAlXgR_483Lf}
z8E9TZgj5P0u2NE+y9d<f1rL%;M=M!5N~fAn#(BoW;fr{j4=BOBi+VKWMDtM5`?s{-
zze(q98aQdy9$c{g@%c}VH?dg%Z}%UM$NvZWzvln7|L1<v|Fl9(jR9J9aG>ozn~K4z
zg|mD~!)TeO<6?(Ra9llwV@Bd8Bye?vlMbL!P+=f?F*$K#w)S0U$|J&?h9WPFaP&R<
z8CwKV1PLCq+eq_AxKu-h^G!p6Vv1v7arE8FsVKJp3pY9^UF&fxVY(kFo86`wO%a^w
z+UF|8y$h{cLvZo#nEe9H1g1;Ien&)T-~F1cWL#&O*u<xm7`G5;D!K1>lDZ`bCTi#o
zVs#B8urlr`oTx^3xI<I8x9>FhQ9FI#{u==dh@zIUwo~K~5p)o;pFgMei~fi9Jqt1v
zSO!t$B#7u9<eG>mRvp2_9Sho~d%I$soLAGRKEd3y_9?tFRtsF3r><H<>a^A5#F<E_
z^z|`q8NB$E-}SHmj1Qs{cU^}Uh+(5K^V^6kJs3uFd3XQ-E3cC!^17I5*8USi(g()s
zh){j}{51dF#%~fT{AB#z{giceaQrZMxg8PNIr}U*kG^kVeR3?=ml<A)HciwJ&H#Nw
z&;S<NVZc^XAQ00i^b<F1@WV<EAFw1-ttz?JtGunDH$}2#I*iw&C{Q1oY<)!_pnHRZ
z^y%OrRihH4@ktuh7sajN<^WYr>Vt%N%<N*;h(UTr8YH?2$H?@pIxD(OgXGi(DQ!*K
zS_<!(9Z*Y`i2VAfD7RzU{t>j&CEkiGtuPn2KOK<|pj_nYnC2Xb?;p|~THjjKcI_Mb
z_Iz48J_3C&)oM5NWiiF*hqE=srA%S$i{>>tax}$mQ6rcuuF-336^d^^VQtv;Hfn_8
z9j@?YranXCl|pfnD^%T7sZiX_6+KzxB-JPsitD)IX4K${6+*F-E0&=Ru2?D*<y>*2
zOH(W+MaQ(GLZ8X?d(ae*OomWQ<%$qg<cj%15yTZuy4UlB;+v0I0OEAT458@diZDcm
zI~OGsC%D3?cW0_l+{G0>KdEsoS3IdJMu{3bxFSzigb2lYu1L@W;3pIdxFTK;GQI)=
z{|tSxT;Z#G^<5ZOgmT3@y5cLL7&yaveo}Az6QTH!E8fo3dj76ZJj)fA>xzF0#eH0{
zM)&F&p}2u7qV>kRh2m1K2-X|FS178vVxwN;4?>a86)Q3{uMP@D5?AEwUR@y+W4L0h
z?$u7AFmuI^x}r`f&i1j8e2}iSQ7#lOam8RimJ7w-xnipBRgO^H!WAQQujUBFUanB}
z#-|HKJy#sn6=Q^A8CN`@do^4r=5WPxdgBhEh~SF9>Wv$O!p;@f=xux#%4f;>k66$5
z>1}*26t8o|t-9i4q3Gs{cD?a`3B{kdqF8VIRiU_wD<<iUpAw1|u9%^>@t9Do=88Fb
z8}|uC23MH$HvTLWQ@G+2UC~8~j%j}pdS9;pE=?QZeM0g1hb#d9)P1;IC{A<5>!`t(
zlxCrLge#7q23M2`#UHuiVtCCJi-e+sE9&4JSIiTNX0EtaZzEPHR&hnS-o^x>Nau=p
zy^WzlF_|l-;98ouVG#-kSET4|^yAGstsGx}z<O@g+c+Z>|Kf_T^%}1U#S>idsvfH+
zgyK%Fcxav$-6KNrTdsH$Uh&2B7NOYA75_jRTyeEflySwSdK(uBg^Me;>1|XC#dNM{
z)7w}s6vMcpQg0($D1LaK_55GDB0(tL=Zadrjj=-UG*?9EZ442Ld$}T5H}xHU&p^?=
zo-01q6`u%28&{l}tM&XXp{V4F(|Y623dJI>ctmgfm{25g#TLDdJA`5sSG4GD+#nQ2
zt~jK((JmApzsEwdNjKFZ6fbbacHPu^p*YGFD|Ay!gyLqdNY>p=6^hHb;xk<lBNVk<
zarqoABw<2P$Q66^#w|iIiz{mM#=jiO!#<uXZqVEKKqzcn@q4|E*M#ESyR7F2^cp9G
z;x(=~qSv6Wq0-Pk%oSI|D_p|ig6DRj_ybogf>&JeTcOy`6^UqrE9e85WU7fPGFcmj
zfnZ$r^|8+4ah?7~LVrK=SL*KWWTn~XIKMwW6O3Mo_w{Fn!<O{Eun4w}k*$ttG1OW|
z;!xJ;XMK0Q!+Kn)M>PbBzJnaSMGt8Jp_g(rLq|=7)^K#E9-43Pvb%2?N3XyyTlq14
zp3qc|J~&H5KOuAqM|bN!d`M^rNB8J<KOpqSUgq!bbhH*{wZ1-%4jxA@6Zb`q9@krY
zlF(xuU8AFaC-jdT{Y>}wZbCabYSUZ0jnM5J-L0e76I#yEXLWQxp^G^B(M;CkHHds0
z;4?{p-NvW=hrSHzaQkOjcp7xd)iJa%8IR+>>f1>M==oybZ&6xvW*S)1?Swub^3Wf;
zbx6lg8NlaH^qK#7vVRa&agT);&gN^mCwPmXD)?~vnP=Y~7~t;&=r<vlXvYCx*e*oi
zLIf^E;6em0MBqXME=1r$1TIA2LIf^E;6eoc{~rN-YXsk_jcdTCYAZ?xzxNTRNchiw
z)ON5|#g?*~qVf$2zH402x|BYYTq0_VOxP{iP*>7WkFWMt;Ip`mtp&At^p)J)>U9wC
z;a^=6E+I=}-yg|uZlq753^v%<x}*_xa+)$)8XGI>nlqIA#)|dTO%+R9YMV-$T53xg
zWkfF1Cxpx0)f|HJZalI=7;m>^J#R_sLP>SGvcA3+fvd<aqYo4>qmR0_G~nyd6(vAR
z8pv&!$S-TmsNVuZxi!_BGRj)kS8dB|T3?3mP;=Y(+KHPEUpy{ruCA|hLxnJ(zUP|R
zNZ)SebvV#eRo38!2<`HU>UH><iryS~pkq=Ufr(m0b=4a(*Hz#n*SrlTm(?^>$$-{=
zMSX2%gCvMzc4j#~ea-7rIaVu+Jgc=q?gm7m$f&Ar+KdXaD9VMIP33yAwxUdnP~jlG
zQ0UX)eI|>F=ENYPxk8AbFIrb!S6`dgl3Cm0fw+m8m^?IDDU7^|5Ynl-tYv^u=GRv@
zW!99{uESTed0j#@Gxd$NgWXz0lGHPjl_q?XyMg$005q1>mDh6&-s7{r)y?$HU+meZ
zej73?8X<43Xk?<&($Jt48L&C0Zc7=bQc+YI1RHd4v(PCGRn<yk^;SNUP-?;KK-P?7
zyc$O{9INJYrJ6U{SY3%Z0hCzu-QmXS^)#lGv&oM_ort$yM1esBP@fxH^#)scgRRtH
zD{rusH`po~Y~>BM(wt!pwu%N@wFX-U8x%!QMT4!Z!J6v2j9M%tJWm&S+E2cNBtGZ<
z6I=k&7ueYs-TBAeaj>s18T=SK69>QA{*#ZYYkIdYtz+<i`Vo3=Ui&)x|H#MPwQsq%
z)U6qxo|cxDx}>1lm6m}LZ5!*hW4|S7LRsPDwZ%&!^3rl<Czs7xKBIL-`kJ!R%*3R$
zip{OLi__v}WHeP)FIek}b0*E2k{B5|rFQDB=;VaBsjH^Uo?W@bIeBj4`s(G?scWmF
z3wKRUcP6E;NKD$eH8(o8c0qG=T1iGm*~W;J*ts|+=T3JvCFNA-M-|OkvvF!c%ADHb
z9gE^u#x+MqmgF}tk6ShKqLovfSu>W^XK#y&xF|8UXld1^Wy@1{#y3XgG)|wmCBZdy
zTgjS+lGwy0bp>nJ&aGNDYuxz6#@q#4Q)9O$k6#><x4CN5#JuSf<F?POb8cvwvTkPc
z_^ho{E4LR#Pp&O2&s)4?Q%!ur%EY?HEe$Q>mrmX^Yx&foHDx7}8(TIkh>2U7J~?_}
z?gHnQs4b}(<6816X4gBnS7&9WEM1YeV8z-cHSy_@k<*jo)=w<1%`e}!ab<3_vtiEG
zxeF84&o3HR;EZi;n6tfX(dMes=v`S&Q^x1cjIP?Ap53x#UY;v!Q}gt^#LC&@#?6b~
zRyQkSQ9^4$OG|aGGbgV#qkLoS(uIjh+2dwzoSKr@Qne+oWL|#eOy~HNJlDpRMeB0b
z)kme4FKZl^*b*68T~nI2vu4|@36YUSvl9w4H;k{0jI5hm-I%mFBPL-@-T3S&Wm(%|
zQ=6jd>$hyHn$=L55>r<<BQtU8+@us|e$mdANvWCHxsxX*OrJ1kNBNrh$>Y})$G2pS
zn;H|pY(-pa#3ppICNU|aerfi~EtNB7)=W&yT(e`>)RMT|9r^QH(kC>QFP<A;)pXI?
zsabj1nD(_zi=vB)opl>i7Zxu|SvzCiyvdbIrk7MM-Ila=?)bH{>#N2uYn?N5d+NHR
z<qLPMDxW#IWzPD>%$+sSE5|QNX`GNAS6O$_{OrXk^Jk~*h$&0iwPwP&g#7Vqv(`t|
zXHKbFIkj>@*^KhIy7FC(i{@m<ZCMpPWkJE*#R;=g)|M`AshbhExG}e3%7(@z%WEs6
zBO~)K%9@=&ck+_FRVB+NPAFLATsCoCYV6vHb(@-YZJvVO<!s%M6F1#CV|sm6(o$DJ
z>BiR9@j2=1k`}F>m6ujKK5lVZ#kl62w250cyJqJlXGX7VD$UEt+p?rIGO}RBju|!E
zFDjT&yQ}1)Mayd%>lRKfTeBj4*Mg}N7DGRMQ&xFZ)$*i^)~>Hwzbj?woU9$QX5-6g
zTazZI=WpG#rg~>(cG7~C%cCZ=Zpo<3**JM_Lyjw{Vf*Zgl?ijAmQ7i)z9}m*a{IV#
zRjE-aTjs38go~e*Iy1hsVs845#-jN2#7#NlSHaAxw6r>BjB`$QRd&|O!pZf88y2-p
zne4i#ET(MEg4RtHGbXz7m+zReY{SlkEz2V#7pK%G%t%Ntn4g#FD%-d*x2B*rIz4qm
za?SMFcx8=^FG14kgexBDf42}9$Vc!Zg|>6^f)-D$zs{WAt&TL#qW@Dk66oP41_ts`
zww@RmXhoikyc4<W)WE=D<i*I3A+JK-L;2GK149g`hddnlY2*pWk3KUnP=ov=@;1Ve
zUxU04IsHG({m73Yw>}Fy$di%3kGu+bKk~iEhZtcOc?9y4$di#DeQsdjGUR)o9~d}-
z{50|&<b6E@1N8sR^t~`J;4tAE0Waa@4D#fc2L_svAADtC;2`C%4h-B!`D>_$y!iEj
z0h1Yiya7IPdVhTx@+#y9k+&kh6L~0pnQ#($G4l73(}veYY*V)vC~B)i9XZ6uO0kLn
zItnTO@qvL-$cn!s+dp(cfPJgAO_?`h=9Kt|@l=uOXCm!I%;~4{+D7<lq%Pnz7TV@O
zp&98Q#w0+1gzZJTM#AuVN;!no35@+;Gui%*4nw+s=ze3GfB5C5O#i6e=K21q#<x`e
zC?IM6q3QmPbbl-AkiFALg&1G_5f<ABJB!o?48O@{TLSIj+aX|YQkk`pMQvnK8!l=i
z*WYQl*S!g{m5h3sPYn!+FtRptfu(w26ua5LY-ArB)olir?1AkCmf%tM5HKe&&0nf}
z1X!*IM*lZz6)^lzfNd1})4-a6g%h&S-%*IjQ2lejcK{zlI5zzJOEJ-01@;aBh{4qq
z=c|nA{^1>_bpNRR=1hO*<(5qU<lR1*{tbo=zW&K+{?0W2C@{d}ad+9-{|lWDThTav
z9wS@O>^Cg*553$t-@nIdxYgG`G?NWdnw{p>4sd(G%_nXiaben;XI~6o!vP(FUcoWO
z{%<Atwck+Se?&FB<11|7XD3QGIA_860%c_pse54~%}!(cCfUa^mxw-7zxEpzkS`e9
z-KHG>9@Xft`m&jmCJcfWg%m=^B#u$`e?4XSJ1#e5_=oPMKB$Is=mVfMY9k!nRB(^s
z_+|eu5fr-(IsV<MVLyD#A`baj4o*Lg?JrTrSir~knUADvgYHcn@18m*bcdkpesN%c
z{a;SZf6UFxjTz*BwtuPdI#u&No6Y-F)S-1G7wZN4f16O}DgzGN4jRk-CXCqSG?u$9
zRC|lc#t_F282Yi33>zU>Q;uS-VgF|nY;+j%>6lok9}}=)^!0T=C{T~)S3c?<!%N@M
zDAPt*Ik1z!v}1t!-3sh+VC?@fWidAHRnZI5T?3s9FQVD8%XHPOf22DC9sMRC17#(b
z#-@?!NW8FJh`@yiT!_Gh2waH3g$P`Tz=a6>??*uUUbgmqZ0&p4-nLAa=Csi_zG>5B
zBYu7>%k+(J+J^897v8);pIV|VWF61(drh`!-}N3TWV$b;4S&nRw*KpRPWK_SY5FgJ
z9H=Mx9PS-x)9&w1iuNrWR*v}a0(}{SwsfJlbCF%gQW2NqeCy(wFsrb8T0nHKMw>9F
zX!mga_#vf?ABo|z=BG;pLcXRZ20o?0{LdHs0U>Mc;;%EA-gg?ehYv+C8JE=;whIxs
z5P=I3_-{qP`TL)*?o8!)s>t(2UMlhik++GwQ{-JDKO*vOk@twaSLEkJuB0h!bBH`#
z<W7;NiacNBr6O+-d3P7DKR+X5W_VQ5x|X`;mhjlvn1mSTwAdCV#_o!Xi*Y7Q7O)bH
z8U|L2FV5ZU*#!Q?R(r<AE$aLGa_O|9GV<I+oIpm;7P_CS8Dce!2~d=ICNqBNpxCNa
zoKs#H?m~?atItVjLagQ$I2j--q)fG*#abR>@Ocf*hnRexC5meKHv`S|w>GI(pZ%yF
zcPZIWEnh-2i~!g>4zjHb=R8x?$iDzeTa5C3C`@=9J4*4?GER5v;eM!D`3B|5VK@nI
zB6RX1oF1uUn*1iPHk7QB?RW%4C5K|Gf;CG*=*^HqOs3hO45QOj#%pkVh%<=FnLZ{m
z;4{IWp0Fky@Q08v2~U1h#WvJLis_@F2yqy<ppvZ^4H~Z=u^6lnhbav1`}`K=IC>nH
z^*lx;+3MqX7eY{zVzt_+%P1~VtTvxS6s<w^iq)4DhXmCrRy!*?g6eV=>kw8Bu}r`S
zS^a#;NT}sO1l8(q{Rl$1<vGN_8o-38pn5dmU}dM`L+!BYHuAw5Xskkosa@vbzoQ7y
z-%?w{m|n3QLlmqdm|(KdU%OgIGQn#3J$hpu#e^Z2kI0MBz7!&dr5Dz$W0(*c<Z>$3
zv8)_!d6_&4XF`<aZeopN^_`Y>Jdm@FXF`HyHuYix6Ot_fxJ0r>GGTU5UBza_I*|#f
zmJbnc>m(**S{{Vw)+i>pEPD`p>trkSCf8JgYn&;p=gB71d$^#QYCjA`s>$>!3`8@*
zWiqV>ZJPD3(BzvO_yNas=4LU<af*|bOHHOgkYZUQRVLH7s2<0J8j~p!(THb4gUQr3
z9<zc8%_h@Vh+rZUT1}>lV1EX))@CxD2~w0K*8X0T=?F$5nF*aH6aBSM3KI^RO!uIj
znM~+1nO=f{S*$mQO{NIU-Pz3a5tHc}Tx-o?rjMFTIf%eqrs+1Bh9m6rnDDsC^gY6!
z%7l|9(}%E@#)M}~rUU3gIum+Krn@0zFyUpB={ZdEOeUN*nc@+HEGE2ZGNpo+&4gZ)
zX*X!|nee{Jv>j95#e_bSsRGA!4inCrOktocV2++MnT+syA=C7mOgj*&T&7XXrtdLS
zc}!z9o3bD*V!{x!NyTJY%mjzo^aq40p9vvmQ!U(C!Ui$aY<dB83YfKUvtxa>vXm*J
z%#NyTrI6uHvuPZ{Uc?m1W>YW1v5aX_&8GdZwVY{OX44pqK{3<hn@x)lffY<sY&Olt
z{9MU2rPNBCvWjV{sF^rrHPbYhO--=2hG|;SH4MmFrfGwlnC&G@v)612Mr=x%rqgVC
z9PN~`uHiM=g#9?SXyQh@Q25BG_{1O&Q4I~iY%gFYSl1cpvVCf|q2f~<jzQ}R73&6O
zOR?-AeI=VWR?9(z%v!|+hvgDju~su7)G`q*S~s$}7H&C3hv6osiL#s|p@uo*w9s2a
zYb`UKZ26oPg*qmrS|-qPQO^XIr5&BIHn2MRmR;27W~M2&JVD2PBWtJB@+@L&ZDK-|
zWgVuCwV4SGmRwpdT3DS{%TW@xFio4~a@>zuw=!X`Wfo#<ZDm4dP<6Fp-NwoXE%#8H
z+nLa1*+xs<MNBwsd5>J%!Gt502J&Vn6OLMXshM3&=nkS^u~;u=!bwX9Y1(YGxb_4!
zu2-y=u-Kfo{78%6rHs{U8Arx;Gfkhxhnm^TB5=-<Mp0{L!WWiVq`8s_{gz%@ul6zF
zdy@^C{mhu6I;vZh4ra@$I?A^wS25h7n%2OVtC<k0nw~<>IvFcmHO)jY4lp4~b>L@p
zzhPykYBHnw-?BE6VYghlhLvZlrpM9xwM<A=9r*c`avc*giPft7jtMT+^hP*di?XrE
zS55T3_If^;s_8jse$Rwb)l>^>H!z_}HQfu^Atp4aj%^#28(Fzkb!@FtZf1ttRFj__
z?_rs6nd(?y+pOHeguTSdQEp|zeiEF@ZA|D?P0^Uow=>}y)wFNCqWpmg2UXKLwEssY
z98yhlVfs%@=u%Do2-9IE+^#w*8XJ{AGvTmmN=3-;V8WfMY1CLn`3no}5!G}vrv05v
zII5b~qS3!HS~r4@aNfm)$5qpHX!LF-oK#H-7=a^9ct$l%LZkOG13l<mwsIdUpH|tz
zb3YS$Rkqd~WkR27x(rS~$O3gvHThzF{5uo+A>iQn2cszllNyXyS*#hW!DPiH$U{sR
zVlY`?;9({>45nx}@Cd6DYG7xaN0|_AU}v1im=I-PXPo0qa2nVd=W!+^8%)P>NI%IY
zQL4f8HLRUrf(ruLImv{4gXw(;r<hP|FqI$#Pcfm?VDd$9o@PRo!9@SP*|SV&FqkGn
z^Bfae4W<EHc0A97HiM}fQ>uq~u-9M;!4mrd({vh4GcfaCWSWBp(=D*}5{pik!E^;2
zeVKLpu)%Z`!}AIg?u399<~6p&9x<5iM;A^r14j*tt%4TD*Nxqns8f?|Z(Io*mPcr<
zdvhd}m7w|t#roFpt5LRE{)ui`|HUI~iKdn6Z6<_R#v}gLUM7TER*=nim@vvRKsDcG
zLbxRjZCl@CLWIQ!189Ap2~n0JS{y!LLbPQ8ys>`B1SdXP3*jRsBm~tpD%L($PPPP+
z180~p+oF;KA2T7<k~<#4Crrq+d`;`~SthtFy=3fD7O{NG=d_^xn+e613A7-6#-dzm
zIZ47fR;S7mOSI3KronPK892{0t(MnthPD2O2|Fyb3R%BkLYsxwC+n9?xXf}TjwtI_
zOxSDjp~doRChWJIqy@U437wV)Nce^c*I15^gYd2IS*)f9Eg#c>e8+?ylMl}R-}_uc
zVZ{Q23Gf4}a@uN2M?ijL!kbppB_Iqir+ckKufV`5DjW1Z>(CQ$PF2}_IA<Mt9Fjq0
z%zo?8SJ4Zj%3^{m`YkYGQkhAsZRkH>#muE4wxO{YWeb-awxKC--iJ#ewxQ2c{J70f
z+t3EY%*Lf~+t4Vq=gXN<wxOS*Jv*18Z9|XXfEdC(bJ`RiI*t3>hsiL3?gst<XV}UA
zKdNHCi-P5+8ovQ0ySXGASI=tDBVhPEj@^(L@IOMqKo$9pAOqXcRfJ;ygobE>8uSHK
z`bMi11kXJxDfSQX_|F=l291T%CjvWWpZ?g|gUakH_9z<TNY1y9rwjN=s$(@(F2iob
zX1yZ!<bFD=CaaEJ#Mn)YTXjZ^o7V<XC_Y0lE9|e6|5H@QeSmx(#BRh(*!_H8C(tEE
zv>Mb)+z+s09r#A)vqRl}6QXaOrUqExkB=R@Ar!r_DDb#()_Zh&6r%=26Dw9@6%vaD
zoBG80Oyrfe(h)gb4JxBb8>mtYOkbu~A{MVsnBubuJKJ_lZrh}wYpD+1ZB1(<G#e$`
z<YB}}3{Ut7lI1KWgDo+T{m3I|Lra}vn-SO!h1K!^j!;`tAbH}j6u}8w@>r@8s`%VY
zc9x)BTZ-{7C=WYJz!|DAi<O7{03b1R?i`xvtciJY6RC0&X}k%=wh_mmE#2skI>V@<
zEdX=DmKjFnVT%YjKuyjM+l2CrPAbPCV7{*53o)p+sc6!7BkP>)V%#+NJ`i>dh(6a7
z`#vh&hTVv}urut%PVu3m!sjT8(+ITv4zt4ka=<B+eV)UP4dF*9@aR&A3@lEz6Ns+;
zm7xC+v!9qWB0l)P4frIsy#UAUuLT7I@d?F_>7zBhAiYVX(?Q8Znxm1Hxse1jkq-1X
z0@e|;61!<MCblMwoc+z9Rzi1T#~L|+f`QElatVV(eR?b4Mk3uxB&vG92WiGQkp30)
z7?Dm8iR$*ck%T8bh?xDoptHpMTx0s;UX-`VW-V{{5FYp2-wzlDozF<@Si=cF$DEFX
z-2OpODlxN&Ne0&clqp<z1hdTkiQ^`O-L{t8IqSHHs<vS_;svTkSJ?)3D6t(AJ+MKo
z2W5~)CB?QJd;7l~Z&3^HQVX9_3w#zF1;PHA;~Ucbh}{sHRem@#3w<TlZu_~wFd#l-
zuw(jZq-WF5X2ZG7{<(h&>F1K3!k0Iwe;GFI=L1)eel6)KFqe>?In9DSSRzreeF_)s
z{|W4%I=|8C{DtbU>8sbFBNxms10EpeG3-X1AQrb0MV@^f_!8+(W5-(liu8klW2QL$
zIttRaK?XXH`C!NB+fgvEgN>FOHhBCv(kpU|I`!W!;XuGLZ8xexSCA9*1rg>1eQ4$X
z;skT&f5ZtsLG2gOAs3)J0@1e5FziM+$qUv53k3@bHLKV+Vp+8ZszGyzn~ojxx|I0t
zR6a2FE737~usZZ_G|a<j1c$2Z7_(2MLt&U2a0OY~PnHtVz#U|1@B~4S@bW8M7uiGA
zpa;P4Iffl;`JcpRW9%Lg#hz^kKTHk!i1=qUz8UMR2Om?KoE*VT+rzM^+eZq)N{hlM
zHHiMpVxL&-SfhnMH8MCM6#J+Cs5FMRZ2ydoiLpX>n1pa)`+5q%IHCCpr(-*=YN&Vi
zTWH0=)d~}Cq<Mnt5hgs0(}x`w8%)?up5n&$6gl@Cb|Z$uT?3sY40Iu7xR*-o%7|U*
zuxpsbaL&LkO8!J8b{WC04cMul&g_br{#)nm0W_#KqxE(m1lA$`7uo2@v>6o*Ou%us
zLxBCa31&B{@3hl#Z!@0Lz>fj$va{2cu}=d-aIJN5FfE-nW3L9z1=to$3!}|=S_7*9
zUJ^oUna$XvfxiKGX;?i2Pio*XfV=(Zv}7}OYv3nFT!4hoQf4zA)xb}|+%tsMFPrg*
z20Bpjaz9#HY{tVHI1Atv!e*BSZUDGffCn{@PG{}GbQIf+of>!y;FbP#P}z)oHITjl
zy3bB)l+D<tfo~)J`vuslf&0Mhu+#EnGd5`8N3eO7uOF*krGY^R#nl2V)xcn!`#bG)
zSzt33YhXLX14HNxZ!_j=U<pR^Hv)8NpnV)3r3|5K1)DKd1AD{q(@+5>Yv9AEeysqV
z8h8fabpnjiz-v+ccLEI8z~6#-P=KKtxDm|j1?bSgc>sSeK&u9Z0K7qfiUux2=nnbO
z<%7-Gf2+ba^AvzL2G+9SIH!R>;?D6V0rqL&D>$BR7GSRi{uyAGuz6Yoe}{wVmLWB)
zp&kuv0P|J>p47lhfVYY2-5MAQ@OD3%JeZ#v_yz{<4}y6_1K-4P{zm~G*1#hG|0Jq+
zY2al54~y0hYG5?Ze18_zJ2mh@ER}Z%aIXe_g<$=~pDutfKQ-`afOiIMXQ6A=z^hRG
zuL5k)zz#6)5@3}Et^o6HVY5^Nmw|c2kFMS@KQ(YFz<UIkuYtgndqs7Z296uAD1Y;(
zOFGO?4g3m|_CCQ(*1)F${+9rq8aNk0zF$<2(!hgYJ|J2T*T7nUM@98e4NM34pa2~j
zn1QA8?;&(5!}_U#^D#sIAwWd~1Hn8d!2VmPe^_Zh#7+N0A#`TL`l*5c0Qj%~`!ui<
z;3GrnoQCyN124hS-z~t?8n_nBM+MlUf$;zz6W~b=EXDcexB$B~a0|{zj|=dq1}+5n
zL`V>8{fGumL-i*Gcvu5RfO$fIT^jfqE~8Hdj%C#kYT)AlPYJM71Fr}8lmPc?U>(4x
z1=yy6833OdO6O;ru~h@_#-aVJ02?&$GJwwomoa9Q2Ce}3yZ}ozFa}^xU>#!?YoHO}
z3j)m7z|(jx^`Zb>8u(X$FNIvhs;6q;r2t<RV6p}-2lz_JUdD84U^Kv21sJ7)7aQ=5
zPk`YXI1?3555JaG57oc`h_4IKp@HvXYW-7yRt<a*;2T44Wz`i8yb9o(0_^W%(^E0K
z(E3{fJg0%H!2H*+yIJ)<4Wv`v+XC#>Kr_Hz0iM>tIXHgb8TAl5N%aUYVlysGeV?(@
z+vmUG<iPJjB3_^qVUaJP&k@RYBq=^6s2}k)=D#i9oCx3tU2g~na|rE*(JmZ2+zI1s
zJB&C!IjA_y2cVCQN`X{bNu`k}&3^{QZ8erOl=)o=yEUPQ4ocE_oJ@*L1*D=Dm84l{
zevDd60Nphi8feGvpCa$T(e69SU?J%;BCy-C$d|D)_>MIB65K^Fy})Ail*q?$IQbTu
zSd|Afn@Ys0@$jhDMlQUl8RFKG>CdD~ViZOBkQ7u<BA>?9rQd$lM$!*D91#n6$Q%}_
z;Bgu`8Gm!GR1tHMjz#_(SCW1!3|4~Y>TvK!Fwk$X^ppX{mccVcJ@ij-{O*8v{f4nL
zKIjnpxuCVAc8l_@!Af0?0ETWTVSS=Ht-7jOWY~iLmiC;$y22s!3*kms3bU?59c)hy
zvZn+KEG?W_P-F2Q)ow@Y9~&&4qHa9Z!KNO%3XL+~J~oUZ-{OEqBaNQ!uPC1y!f%2e
zm$<(tHX6TC7`P3v@oz$S3H8<k;O?qwq3TrCC4;q8wSbUf)u|XC5_M<_)5~(Hjxy%B
zP$rWX6KEPz^YJc7sxpjR4Sw5zwZdRIS3>UKYDnd|dMIZiJPIo1Vq`e}Lpa&G%{aD2
zFv+u2GD({;s+8pl)rg!R`C~IC$u)XKgNkvkRAae&4VPRaJeylqjD=E-cJ~@f<r=-(
zlre6UYW%~!##*^XzTQTgRO3_k8u-=&+t`@&W^q9=UMJNUi|MD2gw43at;T$<#+_1)
zCGIsk<Qn06bRUyy>~^nlt6W3-eUV~3E!Frh_Zq3&xg)F}J$kJEE!FtQy~aaw8&16+
zCNa$-!*K*fy7l}Cxkjp9V}w*=p?i(z<r?{V4X0FNr+bZm$~8*$8rf2fyWMNNFV|?$
zYpj-PyyITuoLr+#uhA&g2*v>;jrlinjU!r(DeY1Xux-XzomXfxny@<2#t(OG*p=W$
zEFRAr{Ir7xSCwp254E7tF=xcFG=vWG;A1$<PZ|8QNrSV3;3iMPYEQrWIc1nckMkA-
zUx?(}3?zXbb5uei|9sPc%p>G-NNt4F)&R<;YPUhWnL#ap78CR!#Q!2_L0Sf|RaEnN
z$WJ5rkHj^`DMJPE3<s1cb49N{p2osuyiF~_3Wsb7I*n|Fq6~N7(($-~J+fhcw+b`i
zNVKVk&QLc>N}9o#L}!c-klsP^-3EkpV*t`ONPfSuk{#BOF=##%$?rBJVGG-!p947w
z$x1<a!oXf!t%ke|X;hf-vjc5lptyr&=G%;a2#nzl!*6o9kWL4v@d(C83^VSx*rp!Z
z8U@G5{3C`56oG9>El7y@BL*kM=>VjCgrotPMaU6IcOd!I;acPoLpEX0LOz8w<!jNz
zpj+<YE5L7sVl8m_4uc<^2i24CU@X2rqx{9d-oA~cljNQJB$<=)YB+AB4D8}y?7Ef>
z$`OMvt5DO_tlZCWoH_QRH?PAEduN6>K*3?Ke4D|dO>4TJqD?)NHyI=Txt+8_6A38_
z#Yo*|m_Q??mW9%|-DZek(mMZ6oZ)ZdomLCOAH}KYc7uhgS(}wV8rZ8t3fkL9O7Qz=
z^9F-OJNO3Md_`;X21D>*n>QH7Np0R>h!$;bK{RhLq)<@RLccDUzQK@3FT?1Mlv|OQ
zABT9b;KxnekL#eh3P}lW!v3JaqAeS+!#4GhV~XheLBlZahd<dnXo%#EDZ^jDfqM<V
z&e7JX!9#hoR{a{i`f<2*jbS`9R_JFyQ`Z=#Q$MC+5>G;6@w~<`oA*r_ei&FMZy;Yx
zl)(miwFWv3!-ap;K&OGd6Gbqnflh;yjAg*kJgtFF!yMj#KZUE)uz>Yvi8A~l7~jvE
zOBKxxhGP~+h)sq4ZvEM>^@sVspZ7;CW2W~TW^1O4O?dvrO*?;L`gYCqUPFk~lf8y;
z9x8w8$zH=W<~5&mdkr(CKI}Ej7Yzlc<KlC!!J?h{-{hl-seBt2C_kFYa}8l+b``?0
z9Eq(xbNSk{8FDQWn@4l`>a!2>9>Q{&<y>m$4#>A7vAH$RU?Y2&TWNf5oq*;zQozRu
zZ>GVI96b;D9ARhiUM$OCCyf4w(1}O^`>@OASSN~CA*Ec3v6#;>Yet;n;-+7MihdOq
z@K1P=!?CX+({F*T%i@#;95O;XKu|`!vXDcg0S!e8*n?Flmt#vIFF~>{i&yeER044o
z(eAb4cZD3=3b~1B3)7Uv9HJoHPEba!lFy+>AwERV5@iWT$j4U+so1bVDd5<DAb&>K
z!s<F@DaR&aeH)Jya34-?g&dm)c@|-RL(qyiM!y*@M4EYHHll}3RrLj&7W5(<&BJhn
zCK)W+#pU3tLgyzoeUr46j7`xbLj;>Se7Q>I%Y+gffiu=LgGIX<9K0W(b=v;{%!Fx%
z5ww=kf`~P68lOL6-cRE*2Z!MOpiDEcIkOp_*CP4R*<qTYdJ{Tu1!Q_K&*o7KZ#;nJ
zsFTmB8=$)$33Dr!$Ka2UZbwpr?}QT{s}}8|xF1$&Qx7e1q7icOV|5t0cntKTNNl2i
ztg@y1dB{&Av5Ee%%FbLLLhdEZ#_GB#E(Ra`qy!jq@GSo|26OP=ygUl}P$W3`8UJm_
zR7jHu!K6K}vNs^hAm$^nN&6p8YlYlQG)&qrICKrfPJ%FLzvR$e5dVn8Chb=odl&M*
zh=xh~HHUtL_$|?B()M#~7>YqiG-<!#&=iOh3BsiPmP6?f=MuC;`HmwiAT1|^Chhke
zqu<Pz5k`~t2aeT3-h{*^?T;MW3b}<anzREP+YR|*q?xgqVt}+si{JXrq%-<^Y7o8Y
z(5~16I1K|C+yOJas=s#iKd8A8L+nq#;p|lts4k{{un&goS=Ey2;`QIr&Cq3JXk#ms
zquJY@14<}<y7inoiC+5{4;&5h-Jm?LhSK|=+u$huPQbUOtWxPwN6_ma`cF$=K=ON-
z3hGFDJ49~@>`1=lEse^H>IiySGzW4bk{|t?^d)s1y*sLeTtd9s3gu-rjNT+2fY?s7
zH_*9P)RFXJ=>f=pBdi-;dQ}}m@0p&1{1jnt2jTIRI+0#GeE|6#!al?UwbSYZdJFYE
z<bJ|pu&}<aj;5DV!6?$ND@Lmj-cZNVe}-Z+I&SA7S@@KT7v962Kfe?d*Ib3lsPBc2
z9-px5`LG5~#x)}D*oA;5f^n3dR4v`&S{>^J+tfpJb3W!R0kb>xTz_;+U87psutL$s
zzR!VNJrtgXA-Gq-)aDM<M!SZc*rUN2T7$-?CDJW|w1blt>ZJE2(v$9_wL0ksiS&s(
z2~UIgW*oMR7sw|EmOt`IG`B}5O_xZC?xf%9B$q^5DM+Gsf6_^166s=hOAqR#?GovB
zLE6GY@T5-qtwcKIPI_M_-6@g&?N0hyCp{yPMkKHh;2on-v(Xu{Sqs+(5_cZwYT*jk
zNk2-Ya(B`koiu#;VBp)`NsDxnQzG5vPFkyzawO7Af+WIKt&`SCq<(kO9-VZNL>isQ
zCxftbgHF0eBBi^N4(p`9N~B76(t|qbNr|-Io%FO$dRrph<4$@_C-qCDSKUdU=%m2n
z!H9q7P8!fjQzTOO47U*YHEOdULn39llSb>L)e>o=J1IsdwMe9^-AO4r=_-kIpF1gA
zC;dqxz3xuJ$8-5+?3PGBx|7!Eq}L_Vge12RRO_Vk63OLG+N_i8cql;I1L3$NawoOv
zqzH-h8+X!`I%$qXdcd7@y-r#tk=}GC{Xr+yOC&Yft#?Os(q4%)(Vg^=PP#=REp#WH
z)JcDrNDc0!9-Z`(M7q|U^sY`iE0O;0PWnP8nOCxavxvX#P8!fjV<eI(#VrJZO}xHh
zOp-{G-AUm(DPJNjawo;<q-u%O<W8EWlP;A=2i-~eI_U<9^pHDgwNAQEBE9QQs?$l&
zN+h3|ZXwvFlRlJ4(e9)xbdqJ&V8n|#2@j6Ndcw{gztg#6CGJjlgSYCWWQlaEJL!I%
zv_vAE5F~Nhen=;6lt^dYNyl~4Ziy5)i$_9OI;E2iNu(L>q?dHke@UcO?xcU|q~|13
zn>*=@PWnh9{lT5|g-&8$FGTEVae2y}^rKD+TRk{BpShE4&0M7z<0MkZY_|{u>!gJe
zDaD;MLMN3=q_u)1PN3s-(oTuA+np4tldh9Uhuuljb<*7u=~;Ku44rgRBAs_9&C*Hl
zNTlI&+>VEIo%F3ln(a=?(MiE;1|wePPFkXqq9qdjSATTXDdJSDld>ezU))Knb<$di
z)Z<QCr<1lyq%Yk`8+B5rL>e*I%_lrm<(u)(5^0`0X_roVOd^%LldjN7Z%Cwechc25
z=?jT;mpkb?o#eN6Fyb${le%=$M2XbzPWqEhnkSJ)&vOgGojR#lBBi^N{-%>QOQcG7
z(%*H`l@e*cJLxf<belxF$DMRaCmoYWuey_-)k&{Nr0?8GFX^OzOQi5rw-CIclYB}B
zBcA0>dS54nOQenNq|bCxibT5Foz$<B3MA5f?j)r};hS-jM0#D2#O;@lPP$AY{pd~#
z&`6_hlt|Jo>Tr#q41X^ee^;!s_x?RgdHd{hmMG#ae>$E6&>t}tsO)idG;Yu*BH`6Z
z0e>j95K<N)_#4auHJ%=j)j_Nxs1C^77$Ex~T|r1~GgWjEb_?VikgRp(%2NK=<uQm4
z6SN8c`$8edUWNQ3QosjzDptg?PaxBu=m*dr>n`J%f?NCVfsJi0D_4p+7C?V-yMmWP
zpm!h{;|tL^Hu^n861s=yk%0$N|7eZ5JqB#r&dnRI1g@26#%qM|LpUCd@H*xLF4R$s
zPf4on?y65D6W6+{LW@{S%!zxr%5vcF#mr)n(0z!;p3$n2gUXm@p?VQ2W%yBy+<ZQA
zjsk9veY_M!_0ZjTa>Yh&J|DR$=z9bbMs7YIxg1Ctgka?6^O388xPhQLAZ+B?A?+cA
zMs7YIxtk$hPuRNZdSHb#n%$6(A<<~Mc*I|Z*h3JFW)8>tAk&{#ve8_?vF{=G6Go%C
zkYj;(DdmS0Kz|IK%dsht$0M;3Ud%CLljt~&rWsvkqe&y<KAJa4G~>rYpwXPh>zEG|
zv)J3j5kRUOuCg3B%Eu*J=%}fUTxVz14&<Zm7{%Bm5$OL^hcEh|9uYbk)LXPR*$f)<
zv{1bWl`^~|7#~g$gSrF&wl8^xL0yu?#wJ-Eu?binf<a$K7&*Fb7+yZ8_HO_h5A^J^
z*48Prc!1VJUxH+9u5D0ebF3M1Bhtj{MFWdP0~s)Htybo#!^wOHDD>$GCD1Gab|@Fp
z@OAKTvhXmU*BtclKitE~+{1fN{jW&e!zr?dtx7ca<^=S|k(h_mICdKHt4I^C7Y%40
ze!x9+at}WT<qQ%$6g}Y{W`qFZ9v)f5J!HSbLQy^RLpG)sJ+2Q_*_+Ctcw8EYggGCm
zPNI%ahcpRkq_rBeIY?!HI=d9WVk8#pVNy>T@&C+(a<{9XUx#FEP{KH}57HHcv?wDu
zLQivVLkhUVfEM@z`M*K_D^koHw0;u)VGu^CL&=v{0Y8PLgtdtnd<H#cmR+CVB(Ad@
zIJt{vaG<}47>~pdjFpK;qAL$pu-UPW(xx8TH=p&|q>iMPLj2J#5+5tG*2C^y)+WV=
z`@RTjdTJaK2mVgfhM!jM=WhXbAO&VEBI5`kjuIvCFG6D8Os@*UWR3w|noATsde(h8
zg-5vKo+JHGdWCThO#KCEd_zsS@+EuQ^CHBj2&z@SoR)~n-$FQ#G;a7<R7)U;K3g&x
zX?&gX<y-<5LdZfI>7+yCD}I1<@0vt|_C0&IT?K79(!{rgPhuu1i{a6vYNekI!zEyB
zLsDWUVb{S`Kl)P>M*-8He2%IU0saFF>DR#QS7VkYB|c&T&Msl|1?L~I01mzX!fhq8
zV*HMQT{RaNnxHAhgGeRtTEXVg`z@k#HU=%0F2oMo7d*pAw&frUnP}i_7)|si;Z8Ra
z3+UM?6wnVK(w`n6GG`OXZIc6kT?~Ln=2LCynE0|VF3w%bGMGjfK4W2sM5WP4N?4MJ
zRwLAUNBA!>0t<zXyPbq7_#U~PzkvJ8J}{1=dT0?4=Hq+A$j4Gxrk`YvZ&2Q2uj?;{
zunmcg@O#nZb{E8h#Hmx>n?b;n5W10gKi=2-aZ0qdoHweBfsB59$ofHlD)Baw5*8=A
zn+mnw4;qj{{J@V5!V<w?y~!Ir1n<`}RchII3F^!$DQQwNuYibBosu~lrA60|VWjmK
zijm4}`uT|yVWRCZo}ur)2vSr#RN;Y`J4lPKg^+&>c4$KR9<4^*3Ly!qSfnL)K`<UT
zvNh`OC@cq(hg3iz+$M}0+4l_9dky9y)Q)X}iv9@mBBa=A2%RWgMM4RLn^Cv{Y1EaX
z_FZs@rY(D3VDybJVMeO4hypJc+l--Gg!hWEL{z#gTv0ZoF;rAjF_ZDT%)m;a>w?bs
z8zIy}xQ-Jks>YiIUI6@W0%tdk_&sIVUxf06u22I1CRq)8ObC607DqiJurydCcj9pX
zofjAzJ2nssx=vd@&oB_AV*_z7qHz~e6b;0)C>%!$8zw5O<l(Uy^>1zl(v+f&IvO@j
z2p8#LRK_F=j3&yc1wu#zoBWH#H_MWT;AD?9=VH89YeAZ$qRp>qW~w=gA#_(LCX!jn
z!69UHsXNt+gH;qweu~k5<~Zvtdx||LI2lKdIe{T6N9KhYP|w0xc0sb^=gtB6V>Bz-
zvVv}%IZmBrCIg1y52O#ryS8CuL-288F#ik(-vwnHRu2<(s5&Vuhxz0OI(%|B>F*B0
zsQ}GUxn&1Dg*#Gz0$mDq7B7RDO6o*#j5;q&Z)s8({!q~;HFSFz#U&&{vD@d`jW{RZ
zq)sF&?;(X~*hCh_;U4vKf@KS#?wy9iVH9>UK7#s61^OwSGK!A|Rzvp=K{tkczf=t)
z_s6;g8ppaDd(=2s8mGmG4W<~2@tPkpI6+l;tRqApSXgKPW;4ekbr``NgeTfn>LR&5
zNpnXw8<n8MYI9&R51V<`I0P#hb7l(lQ*>r(uxdB+VH#&6bJ4__j@V8cs!)44JKEKv
z;iQ^A7-oCQI4f!8I6JjCKuI$YVT&~nx)GPe#wwoKhZhM^7|l^AB*F#z%yBkMYVN{}
zU}$ECVVTq#^u>^mvrG!(F-c}HCYL`|VY8JhQZx_5RGvv4#UjoFZ(wZd(d-yC*3EVN
z3TqyW*ZI@@qCnxzaG)CLA8a>{9XXad>A=A>PfSHCzD|b24z8!sp+gqagYjQspcfcD
zic|)3M9grqj9(v_)VX#$;=+$cVr30hN@3NqVa^J@aDTo&_83~N^DZ3}6C#{d&7mv6
z1=J06fmx7;2iHtu%&Fa6cN2UJcd+S}r>j_#Il<xf1g2QTX0+%L&TBNku;AJeNIp51
zuo;IjFBlE1Ks}mnE*f5JUtwO({98I0VFdRM)HN$sWjD&Is($D1M;UscAfYhU2SxUu
z;#4B<M7j~_VWjg&26e0<Ab?KF3fuWnrD50^#?medJ0%`F<1Fk<S=gCZP^la{%VrX<
zAn|Gve}|oFo~hUj+HX6IZyQyMYBpLHn*EWhH=Aczrkjlm&Av>sByeSLuvxVPn0>x9
zk6CDrYBrBqVD`yDtvTi>i@E@Xn}f|$Ks8%#HXDQGEkqWAgF}KtEGN-|7348S+XAz#
z*gR&1dGreNs1;_L%WO4nDK^??8(WIaKE@%JLT)ZLTV^u{EzFlnGtXCuyb_I42P(}z
zA}8BM%L>+p(Nble1bNJ*Xr$6Ss?uyLf@X!;2hGi=YVgTCggRkX@S!Gav)M>F;$dew
zOtIYh5AzHGL{q-OoJ!e3+_58D$(2Rsc#;-yFprlRRAipQCDzJv{52eSEA+OqNQ9e=
zTWn@xv3U-!l0jLqc>$MPToUG(onm2U1+Nb~Dc+En=Qo*|dFyQRII6RPi#fb^3%Cn^
zH?QDrC-Y2h8-q_ED#o!KbBJ>0dNGy4%`KLv*@!F*Hk*SnJ<MYuPP@q*wSZ>MUk7K-
z1d6}{vpUBdZ#gZhgqQ<R3Trk8%r;NKJh1LI7g(wQMOB*%QZV)iB4%d+W@l8MxnRCI
zYJ$1oV)K|hbJU$|cz<DD(zNuLn;aUPpF@oP3(RA(&7&7$8n!GjTW*5Eh1&FE^E!l0
zE~BLdQ)n1*F`Y>o4F|2P$R<}apIj&P@%EToEflC=_g#qDmjDU_%U5t47UMRRYnJdx
zv!UUUCSDBNnU`5s{(rT73v^q>wf2#8jw~F<@gt5M=Yf&{O-ucd>^M$IxN)3!9Ajh$
z9NM)GvUFq%WJ&0e*tCBeLMi2SaSHt#D2>f6kHBr(EJ_Q70{PcU)*o65p+E~i+`AHP
zDO}($A@mV`3jE)mhaM*hcht4+$%^NkJ$v@-*?Z5PnWLF~P|^u@UG7?AllzzA$$_V>
zg-;uemF{omGdCJvWRDJ%{0bF9i>am1SrzN}(87gKn`e{J#niy6I0J^m?W&J(8gL)g
z5yQ1&4{qRo<aYeJd@eek`$G5f<xnUw8FOb@UJvdrSq{Zt2uE-(m6bcqT>t!aa$SbI
zrxqDAj0K*5GS0oq@HiLOky*HR{%+?^!|}~Xvtfjp0XwdNY#bT(hjmAX+q0ABfX*s{
z4;s_mu4f(2c84+3?Kh_1xf7&3e%%Rw=(^*p4jpjUoBDDB;Q6q4{)2%SsfICgCg2X&
zcfJa-)sY8%Y!C>B+w~{Bz}RF^V-3#=qwWIOB+zT!8F$MhT3hcq-0U?Lxm__&Gkk9?
zbH$5NMrEt<B{zB|x(s~U9>d)X>eU<3Aon3EHxn+x?Yj4CUbn+j<6Kq)34T!P`Mu$R
zzjL@VI}aOG`;7U&Fe*10=eREgjG2&Yt5Jovy4;SPze9<cwcOWx7I=1gLC>=S)S8`s
zJ2lTnqv~2^YHt{8+|JL}x*g5{C*)r4_UsazP3#=pX;iIHAy&nVz{8z><uB3le>KV-
z#w-V9RJ;5zOv+tH_Tt&eB=`X;TrD_%98x1epm`gNx@(QI_8DcK#|+TI&}G2BX`DC*
zlE7gAV?L$O$tL40G=Gm#UIofG(5Sb<^O$$gXeG;&Z{pA_qsiG_Yb=FODlRmZ?$~KG
zZ8YZ4FtvjD2ZyU}kLM3~bj7`-&g%@<_uWsU8_^)^y2_{oz}=E(aMGx|)hKH+%2Q7o
zWn^=kjM?s)hUZHd)c9C@76!QKC_DqnrC@R^XMH;79GYcl;N0bIWGu{Z?}UZH9ee7G
zh18jxE4W}F4`{-b1_>JEse`iuQY41^_atJk!>D(AyvDSr!NcRmdFcOj7b4j~Ib|3g
zVLyirM*~E<fP_<ZF-=kMqr`t_P1uT!ngY3+IQN6#t=eY{t~09dHU_UUYDbL09Y%H3
zsJ*}#Tu(RU*BI3h5KUPojcPXyQTk5$GsuU%J-DwxW6wiciLOenx3bCbQL>3cmvi9z
z%pF%k8w`h$s@z0)P#`6j@r=799>HpP7NqT_$3dfE4~fv_evD08Fu0dv-UMkg0|APT
zc}NRV_hG>we~JK90f#_NzVZ*oVoD;+b0J|*k8x%<8W&RDMgvQ5N(!GMoWVghd{hn-
z0_vyaY!1pbzb7{e8$DdaJrMDmMh_El4-+wDTDy@6H!tGh5=6`-zKOZx%AG}$hO&sM
zyrjkRWU(-RWR#O6J%hEz90>UH8;x@6hrq&9<vzoG4W8KS7S!|;#`#Fx6UIDB_V6bu
z)-O<fe_$O8@?<iL@TbR&g_O4ub7z*&1kLuODpm_U+tWi+?Ah*Wqp`)9{Sy6MxWbs7
zG8$cHd5zhw?HQwSt5Gp%EX2g(2jG7Ny0F8j8Z;JVjVcbH$r2QIZq}&efO9#<=W-Y{
z@2!S`XN_u(u>gQsgGO!En8g9)A23Zp5f2B5B8|Bsjlh^TXw1nP(>Mmn7yw`h0AYaB
z*;%9c9;3oFPUAS*!qMt;X!QVKa^%k%XK_M)05E`kK5NY6fX@R^J7}DpHEKEFYyi;0
z=50nLdXxJ?qZxp@LBp3d>NtiEfQCV1an@+yfW-jJ7&Ok$8Z$WHd;mc7i#eiS1OU~Z
zm#g+X08mS7PL8brKuQ1zDG9(AasXQCKo?+L=z>_U88m=V!>c6>RC`XY+H(My&Z?cx
zF{oN_3Jk%iz<|aD25B6O0U#7C0P}MInn#y2^_6p?LVYMAG)_gJ9B?Ykh&Y8!p!otr
zG#^GHY9S*5r!X|Z2N@bT1%}{MV1N&SA^3pQ1s^1J2pIsOTLD081wgbGdJ!~)UW&~D
zoB~5|DllNIf(Etrbjn8KgawJlfe+ERGw}hUf)65kI^BZxS(8Rhidnbdepas3ER=#K
z01y%ofQB5P0T{4<0EGPuz}y^QE-)bYc{!rAh=2hAAh<sh05k-~nP{AXf$l0AOg$B>
z10YxzfVv!j=B;H|re}>ZZf2x!a0;g&XpmC?jpEkAu>eCj7J&hPs6_z4m;ey#r+AuE
zh=LDrD)=ByAyW_)G8N7V5);meBnJD3F$?<_UbEQTf#J<zU^0DLK0s7(N<`6oxIoc-
za)IC!&Q4&EvqSTNA(}5R01y}gP%MvA{OT!0(R>J5G+$T{h>GSDQP6-)Me|9fXk76q
z0+T{(g~3p3fdQv01Vm2vbbJ)s2QUB-wFqB-3c%%_GHCu@iDk>A(cpfX20a>b+!@0W
z#UwXnEOK4v+{wnmCk<B*Mh#567rH%<d$xY|Kg`Q9VLBT#PY;a&E_V-ZtGrkxy0_MQ
zwz9>M;hK+SkB9my<<rdj9IUAp(iHh=Hu=ZAizdCUUxFN7Bb<%jDI5+nHO9Uv-Z9j^
zbV+!~?;mOl`osPu?d`r`dz;@M2`&rygH~i&tEP<VQ%Ch{R|G?5D1q<DSSj^l?S#6{
z%527$&|^K(p@H$FrCz#29d)S|N^CRJ!O>)_gbJ2X1%c5b?$Xh#EgWG?Ct};Io*?!#
z@=v*k$HN&jmQI9&nrZe1F1oaDwP|XxXuNwg(`6!TjCu9d{h46Kx+IuF#^_i&V-d7(
zbpbRJ1HpbRU<Rz=C=gQJ-N9HaI0T*p0~hsZG1|~$nyXi5a)nw!44Wl-qVdha%$l*}
zMX4}Xz*;(##(9gzqiHP^K;>GHps`r=%KqRsOG}MG?tU{8Z;fgQ*dK`vD@(MkO%NZ=
zL~&eA)u!5(X%QGnSPNznF~3G-2S!pBc8;0R%=VEeS{kyj!F(GUWTmyy;8_1iG=i;b
zCK{1xHCcnSY})G2g!>ZPwdim>k+Q<86KN~mt3|^|<J$VoU4z<{w1tE>FbeL5qVdG2
zwjmzLrlW~Cm&H&bmC;OUI6@uAt!>e;UyCP((ta&806F1-V9T^{!c0PIupnxu8B7ny
zwZyg{sbVM_jb);7bJQA5V`oErlVv5>3dumWCNq;1Pr|`WP-L<C*K{vw)vU2hD%cN3
ztgsgEfnuc1gtrCb{Z=NMB=ttLG}+ETDmuDFS=OCeU?`D@S;4qAvKl83Gv=_B87M_t
zOF9I<Yn3Zqf<9AZpne#rurso|b-~fm;9AH#o$i7OXk_sl(*2NmAYvj=L&geav`{uh
z8tvE8p&%@R_o7wkMd?fw=S;KlH1uwTy`d2(A(}R0iBLo{gV-&pYKQ)1FsZSa6-QU@
zUo&9#4+J*!t~1y43@{5Q(CoIL8AD$LWH_D;nYqVR5Upm81~b|StkkbubBA9WABhf+
zpyi`O@D_10s{#Ra2eYA(agq@4*3=)+>qnT$#8*et$yji_FO?WhS?M$ykrJ*VkjTOi
z(Pv=K8{$zd7!EUwEg-$h$}rJGe511WlH*0AQmn6a#bRV1X)Tg86Ojm6bTX3)L*@`u
zCK63???g^%vy~Z1gjZV;^o|Y8|HwuJ)1)0V06JAwhx(wTJ|?Nx-loha8?TrT`me^>
zCCW6_N3o^o!hx=VHOi!y6%7q+a2*Mz*Dyy1vlHHjx*Gl0;0AkxaOTNGIFWOm(MTK}
za5PFiKp4N7!61U}8cp|RW9S6gBwU-td?d^T-IUy>HXcr7q)mhpqwt{C+F&S?NL><)
zVaU<KRuTwQSHG5`v59=F7@=c~*Y%q>^rFcyueJH{7K4V_)3q+|5PNvv@MBCJiAQ7P
zJouQQYJ50#tc=>h9X5%JdddcT*k*N)xpqr$x6SRf<;LZ{k>~(i1#|RNuGiW=h`y>p
z6I>qXdWo>lG<zUb(OV%N4kN#<%ABvYO-pA-ndk!Mnla2D;=zKB3O8y&c&oPrGz<iE
z7JXgs7H_w2srbU5w`nw!jaw^*t+<tnhVU3ng+^A4b#^p$w0oO|y-gRj;lJP86iHHc
zB$ytLhelF~cp{r_%Eq_DVKrrFwvt9B2CUGPO~RumQcdX$5J5v^gy8r`WHv@*WP~CG
z&v{As^S5nU7fpkT2N+0>P3~ZN+E6e|zHj@8uT|ZpVAF#cw}^XL6Ae^-iD;U<u&VTz
zLHIN$ENOsUrtT(|b!tf~9*)L`$%xXL8S2jt70rL(y%V9{A}j$)$CD|1RxwgwgcvzC
zq<gy{OQt9@G;*aG38MQz448YtXg6yyQW*O~-(=iQ`s?o}i$oa1Xr>3#FY4DZYs?!?
zWNDHLn<WR*2Tjn(7|}w((1^8p4Xh5V(FBJ2OgwCjVfgolg7J_Q1B-pUU6{TFtYC~h
zOKg~?3BDw_><uztJj_}hFthPkA{Z7E!GghUOFSB3`YM`znA&h|C=;^^q2#FFOhx&u
zFanBl9*+q}$fVD81<daJ7$@o}n9(I-7+z^+vMGoOz*tmEq{17*Xu=pe9s;A8t79?D
zHl-gm+1$rUj1r5yYtxE>4DTu}9%M#SIG4>c30IKVW~IR6b~Bh7#(yhFBBDW->k}&s
zEe@&nV^oiu7@N%u8^meS=Z8Kc{rzmQ!KRQK*n}C1kwbV|i;jgcYarXjEL?FJ*|Ai#
zB0t)*$$CGg1DJwmN42$GJ^kt<m9G_@kdC{=yJ=jCr)hePR)Wn$=n~8c6DiJO;kt!n
z+C36YUqwRYD#93&wlG7_PeGXd3i}LUt&qliZ?JD*Yu_4(rU!~<iw12p8t?CC=9Y<W
zv-&X6j?zt6YMA-c!miiLmj`@Bg?TGmnh1j&PohtSsE=S?hspw%2<>i&C$pJLgDINU
z!78HZ?hy<M855lYM$}79&v&=n1ef|<JdJ*?su!QHO~Z^3<LEYsflP%P8_Z*^=-Rzd
z5t@ogD_?JMgaXsX;83t&iCf9En1MsaECW8sOe8b#+G(^stfgbyXmwQD+JC8^DI&0@
zTd>G_9=%22tLRhf5_pi9nGIt`%H9>^R$y>s!po5_Qx}-}IN?*Dd-~L;6i}G5a@D7w
zKJ|gcr#=Dop_3{)2w$f<>N3%JnUh>gojFRqg$VSE`3!k9m>LF-6nk+{;P~;<gB3~B
z$d3sRCZviy`qg#sSJ%B?UH5)<-TT#b?^oBozXNk6ymYa~XnwCMwoYwz*XjW|MA2NH
zuMIKvNJTYz%Mcjrr||@<#&8ht#iVb5r&|@FH&CK;TN=iSG;^>hL^{?mVGN5IEZ*|0
z9*+k{qoMxs(V;{PgH$>|?+gITy}P0@U=_XFAsbL_L7Teb+td}`=4VS7nujVgqOR38
zb*;9k_vkkD9^KZ7H5b+e84MoVg6y40dIT@7yfjGBn5x|BQYPoVR&@pVT61d!GK+yy
zOu~nq(W+YWwW{mF*V-XFir=KQw0b2ac+l2m7|)@cu~Qb~2*DOlCAKH=b`fuXw1L1D
z^+{Q~dN*lT*HXK>e{5IxkL~K}!+|wbCA6#eq;~b5)UK}C_GN|hN4ztLhhrAKz|q)m
z1zLPF^O~yd@%hwM>MLcwg9#7T-puJM3)kmU*O?C!nvBVtZ_?|f9$Fn_G2`QJjB{^S
zy7(eEPD{8PkPFOYGfe(_1)o1)4H$|0wGsZdUsZ^BYp;x|-Yu4>cMAlMQ)WP2#5kN7
z#{gXd8|oInq_c2}cc~Q}!iz(4&sZs9iItYQeBox&5XzK-brSvV5wPf0H2nfZ>+-c2
zbW*{1esxZ4Gu{JQguP+xm9NzJuX52LzNFwQ9l6d>mSKmwzB_!{=D~{st9#XntLw5u
zU6(kZscIn|9h~KU^%CmttAl<GiqP7Y{SeB2!$7|s#j+ji`%Uy3G=|?Y@WW1a=nAaz
z(^`C0ko}s+7SVo<&oGMC!$QTF1Ln+zM*H4GO3nc{q`R}J6n-}8=8HsLc^CbnMvX*;
ze2S?mqNVBvu+*oegAwa83n9mq$*Y^dQgstps%`>H)knFd>V1Bxx|u9pmiN^Iv<hG^
z7}rPX9r=hhj#ppG!t7MnRi{tR%-ATPs;*9Tb#<z%t8<C)U1{~Es_UXtT^F60aab0H
za{M+<uUN?AuCp@vod~qP>kFpvo({kCu-}yFbyqlnmtB~h;74{W!I8<{b+wEnMy-}q
zHjZBuTDB)rSGMqhS0>>^DBY5>A{KgU$Z8oK4<?f><ffa${B2KjNNdTaQ!UYWD3%Rd
zE%-6Q>}YR7kIs%Y;Ro54AsQ1HN-R3mG8_t>isJLh;zhKy7QDl13AD5~w>P)8Q0aig
zYYCW`J7}kJZ)U8~WD7Q`q=Kn&(TF6ya+D=MgD6%{;b_9&JVG~NE5dK^N*6cF**=1%
zX#7Hs1rYtdjGyjtSMX-DKa<LaGJplb<Lo^+-D3U4e(6c)aOq_ve%4LH$y#ARyHe52
z2%%z;Kq)2k2AE3xUz~fUV}A0|Czjv0+_6Q!Z}|($bwo*+M27y;<&P|P$dlXpkqchC
z02r@b@W};^OY~1JxHg9f)T5q{J;?mn^NB|mC(?kw=Wz_^?@{)k%pM@>bRqkM$8oWK
zg0cr>_QgCKnIC!_0sTWxGmwM6OQhcA(7_xO6t+qVTLtuh=Vgy0sK4wv>T$D=Sa3Kh
zKHzzl_<fcT1{p#T+Xp-^5<@Ri_Qf(g&-MY&^8|mMvbV@=!8V9L`x)8#l;`JE=;xkC
ziX}DWd624nkg^A5cD~9f&pxVhA7u~7Y$+*5Ga#Xm6ppsDOaF7(M`eq$n#X&N$9rD+
z0q6YOr$XKCg^NzQ5X0o&n&yR*`bTA-s3Cq*cAaNlR{Jbk*LlA0spBYDX+_m=rVf-n
zUq<39d#4OiM<a6V9xVH{8W+DVdsYqc*|L|){tI?r0w15n?o`>{GFZ{xvIoivtw^6J
z+fDG@Wp~)%#%l&N{F?EuJ8x9_1VMugT1W-vGGRPvK&MX{|LJyoMgLDXE0)*%gmIq%
zY29bM=|;bMlV?lKA2ePyrX{%^t4WWqqMGWAh7c%Hy#&VLUHt_dkZ4MK-B5XYzterY
zzIVyq=C5fKODX+rhkH>K-ObHs&7ahqyp)O0zua5fJcprqS%EVWKn06DKv~F;8*m5v
z0$Hkoin6DLRLdfrvw(VByq=CCETTj_Ej*=m#^fSTuK0P}te%A|^d#bqGa(@G783$J
zKMtWehJdAi&;2;R!~I9bBW_1ne*|*Tsb=kSj1F~RP;wKcMIJ`0B2NpD7u>J8J%Xu8
z{WYQ;B-%d&oZaDmhx-M$BgACxX076`R=?i;pnFDAD-`~u{vbH0&I_LFgmlITUGK)1
z>rn7zH>ZU2aQTdef{X^%k5Dz0@D^lvIjJAAx+_H8q?-A>?pmgf;tWmCSF*T()?y4B
zIks90GFmlvKBKcBqZ441n9_f&!;F8dzo7F@FQ7ly;p2bKq0DpX_vpXYomp1@|I(k+
z3(cF8@@_=;+PItOuz?EHy-D<~<Zj*8wY&(IrmwqAN5{EMX9e>8%G(Wgy1%14`t|Sd
zJlUx>_B#6X-Xwi11D8ISC}crTB&mOWt+;<70u7-KCAa}5l4O2_L*}>7dCUd#WBR<s
zxzFhs0_Yl7`8Mh(_s)eHc<CQI9&o~F9zdR+B!fJg!5Lj?j)^20qo539{FU=#r{~^=
z=6mNg&zaOecK*QS7~uwT3vwJI08vfUk?x(HtLb6Kubhri6u!pgrkW^N5-$A~=PoA%
zxy!j*ckWuxVZU&^?L;?uJBLf*Qh@uE;}xfK!mGcM%Rq<$O+Vz=tvlk>rEdWhM}%O%
zCfF5pcZBXJascjs=lFN0<4XPCbLF|EuuRI|an>_uLDA2gwWA#R+fjZ)Ii7APzp<R@
zUB9t>4*~X+A49e*#8O^z03~;p!=vt`k^{2jz$8k3j{x7Jl3&S{<fSe~DOSV9g!gzk
zY~*p^xpwXTh7B_yF#`fKATX!wn1d2?P+$%U%)NHZtrByqz}zY@_j63}c;YPAu9xS5
zH)87Ld9uN81AJ71kLJLy1N@c*zm)^O2=H+UKAr>rKY&k4@JSAaA||{iypZ$>@7;@_
zh`Sd(yeQxESoRSwjOK{<PYYZKs;K{I!NZGe{g6=@0EMYV=x$S-g6N7M|D+dP;z{qr
zi=4Yw3MunSLn!jG7h#ybp?{2QqFUQ5Jmc-B7SsPooQ6^8;YIjphmhG*3tm_N`Y(Vj
zqQoWkGy#JE{%JwZ(m!2r!$QDrSoo8L5{Bj-nfuT@wEm%ae?J?o|NGfDd5c<qWbTJ^
zQRu_DZ_H<{e`EelUR&!Kg#l3b=xnt9BThlK{>a?h=YhN1=iTJBR~kZ*C+4x%KY?u7
zdTIdYDhvcF*2jIzi>H#(@0$niErGX^xBc4r6Z28!1UStL<lgxY%!j*pU_P}2q9bob
zm(DVR$T&FPxoc2AIR7veK8(V6<S0iP56*>%4$eI^SIP!chyOk2l{t{iD|3(;1U&#w
z%qicKUv$tohC>V*P)^Ko`X==gbFQBY*L(fkTj#ox-<Z^IoqG?SMI8+X8>x=QLyd)X
z%zkAy)iE2HypGu?W}m(eWUxABU*AY|G~U|C>u9_O&%BP~b+67s9k0%Mwn5gxGBo{o
z-Fvg(GTxhYYdutWYyCt$5a3EC>Yr_J^z+hh%mU69fS#(y(81-W|D^u)`W!#U|4Dzn
z{`v-pZ&$;kNVuaeLV>_vuSZ@8H1z}ZPdBh;K}_`YdcCuC68SebpdLnI3!ghXnJ5cU
z{#K7h{;hsj18d~2hVSCWF{yvI;l2iB+=uGvnYT{-)sNO5n+d|lW<F5|OMjy7P#vBQ
z)xAMaZ`A#<j>$y-W8FId<m`hA>(9?THnXsm$7X)B4z2uV-EDO+rmyR_)q&&?BUy-c
zsP1imc@sDhY9Fh=pCRK>GB;KIvl_7Tvzl+!Vxjk~+Lvl=Dw?W(q6R8@qUJZX;P*GR
zM+gBbIs#NVKtN9VRQ28(c=f$Cx79+Gx79vetEV;Zr2a4`DJB7|SyR>fYcM<6U-Mio
z#*ybZNw|APQjQ!TcwG&aY}XNImr2g}!qll>U%Rgs2D%T`kR+JKNp|{V%~Lhh+NWxc
z)F6Q6xHh-#Tu`ogsivGgb+YIGQ}a$fmj=~6AS4V9M5bo!uYyqaSN*lh@n!w5Rkv3c
z3I*9eodJ>jbOy37liB$h(A13YS3wfrr_uv58|~-)669~MdZ>zce5mSh6?nXw^LV)G
zH&x&fPn^f!R{cjlSE_4b`n{Fl?B2@vD^d0PmAhsXaW*l1cO~lFP1#?T**VTo^~ChM
zr-Qe<r{6RKqQ7az%~T#^^35}L0g$g{q7n+4ATk$AGMGi?T5v<<{gu>$`zsGr=38)}
z@=#@A3w~AkYCe~u9BKM5rhPcg>7Ud;oc2(Kvn^+ZznFGt8v4(nX%AMw$o5q{ghVjI
zEat;$J1UUBqvDzhwEmikA60<+A64ulAjAF&De)K6el@K;7vGPqe+Z-_3Lsw$1LJAL
z_R^mQ6s=_(9jrK3K}|VUF)<yEtgzN&74J|5?^L`?wZ2>N0o8i4;@0U%Zk;|c9sUh5
zaC0ww^b_TOFW0{}?dA$Bh^E~+O}~HI&!;h_H1vl(wamz<KfW|m|2H@T@+&XZ>u=TX
zry1U7Tx^D`z5l5l1tIsK62yt?F3tgj!CP(c1p=P5!SP)Z^qofS@Dy%vaUaXcI<%SE
z$vZe){-r4=)w|zDhXN})v|8=B4gb4<>!;!W2MoValV8V@L)asNzCGOJa}2*ulb^h9
z#WP)JlFz@fyc%t62VoUor{m`xhElDKU0sa(2_Bu=3~lC6F<eK*4h_niDuy>Q`0SiK
zW$t+be!zxLp}X)AZLRfAiOJLDfIGEvO<P$EUj=wI^p(7u!=>Ct1>Ez)V)$;rYsgMS
z>_$E;YuW<>zt2YJMS*V*b^58mAGFc`0{jA^x}I`%08V@!_?KdSh6Mak8=afaKxgU<
z@SiexrFL{3r;mtr>^dmuFa8()Eamf_fM0H_x6#G<+--xe0sM6B3NSe2xt`Pa@<fXW
zxcz&n^l49$!?$hx+$G>sHu=*>JPCiqhJQl9ci8afVNp-`o_mVhyH3Dg{$VlvDgj?_
zqkk9Rr)%%O3jCus{ND@s0UJL!L(A;V{(-tC0beQfg-3QR#SQU4WuqSg+$(Sjla!z%
z!_E9fWd1<VS#0Cyuz<g1gVP5=soujj_<X!_A@~lP+%^jMaT}l80H^+SbPZ>uP4IJ_
zz}IZ__Y1hkM(0%l_uBB?Wt^XrHhz2pzTXD_GT^7v;}rt`hz*~<eo5^ewAK5RfVbN4
z-xlyH8@vu3jpT47$OX_X<Zupyi-&@QzEr^EZ<E7i0zP4*e;wf7g7WhAZh>EAqw|b_
zx7y%;6YxG8eBoJ~pCeP8GSJwyS-@YO;?EL32KedZeA5}=cM3X5TfK(_{M$D99~bcB
zHaRq6Nbqu2ivG3(PV%n`aIpN0f(baud9RKBjRHPl<L74*-v&R%=oeMV9{$4Mm=E;w
zJQ=0li-4h2Z>z1|O$>f2odLj4C(jVz3r|JYzRvI~HTwsbZ(;NswIjBA9~JPEwsz5{
z9EeYk4SxnESOh<AgMUfDwR<>I(jF`UUvJYFeVd&0a``%r54E!EPJw^ehJR4NFSphE
zo`5g5;Wth{z1$W8PW*dq_?rd1(*_@7bSgC&7!+~U*>#(M+rLvjDd-%r$@A9&{%srm
zzY2JR4SxZKQsy_hd400pm4Flfoi_X{8NOhyAlWAHk4|x*)XUuhK4_!!%o*tXj=|y2
zO-@JZ>rDZ-$NjGX!>6;`*?<%OW43x%3;2XhzX<_<%O<xw1bn{@{~+K@Zv0J&c8`Pe
z`<}q>wBa{aaensO_*n}$(UDOVk8v{ERRVv^#?KD~{AC;c!-CEZVdv)yI?pq>cqmAY
zo&o-mpnuXv|2+J<PV%g>!7mo@<5Qft)ce;2ywAq}{Q|zyCWqe&_yHUK-vwN=!9NG3
z6F(z1egc4#Jni4A_#Xm)y$ye_!1u1_{PSNu@JVxlztUFkTLOQ-P0k+(_(YJ?k@FBN
z$rwNOFD{%9IPveb$>9<KAF<VYjex&plk<-Re9DIZdja2Jqwk~-o1os^HaRp2_~kbI
zK>;7L;a@A@M{V`)1013_<vJ|zAGOi{Gs8!J@8(71_sl;P_%GY=&&Lv!@o$4)D&UXW
z_}mFN*~4++PwsH=dL{(i{%z*R1)Y^PI)4=SReiktUqlP5XK{Xx+VGbP_**tUM+7`+
zliTfp6Q8ZZt}s1j*Ifc`|FZr4f{tdR|62iXu<`S;fS;V=bY(wXjD;J?bFYn`egSW=
z!L?8-lg?x#k>(H%zRQ>mW@gk3;n*KGS(zrHj10#TLpU5C#%ag28O)A}^R<W#(u{3w
zr^7HK*cFR|eyQNNNpT{^wMYs(Bh7Gjbab5dUgn?pF~veuwjUM@6>KmnJUZFz6DRKY
zZXya2(TvSb8Sp}zhxq>6G~;Y-plkCQb4@SByfIB2D>v7yH#u<q>VU?$)Xa4~7p>~*
zF)v!XmO^d}bgk-HL+AsWyK_au>4ezjjT5w6d%HGoz`=Uj_^Z4t4m;*mck*E~BNKtx
zF2$l?sFp3^5J3rN#}JAD+hcR5pH-PD<v<Fe!ki|&5RYH8a$V1cRo$k)+26cGh^8>E
zk>dHVPWA1irJOxvqRH?0=bAg+25+{@SXFd5R<DYR4f3j}_{f1O>Js%M0_s~^)vHH5
zFl#KCSG;=f@!`;sawFAI5siVXmcn7E(r2f{%;C{QoWdETGUSF7t+j+Rm?fNtDrK*}
zx?1TJjB2~+_zo9j$p;=vIBM9T-hB}IOO^K$Hr!Jjr{e9&ih5U<WC@4BO4*-Jj<_h8
zMgCxfs+{O>p)xMEt#26EZ02Vl7<XyzF~H)>sgfyW_q+Q3^kwQ!u!J+{IOcnr_*ghw
z@Q=8=5<<0<aFh$}CZ)&;;XkC35{`x=pqr|qm#KGGI(G|OKIOQ!YE^!9CxUaPxsWin
zxNW8x8Vk})6K9jJw(=^@O=XxLq;r9)eBm^gDypx9hz}(k&o1R0m%6@iNQsXObZ|qp
zh3b}5!bw;hQi9iGp=?z7z*!+xRGbCk?JNjx!sFcJBQcg1zLJHhQ5;}OINgeKLZ>i_
zQ#J)rSd^QLQ6(QoQMFVFK@OI&X}E~|#+lAX+fmjGqJzlzQ;O1Drfz>F9FP5PIqX_O
z{EZUM!c%lNRed75r79`~3{u7IK#V;WsSY7{U}bm*kc-V^4<nXyzv?Ry97I)Rq?D)?
z>diw$B2@-(NKO@1eG*i{G0YMo2$T@y1~C-4e&vcm1~gXeqc~8>%d==xMRBoIrR7r>
zEKW=+Q&mSr5Eo^t>eGW|>Q=Bsy~iNrpenB=#4G^OqEK47uH4qF&KiYN;kLlT+$ma+
zuT_02SV|Nb#2+Y*AgwAl1V~avEg@J*2_d37)OB21a1W^;8Lb*GS<}jGMfg8eu^?=a
zGHNOIrA&2c8$OSk9v?+yQlyy_Pe<fk905d{agw+hp+peBDl@J%=RkBSH{66^D-3}h
zMsLQ!TM*M&iZB%{2k{vxs(8_ZiM*5*3lgHZOU5!<GkclcjN9g6gui7sX)B~PXXwOn
zGri6SHp0jwSWmMxB3|H+gh5R{aT@%kKQAHgsT5TT0*pb5odu>68L9>lMsx^aqni=9
za}>W90A>wkDX3&HK8#->#3Kt)D6@u!Qr0$^gM-(WyoUtkErIdoE&rSdQGlo%!Bl*h
zKl5Kv{_($aP_zvHor1((aU)Z&fXhVwUL&8IaZ_0SuqZFnb0nfjv0bd_>cZ1sKm9Zz
z>pw1TWO_gpkmc<4Z$f!&-N^C3pP0cV(+2zwP8a=-<-mVgUT(MT2h4^4vi`&PeTpub
zt`g52%4x`ZO23T%pBC_6-o$fdO8YeFlKjhhWIBSpPYTNSkw9_D^m6vZ`KL0He*xCE
z<A!KUdizCrnXbov5xQjkvb=2nH6m{n!_u^qU*b2pw19sqV=sRbV1$BgYy4VSoRWPN
zUXl+9`xb5r%X?Sy9GPA&>X-P~7Qil<?!r@H`5j$6U#6}6G51%NlPPVtEiAuZl$U8A
zWfokroJ{Fw$-?rvec0W+Q0>_Tg>fs(N$3N(DJ(Dd7t0jeWr{Al{C*<JOZ{s}fiBW9
z?8_>=r2ZuA5!_Jyvb@~)EYo&d`|b4~vX$TeRZdx^&)5ps>wgaD*is`7s%W`=<Hv8{
zbnN)@S*EW7jk-AXVY=i#^^=O_-$WT|qAV}>m1{Q=HBD>fN$#&VcgJu3jD&1omY4gt
zTSfV!LVoghbji20gYTkjVf|Hiak^DE^YpkR%pY_8Pty6=R$iMDh!Z?LARc8R%gGcU
zE-b45z!b+kAj(TTdwKa>f%1eS**rYOa}JC0eKJE7u+ys+^|y+AnaJ{T|B!@}tmu*{
z-)GNZg?~Q~_1o#nauWCRqJH@-(?#T1a6Q_?Q~B3ke#sfize)!WaCxNS`D2c6@u+p5
zq5OZm&M|(B{WFD^z5a_td3!r|{gs#cp9>31=jA5pTy}=?pPS7wekc(I8TlyFD{v$4
z3v2onDF)a3*f&*p*>SHvL;1(Q$T8O!7RJ53JUvU_Rfy_o;+Ts?{pTVRTS;>LM~;8A
kug_+pj7;l=y!_XxP{C!dpU9uC{ExSA%o}Y5Y>(Ri2gt%97XSbN

diff --git a/reference/myapp/rx_shiny_data.rda b/reference/myapp/rx_shiny_data.rda
index 446e0f833d6d9b32e823b366a06c09e3975e8902..b847a70661c045937073ee16bdf50629d3410f33 100644
GIT binary patch
delta 5433
zcmV-96~^k<E7U75ABzY80000000Zq^TW{P*c4pr%mTZmf8G9z4G41u;E4I2N`5I?O
zZrNVP@{DbfFBX5r?gslZNoKJRc}@@@FL}ui$a4@Ne_((-2FPL-2!ajVsbW=;#bU89
zQp>CHlmLokT~3{<bE?i)RfpA&mVYo`{y|xiq>Pl!g8w=2DCM3!zCC+IlCo*=l9IAg
z3EZ<CTd%^qg#La5u2<kX1g?ur+Co>=Hq@4Zli(|wMBjh1R%}%V!5NC`|KfWZC(ApV
zwr<q$Tdt;5m3f9z#_M!L3h6!*opUyq+ig@#u8+CNQwdUk=9d=@%)T(abrcRM#^Upi
z;q@E~#gqtHUPbc?W*`GwssOGTDt5WDZg-4|)3YqYsl2eAjS8wUx#_m%Rve>gIEJMe
zmClyh?N)!FC=}hlttqa}(5&yArUvRoll-DyyP+AD=Tx57u9xfcH@b#nR-VpZFY6WE
zJEt36+dF?__TJJ3{}BiKInLi8p|kas3-#LUL-6k+_(w8BQs-|x#8D*FXf{Dk)0(|!
z_q+?|s<ZW4<=n0M;@ri0?fm%*Ao_Zlm9(XH43~dtjFM$Lo%^8LZ6q6C3P*6FB2JWk
zS__`$$<w_h@>E|Ts`8NN9N>&-3S^KxLk2PJbdWosU7yEsS;Dd3F66~;E<y>m3MmX!
z7R%!ulGYa!=}OIByKP#8PTr3rk!AFzZ|Nj{KMK**wzl?}s&PWw>n;wRL7^SkerwNF
zhlGFcL*d@K;i=JrOv_|4!=l&PuC0@lhocNc;@(4Xpu9ELZa+61bko(KoDX6tT80-(
zb%-w8Y|up%P}FUrAn0{PZAB~cNcD_plI5Ngt$xQdJ4Up!s_U99>vqo~jZj44_ish1
z-?jAHNPYMBu~HD(76HEsRAnOmP)PYM=(T_5CeaZ}s6|GcMMjY2m<V&U@UoQH$}xfM
z=Zfm^03jV6BA#iUU{XXRnNS#6N(`A$06Dtp%dtjJ2hE*D4GnvKn(mc*bU;JQNsML?
zrm3+E>w{_OE2f3(x1W<$!d8hqll!z3+Fa_-Z=(#MUQ_0j>Q(%9X5I6;?zKv#1uTDK
zuc2smr((4`=B7jJN*5>>k%_ETuQZ!gwbraIR*jl|rEz8PQlnwis=C^&HH^y)pQOwJ
z&=}XWgTd7S?M%eq(HE)OY-bGhtDT~nfcH!{5&S!Zn+Cp5;V_am`0Mxkzh;6`9C1t$
z)-0_<cusEwy%3DzH-k}J-3<F@5xjrV{WZU?y7hoU9^?z(O6cu&oAjZA1v<g1*&(T2
zv(pSB^WM5+02oG0=t|q%kfJ&>&AF$VPadb3J|)l1>GKPVn$}dchPtqL_41{rQESdM
zu3pl$#f2u(B-yyCRrSjYH65g!yQJ3^fnr=%=NpTcFX?kv8jA_)Uu>FfV-0_50q*y{
z!s%;)UPN*o3C2v-Q9ER0$b>^;&h^5vCFeP6Fnr`XR3FQLp)nY6oB9@cc8wtYlut<k
zW036-f{Y?h!LF;e6}rr#Wf=OZ?F3@9?o*>$(bxn$g<>uHS#QC>fSu<i5gZu_a0Qxe
zwdDq!hIP)W?O<rl^z>$+1fYN1yMz|%Xi;dTwZR}OMWNL`XVV{yo<gI71VBOgJPX7#
z05W6zm6rqxo-hgy^j-j80el7U6~I>jUjck20el7U6~I>-8~7ez(G(x+&craTqA1|@
zJFG}=!yv3U1>B2{;2t7_>x_g}mf<^GdC%i~u-0VN8NUA(D<Bm=#m0Z9qVXYgSL!NU
zehw0Yzk_|8=oRM<2pxj)Ut^+QhuWX<EAkS(r#V+9q4{QvO<=)Ys248ilg1`0D~%@u
z0-FkKDzK>}u&KbN0-FkKDzK@*raQu>k?89?SO{fE5flR|E=Bt$>m@?PM-(8>m8Nbw
zfnw0cG}p8pEf5~a@=<?ifx!^d=(RXHmJg`5NdYO1S1cs>l(M~wrUp5(ssm)jZ*SsJ
z5_9#y6e3dtA8PM``M@DMUfMfTfJg%t6(SNQZw&D=zDbJj1H|21z^Ht8HXJB5&%6^f
z9v<?SS605ST6)_^Xc0Prze3MaD1Zcat)`vSqB!~yVH0N`*7tw16pqbyRc}4Se;~%x
zR&0Bt*A0!D&tU*bo5b`_`JHohtJARCNnItZK}^;qrJ?FKduCftnoD<F*u8(p%351>
zG^&;<DePPCARU!DQnYWqkMuI=Uoc%3Uq`Ucz%)gira;BPGEahQAHGlFG_d?*I1GZk
z<KQ}g<EA;Z-@<=!Gh@;|#Lzy-q5U?FJB;HF<2q(=x=+U_?_H+6qnz^o0LQ(D+p~e=
zkKy<aaQypJ-sAZF1h`J)_meoCjOBDn8f(l8bRzwSvr=SqJ2vX-NC8ZQJYdv0*(E7D
zDAMH8N#r_UghP&xN#tlU8H&traa(cgckNn~V%SNvwwZqrm34al$U+w!#I=p4hvR0b
z@M+i6!AwX7SGVUenF*JSWs=#p&L`9&*HvB5U=kM+5|<qH#bdj^t~yMHVnT-0sibyL
z?4Fkpmq+!bD#~lEjB9i`jg=}V<Rh$>C{`(!PcWwVB=i0qMqN&kce-#WDZ_CFOR6jl
zmejsz{AqtSekvM&hK-+&#-C*>a*lbAG&^9P-+UK%)g9W_`=x}gV>Y8qyiYSVXs|gT
z?Vj|Cn(W#=M`QaZDSJj5EV#F%!Gb%O&>AH;tprmN-%XY)8AZEoCJq}#&2G2Vu1n(S
zDsd&m)@@LGACmYy7Xd6$+P<qem*tY80XWWZoy30tEa-4TL0Q+cyU{`dI}$p=0EDmL
zjPxXkEf9nsry(>n(F7JuVJ0LTmqGVqbb11{voFJR_!<XZN9bZYGI>vr$ya)^e4QY4
zvp5f(4;wZ%m>9LMF?(!$%z&kE4}=+VRDGwUI7Ec?JBbM&$9$ck+9QwYpbr61uOsw|
z3G{#H{8^m8fZos0`N?$fFiv|H=Swqri|E}STEYE%Uy4u5h)kHjvE)N&7GK5*?S$@(
z#2GeHc+@9_P7N<~M@+zmEOD5L8C94NGB^pX*(dExYo?^_YYkpXC0cV)Xw5d5NMY~*
z8+1RuAK)-Y8~3kq+$0p}Eoo;8bP&DI()oV}4Esu-$~UQD|8*-!<uxnF9BDL&W=3e_
zDxH3UwzWp+9v+R4V|4ln@-a`he+Sd!EgQlsM_q*hS9!=aJ>;4ha1En<aL9Fd!1WHk
zj!@UI@{SI<-b=U${r3|t7$=W*ZF@Plpak<gloeUsn?2XGT5?_1>`oWJcITS>)ysds
zZE2c(?n$F(c|EyWRTh-F*=o;!t$tpsDRT?waqw3!zr;Sz(46@$BxFsUm@m*UsSQYQ
z_U_4w-SsMuDhn`crENCAvuSDVo^G6!XO}-(zIF4-9a#n`D?Qh#v~2+JNdg%B@x1I)
ziLDo}c}AyOSuq=q>THo^<B;;GS^R(PGAduQYigjx7AUy}?wX8U(dAhiKiXx#X70M`
z7`i;W!l@wm4<o_~`2mE08&HyG+l{vBf*Y9CsoDm3Xf<sIZo7fU4!Hibr|KQmTOUwP
z`w`L8i3v8g1D%KaY&|!8(?kEBa4~_^zp~<N-ZU+^rNn5j6x+E$`#r)TNjQHbg`0!I
zAxStS35O&p_9q;YghP^WND>Z7!XaspLlU(CsdZS!k8;p-!?Y@oVtM_qW+Qq3<vXaI
z<ooJ|0c~~p!M!_c_1h~;cUEt_<_U{m^vd0vt2a)ayScKo>OY^CPn}y?dazW#7rfr~
zA7UrAD7w*PoqeR3QxIq#AL)NYJ9-4&5-%U-=p7`~H{Pg6YCWS+-ET#8Ju#gT(-|?H
z!3JDRXWpohPfTaTbjH^bF`a?A#B?U$bXULqjqY+-#tV=@icK;o2#ai3jiLOmFHE2K
zLiOsU=H*(odim1g!WxTY(~2pfO=qeTGATYAk|LvYo(m`_prC+)0t$Z$C@2XinD9(D
z5&RQSkaHJMP(Z;r6wE8A8IfKb#i6I1&9z}7Prx##N6AX^o*yYOWi)3;%D`CGF*1ei
zUKuG<%tl`rDH}a7jMfH}wI7YM7$le<tLY+Hua1;;0Jyj@E&(0_sf|(ZC`>ssQci+p
z&W)5h1U#b&kU~YH6%c<yr_pqZ##f`MNDPFX9VuIY)JChg59fB*+>l{Q;P8p#@G<Oo
zU{gv8+2&(x(T=`h4{0(Ra&24NU<&&F1WwY$bR&gLDG&hEk*}Z>`Jy|LZl9vI+xN-A
z0Ac|{EMSNQjBnY%DHbruGr5Ze46%TL-oye18V$t)#<mwQs%(F<37wAwX)7%4$^0+_
z<0S_j_9p3N$ZX%;F#r)$V~Qv_n*n-d)`99ne=mdCL%eTz>o)5p_;yf$lAJfM+;Qw)
zm&qzIsSc1jz|(+lS<j{TFXingNtTVKpCa}X9=Kl#3@$LZz~BOd3k)tWc)~N?MDR~w
zaL!#|aDl<&3%-9Dgk^KmB=IcuJ&aF2iXgE?O!&6uR0IN|FJKUV^hY;IzHh--GVv<m
zZ*x{L%*^z0npa_~z+-B&Z_(wGuWXp|ANRwIECF%;u9SFSgZScIKzae`1*8{{-v1Mj
zUO;*Q=>??UF{J;teesTpdl@gT5R|avpX0#kP(Qw(Ac}v)U*{+RgR7Vb`NAI@$n`lh
zggk}x!Sg=+oTkE_;OB$Pj3YXw0`)TBVdtczQK>&2M0FZgp%W-tdR*eq1`#j%gqL@7
zBc-HoSx+o<mh-s0|B%Cbwm&|d?4UL%Xu*EYm}Gx6h-?l)fU${wGLTo(eX6$2Z8YVl
z9I_vQund1K_C1OO2amsx(N8(xIz42Q1%qdF--`m}T`yFBpsxVFuzO(TF!hj>1u)_@
zk!u^x#H4<Eh7w+^Jq|y@8;7a~Z0&v3Qd>T|IQ>#zO!&owUrhKh#1#|%JvYoF<|HxU
zC*l(memshZ3BQ=|@5Ko}^U+3kqqXYw{Jxg7GnRi>{F;Rd#`v@uTraIMngD(j9*7p?
z=x%iNHy6#XvdQp5NAwts7wm$`!gk;`v+3z~ICKPEHize0|A#n>q{|~!l_J}iFvVlC
zwVAME8lh5vDC3V+Elwt62qsyD9WOb}H2JY6WGIK0pS1l7*G3|Q%V=DsiihnT7%mjM
z6n%fa8GVIKl&46hU}e*c<N`{GQ8$yL>OAoTC}Ct#fWfO_yR6)kl9%8F<beaHAn`Kj
z`oguSkGpVg#;L&!QcQ&=mS%{tRm*|FB5)hPcqZn)nRPrHDuRDX%pd*Jzy8nP{qukQ
zf=#2+X>M)()mL*rzs3$+9iFpl4nbI^KQ@0YCfN}*pLI6Z+;!91T7zI%abg*Bz}k2&
z26G&MhW@_y-{dJ`@<xF~Dhr8XOL*8`A{L6@`}cqQ&p%oE>E#b=U&6m%eYpHCx}Qc2
zeCYf#!V<ZEXmVK$X55_FbvO$T=GozNe8`NOt;IgyS^_h*O08_5sN*O7<~jLY`J^V_
zH1GkA<?s%g<Roj;2qogzydB^PDW>`yvxbdVgu(wP%d%Xr$&*bGFn<j4Q5m4IgXw{6
zK->U}2Qsi~271FX*W)LLO%XQ`_WllQRZ`LCC~o+8`4*U~SWUAPdVPE=n){p9e_!j6
z#~s4i(Vu5A;=i{Jj23sQfrs*_?Vznq)H*RvToJoD1qg#GDNP5#Uyj4qqrMa~mF%MU
z2jn?0Q_iLoE|b3sQ-8wi1e-dK3^Ws3ayFN@tQGV_m6*im{U~#klb^pwMWp;2oD6o$
z_c?9wnl8xVr$hEsrl{x8%0Pe&rZcANxe~YW{v6*+Yy$@^(lQ9piKryr&V#IV`;>q<
zteYIs!|+dVmC^}Ss`WVqA3uA_pH`ohO;@&Txoul5!;#%yw;c;+ss^5D&GWP^CTs<&
zSH91Q3TDS>1tF+)O{prAm=QQ9y%9W}$V}vO+D#{6{y`u#&FFIIP{o^b&<sx#U|3-4
zJ1K`0n&xDa_Yo`tOh1z)5;uQytjAkj){!0zK%N87!-}rw+l|i*%?pcV*YXFh)fb%2
zRm2$K_~(6;<u7^OCy8z^xHd5SXYtQlG2e8<{BVdr&vteoeoCcbw;Q6KFR=d&o)pC8
z=Z_ORuyhn7>j@q~*Wa-3BUk}EX6g;Yp7f)gB6PlYrNjdexZYk+m$844(%aPGUvjjj
zndB^&*E-$%Ee9Beq!i)Af$^h4lSTd>4V8%|`Xc7yIoT;PEJJ5`+2&U&Y!g1+h0Y$f
z7mz1D)nZWRf7T!WU<w|OS9q$I`UC|oQIQ)w<qN#`<7(+rx=Tmk!HVuq3`?QEsL~B%
zzhD{W^N~RFCsa-LQ_p{A`_Y>Igx%In;w=aFqdjRfndY6%yC^j#S?Uicf?|tQ_64Is
zUyPA9l<Emy;5klU_Oxl89=!z8pnt>J6CYH|12kQdh88QBuee?`{}h{tGcNyiA1}wH
zq+K6lV&h#*6|<OIpQFs~a<qMOJ|iW92ktzYaXzP~9q}HX#mav%XaymPXuuGi3@(!)
zS%CP)0FS@uhYY+Xw3zgfW!&PLE_8J_cGJJ53<js56rRb6;(2P87?U3y)`atmaDEXR
z4`Sm%Y&?jK2eI)WHXaC&7aI>^;{nHhGd3QC^NVnPkpu=67*t?Tfk6cZ9X^H%3@Ql>
zDlq6IF{p5Uk$!*w&Mz>*k0!L;`s0Mnh_Kxos%tP|)9}+%*kW#1)u?^ps7Qn3a+n>r
z?@~p6Nj?7se%3jAk&!)<VCf%7SJA+DM)q~}KC?Mcaty5pI;QV{lXOJZl)G-fc&0j*
zX|-HN134Rpp+AIe!Lx5XP;DNU)IQ~p*+wlYTH8<^fLnhO_az7BFDcIEGhmpXQ9C4h
zf~2`p-ACK-XjDN7j<dv+0}l;WfF#R-jtMZC)lxdakNyo%x#>|`iT<H&QvOmubai=*
zpEJCz*BvnM5B+#~AA|b?PT6qhY;_PHLpIc>*-D4wK#-LhgSd*AM~GNT-XSC<@pcL!
zsej~cFJ^y~&ht?iCIehBHVrzkiQR>zw$>wG&5wJg*El<kFR3&<SC$s9Tpag+9hfKi
zJ1Izx58#oFgU0q@rvog7Way?CZx5MilQ;2b;lJm|nVUUg%80YrH|=~uSk808_v<`b
j)r(;n>d06ZMnaC5Q)itpJUD+)diDRXgvhG7@Bje-Hg{gS

delta 5416
zcmV+@71!$2E7vP8ABzY80000000Zq^TW{P*c4psOZOPWyp0Q_Qk7=#%Ua{3J$=5hL
za?AEQmd6*7FBX4Ac7uJHWM`3wJSPZ{m%QW$<T(hCKQKTZ17tA^EP@T(sbW=;#bU89
zQp>CHlmLokT~3{<bE?i)RfpA&mwzx{`awyOq>Pl!g8w=2DCM3$xjlPDlCo*=l9IC0
z1h{A0wqAjE3H|*VT(7}(2wWGLw1tkUZKzEHC&5=ViN1ejt=Os#f-@A=|Hbz-PL_8z
zZQZEiw_H`JDDw=ZjMwgj6w*B=I_GRIw_2!}Tn}@TrxK+8%r7q*n0;w_>nI#jjK$|2
z!|OU0ikTo}c@@non1KvzsRFpBso3T6y4^O)PS>&wr~J}(Hp-~R<fhx2TXu|w;TV=?
zl-pZsr&E81qEK}Iwx+l?L$kJXni{AVP4bIc^@e6xo>P8Sy<V!#-{=^QS$;Nuy`-0Q
z@0@OQZ14Py*?UVD{6`$@=Qw|ZgwEDhF4U^CkHEi+;2+5hNu9s(2uG1nqtO64O>6d^
z-SsY<tIXD_<#V@ci*pxi)$`{sfavQbR??Q*He7$EF(xeAY2OFcZXwzDQaFMW6>*~U
zvuf}(PoC~Ak*C@UQI$tT=KyC!Qy_!n88V1zr-R%9?b<w!%My<Lb|Ei*a}i3gRY+l=
zvREGPk+i;;NLOlfTP@Qfbn<>2i7cTveM=|t`%#FdwzRd+RE-nbT6b~iGzx9Q_FH?Q
zIwXI59}4%@4Nr|0WLhSZ85X_Pa&4WYJRD^p688a$1Ldu`cI$=Vpqs7+<$M@R(KNhR
zszY?yW}PmgfTC^_1wpSXYBO4q$Es&UlPq_gX!YBk**2n;RbAI?TDQ9vX@nvQzke%A
z{jR0oM(Vr2hn0fJwg~uDpehsbheFDCK(Bu_H;Il=LM<}lEHZ*D#YC8+g_otoR*DI1
zKUY+T`v~di0P#%m1d}2n$%MklQew!20?5%#Uy3z)I%w`JYG~N=({!)gqXQaZPGU5J
zFinkRSRYJFUokCQzx|x761Ga@ncSzO(B@Kqej8;7^{O(bRIcK;GwYt$aj%ujO<;c+
zyLCmg+hwcOHa8t&S2{qsh)iUya;4FzsMSVgv0_y9EA=ajm+Ez+TG7=;wQgKy_#|Z(
zfX29{9Sp7xXlEk+w!TQ!W;<i3U+omt1iWXuiQwNM+%)ig5{Hqr!C$Y}`#UBm#SzCO
zVa?Logy-}|&<nvRelr-w)y=Sf7QufT-Cy(Ts#^;v<UzjhZ34aBZjnAzus|nRHQOY$
zYqlFfWZqkM3;@H330-NK8&Xtfra1Rh<LQInGJQ(ERA0EfsMo92`n+1ztBVT@6?I|0
zT5YITwW?|~NIGM2v2jVev@l<p(=O=?jkzmYb^fw`wX!&;E?&8`n4tc}hS`5I)}R*P
ze(x)sz82_3B-fE(%v2q<O-6=HI3(s=FC1HPo}&iCN3Kovu?!d*g8{dpZ;@x$2+~jb
zloT)q*$yGdDDouix@t?I%Pd-kp|9FbAV%vxHL4YjO~6wq){>v~77Ps7d0`U4k&ysb
zpwUvBZop|+=d9WahSp41Zv=lz0Lr~fXrYc4g;rV%46;%bTJ3W-{lVxNG%82{6qL`i
zKs*B=Gsa(eNub~{qu@aA1@INXR{&oDd<F0oz*iE$R{&oDe5H|r?*SG~@v-i74C5+_
z0&c&<iu5iF!irPCz32$;Au_nmNN8mlzQdLGJkAGeO;(-Z`){xUQt^M2Y<wyjA3}Gf
zro!dtATjtm*u#llbMAo9AsGKPCi-=#{TaU^uh4s%b7c~mZ`#-d7Tkq;;etMCY@)K#
zcrqZcslcWJn@R$k3T!H{slcWJn+j~YBWxOpzP^KnP=*viF`(j7v~RLrB2;`t0rFgF
z>ZTJY25n4p4cpNI;emfFABPqg3^Da?lcQt#fNF~rkkWX?LW0jI+pB15kRz))Kvw+r
zCLSd*R}V}fGBxm__8yoI9HQf;y)y-fG+<F7B4P5z5HI7Kr1(BS+`R>i%6DhOfl~9#
zdqLyjA%A&g<@>6ow~T}qp%eHk^elw}NO0F`*hwvlqaPDCarS>feJ@Ml*lb7j)<gUU
zVoYttwl}(+(5U$w29UHtO#h_cIajyZb-R_+Rl*v?WIdtORsCkyZ0SjJ>9z~I_wQI)
zYpae%)iNc8ed|4>qf%Rn_O0VcFN6LC(`E5>1nUe;Q^aWsR2(dG23-5_eG;dE<sZXg
z5agWz*8v<i#i4)w4vw21k@g{m_CXHqcX8Zd9CsMkF@w{6IzoBxGvyuSl=lZX?gQMO
zbsT>T$A5_9k5hS1;P;c@I*s3Fa5@>w>6A3mm>1|o`VVKN$mn)#*wv8&m<V~msB^MQ
zQgl$H$)%IXwZRC793PX&(PT0dncv{H;@0olwJ61~lW2b}Ga)MLbp4TqE;xv584VA|
zO;h1huBU^UkPNO)*JCmhE*Z-tvu&MEs70=$x}L!#E+izLaMYJi?Ap5OFd2#o8B(W`
z+Cj0qUP4?R)t9O$ueCC+(cv^!s+5qAusT7pO0j&BF~u{?`*#?1IYr*-!l9%L$LTMr
zlGI;P`=WpGr`h<)X#5#AekvM&mZ``&<~`EvfO&rNUEo!BU|a8(61tArh%#}UW~$#{
zb3ocX=@m8EwY!eS_D@pwwA5d4?@0XxcQBzfCg8LZOi6q<S*~Oh?Y5aXY!o%S)lxey
ziKnZ?l@MFEe(ilo;`3YtutaJ3uHsylCln39afW~EBnDtXhZ73Qx}M#M782Nz&=Ce8
zd<AEuCqZn1Ap8Ukp`nQ;uwW81A>p_Tx*wy{6R4ei6{f@2DDXN$7t@i+dwN8^(&Odp
zB%zzddFXuDu(83!sC|vuW8-56EQNa@%#fq%J1NB>BCOvGCVT?(b%tt>Jf?#_1VFux
z&?|o?(4+HbasC2&KSSpy)4{_y?OmKN&EzejcYkOF_wzj|J}o0MVg5#v520Cn6(_V4
zy3-P8*ht|~pA<SdxX>Lj0UL6H!%WPm!i12)acIpxX=hq9DQ#bC@KP$#nu|hfw!uUS
zg9q54`|<q%hdJ7~e~shDp+N6QJ5!*8=zV{d&fjO)S9(;waSi)#T0zQhSV87Uqd_z?
zL?c(|^kcNGRYLdhaC{u2(~psldAj}Em>zH05MDXzD)hNZ1Foq7*L0t25bc8luETw<
z_waRux(1bZbinmN!bRvGPq<*5JlwVI<=lc2%=1uIWOZ+LUDImHHCeOU9RS;%Yw~|r
zuYTFoH2K`qde`#0a;2gyD08!wuK!y3yjoS}7S7|~uU>tLeV(B?^Ib^DnmRFGpkY!Q
zkl^gynX=vS%8$znFl)JG*1@xBX|1kq%*eCLA1~j!`Sgw~gOug2>y%qI0Qe*U41O>#
z`&44<#cQ6??vz)|x}!Q<WZ5{RJZyg!f4hvz*XWoUD6t7ju7bNJV^?%}*2a%^$*-Ba
zt~!P;&#rJP2>!!}utI(SA>am-<k?oerMln-W_7Bz0UnwS+kx9|;IR#^Kkce|TlLoa
zl+$`l^mJ^3jqO0^;XYf}4d3+8zb9NwVD+!0IGZ<33vMYf+AGC&ZqR;@a7cd=4oTtW
zpm0bM4oSixNs9dmha};UBpi~2Ly~Yv>gSL|Z9r;mmhqz;G+j5X^5a-u|Et+Z-hcfL
zYA5-=x?w<DU3z%$&RXsE%F><H8*g~R;upPg_vY%2Q|E53EUo&_=jBu9R+b(v)$Rqa
zxBZ9Mi7kq5G+1XJDdrRen#X@fI?;|ELAS)qhdFu&3H7Zv>XBMcD^&MeQC&|=XT)?y
zOlPnG7t@)yYUC5s88MylwM0y3U@kG82{_%=Z-1k^9G3ACB#>g03<|;`8&+i~zwZmv
zC%#b88<#K5UA<brdexuNL<po6Q$m~0R3~Iod^R9OMrpqgP*6ZY0R?{r6ckWU5>PPV
znQkKZC!iqbE})=*f^jIAS5Pw|y*P?PPdS@wgG8QyWlj&1mE=7?RAS0#&JLA<v8*Fx
z3fsLrRHm4XzA#iadSDo?4Jd0r9A_~|Fh5e$MY3KUDr+Bbabr{hJOol5q25uLa%QNU
z1k0QoDs>2Wh7%x#iiUqHAcRiC=@gBxhEtIk2s=Aewg9OOS8*TC?XI~Y!<N9|W5?km
z*zv%oOekcVkFiBN`i4EE$!y5AY;A)n==)<hNf*<N6gH_q08m4|f>Pv*?nt_QirQ-3
zCkF$F1q`u(Ar>&cWdo;Jz#z}$E*3Dv0tR{$3m9lL6bl&JUci5-u*t@BJ`$v@u(ZeX
z!wig<9CX+lr<WnKeRszIL`;n-V#3)B&?~bJR2%qv8O$EyeZyO~SueqNg94P~ym{r0
zV|P1DR*6Y<fYbq=27JqUA;o_wZ$C-0Y&857v7hk3{YqePfx!g^7Z_Y%aDl-Sp6Mon
ze*%MZ?gE1g3?6@9@Xa7Bo0G<gXQ}UDeDYBQi7jHnw>76C5D<L<gZQIAx=HeV3%-(x
zR|$Wcvx;G6riatK4qF8tQ=5H<E}wj5!<7G|7iMG$i1T-)!~+||7w-bn3rH^@y@2%o
zpMdlN(hEp0ApMRZ{kQFlcU0V~cyWcGgdP7J2Tq6j@%?`UQ6&C4#{@9Aiiwaf{K0`-
zk26EaQ%D~?@59e&D(oqKKFG{CqEjkRF9RNSPD&b<`m=sir(hM@fuf~HCH}l0@uE+7
zc{evwCiE@qsfEsR9+meWa(K`7MyHeQ*9HYG*v}b}?2r19%^?UdGSN@^@@l%z)Rwu8
zru>vc_G5n#mZ8PIN0DIv@%IV(DF<AqM{KfS@Qm(zQJ}ozh3XIV6~GsE53CfX9+0vC
zM!W`cZKIhO*Kf~I!i%-X;m3I6Q1yVVy{}qo(`Of_U+ReoznJig2|tFoV#2@YhIz!C
zBqsbsd}6|nM-egM7Zd)yIN@hL+URUFSG}&^*OGsB#`20^vrxeppEiT*rBy-`z>mWN
z(SjVEjgJ2IqWM)e8D8j!9)t0cT`*bL4%}uoJ^c=cj-bot@LcQv5NDBed8DdRWE&Hv
zctW-|6Lw4^R0<Gf{IROV$%G8SB&)FFC8wDpKh}f{<-qcjwqN1eNQ7_+jjL4gu)PDr
zg<^k~qOUijuds>o6v-5<Y?_f=Kq)cmW^z=GC!PQ$3@r*Uc-3u}m3vb12{-|H=)fsR
zybQX&a4qUV2hPnnHJC<<snEpI3^BH9IWSlRZUY$4#N0Qtj%Pzf@K1^PqksOl-~Ro-
z{P)k<G%B6u*4AHtHTSb??7-FLIjiOngk^vFW5Z&S9YOP1XLHS6H?6HT2!<6WmN5sc
zjpt%8#{p>Q?|c74o+2i17)YeDkSMl<hwUX|q4>T3`1k+%lck?t{;2vT{QK2M%kQK6
zX~e)s&MzV?k^4s`m&IVl&6!<?v+!V^9Ztsw%(&TF?DMT9FjK44$_9!$G2=JS%y%f|
z8NX@Z102iY9W==dYtslN;@7+#;0Y<FdK|NcjaP)h|18O}T&v2HOAs)B3i44Ips|DL
zfoworABzVvuxbW+!!p<7Cx=ZDHxTyzHfvQ<(dQ^`c(8m6%vG$0*$lltz7@^=P3ymJ
zbjafl;q2(ovl#K;TL(sqJJrBLdDM2$)+TD57$>fX-JAr3L6ww-gWxa6;p=fvikV7w
zQT#*l9GEF*Qwo>KUxz7w;dPQtoks?m37v2@m$$4H^h1@H#OJ*zbCi>xzeh!+{A-*H
zcFW_OHh4o9WbxAhdn%LEb7*BCKnBwp)Ad}5+jxJD?-OhT2QAVx2+)bBB;L-0taW;n
zfH<t19MQw@PjHpe30120I0c_Ld&-|ypOsBlwrsg&TTR1}-EOBH17@lQo@mYUv@IrV
z1*%uR$B7DN$7ls1sC8ATD3g{EI0D@vlh6?v3=v>hVCp+AhZLITc$4)JECO^dlOqy0
ze|4-UTOHPs9t=R91JA>XuIJnJ&kW5Ai)GjH`>xd&oXu6l7~%M5J(T4ydEO_9ZZEht
zF#NCLpSNPZ>4y2?0DqqC>_GgKO2ck9L_J?%|LZ>~h|A9(Cw5@zC`Q&3Jc6#jW#31z
z0(ipI8-_jYMLR|4eD6w$2Ox00y`U~*e<7u}sl&hEXiX!@SuU@&JNKIoFbqj4!ifXp
zM};Pf{5={f6HWAG%*AuOQ)F0%&hoO&uT<D3e6|anJ!~%^PkgFHzs~=>H~PU8JRYy`
zR4?@j3S6QhH+aeyc<;y6(xr5lj=+Nz-JKYgLVr=E6UKhQGR)^Af#y%Bn(U{Zf6w-!
zHT?;@t((MK4(>;L(r_}(JDYb=YD{vX*QW@IEmGN+j0Sx%LfTNOCwYPAIDy&IrgeJw
z5=evoEoV=>Uo8*NbV(XotYE(4deQuoY#z?I{5L(k9G8-IeT<2XcQIAWVs3qoGP}#s
z_RaaUln5TU^JvEToSt^Xdw3Qre@DL+geal`Lv%8@Oon6u;u`@x{=645@P^Q0(npqY
zi)*^j)!o=l|CSOMoPtt#CMSyLsaax7esEY5&M(6GMQl8XjR&#uAT}Pv#)H^+AV6Mh
zJcx}49Q)1Kco5Dn!udrK7*t?Tfk6cZ6&Q5z7%DKRBrvGJpyR}#!udt|fBieZzyv=U
z({}5R6E-8lcCV|h!Gul0PfuZsxgAxb_JyM&4UWrUcHF*875OFg{1^CH=j=sB_Dq7M
ze;{2&1LGOl*VTK>=D>twXkE}TeFvPRBeEvlb^GOW)v-*g=`tF~*)R<KA#4ktedB>@
z^SGq;DSym1YEjWzhUx&^f0DQ_**E`$;%q(#hWR<QL!u{0nk&_Pv<;6&6_nsOOH4WN
z&|n2fvK(le0Fzlwr5*g}Umum58n%__AKE76FZDuKmnZl+!`ph@0R#Woi<kER+#hnv
zhC64g{rDKLp+3b{IvfXrtkf99Rm40(#8UDOAt{NsQwT}@6K{Jlf1|Wt48t%P;DWKK
z-+@i+E-baR9{Fm1)HA(C*=c-5rQx}<G<oIXxQFb(JkH-qL2`Tuk8JEWwhuZTU@0U+
zH^pds$V{8OjYkXr14qu>>=9E&oW;Ir=L^DeUJ$<D<k6~L49ieQ#=0;Pa>SfE>xALK
S`GeBy{|9^*f6=+{0ssKP4|qiY

diff --git a/reference/rxCombineErrorLines.html b/reference/rxCombineErrorLines.html
index d003e17ba..f034b8ff2 100644
--- a/reference/rxCombineErrorLines.html
+++ b/reference/rxCombineErrorLines.html
@@ -383,7 +383,7 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-msg co"><span class="r-pr">#&gt;</span>  </span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span>  </span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_be0b5dbb4a0767d8702ddb24bbc603bc</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="font-weight: bold;">rxode2 </span>2.1.3 model named <span style="color: #BBBB00; font-weight: bold;">rx_8d950c43531f6895b2b69a6751925bb9</span> model (<span style="color: #00BB00;">✔</span> <span style="color: #00BB00; font-weight: bold;">ready</span>). </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #0000BB; font-weight: bold;">$state</span>: depot, gut, center, effect</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #0000BB; font-weight: bold;">$stateExtra</span>: cp</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #0000BB; font-weight: bold;">$params</span>: tktr, tka, tcl, tv, prop.err, pkadd.err, temax, tec50, tkout, te0, pdadd.err, eta.ktr, eta.ka, eta.cl, eta.v, eta.emax, eta.ec50, eta.kout, eta.e0, CMT, rxerr.cp, rxerr.effect</span>
diff --git a/reference/rxExpandGrid.html b/reference/rxExpandGrid.html
index ad8c0289d..20b41c813 100644
--- a/reference/rxExpandGrid.html
+++ b/reference/rxExpandGrid.html
@@ -859,12 +859,12 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span><span class="co"># \donttest{</span></span></span>
 <span class="r-in"><span><span class="fu">microbenchmark</span><span class="fu">::</span><span class="fu"><a href="https://rdrr.io/pkg/microbenchmark/man/microbenchmark.html" class="external-link">microbenchmark</a></span><span class="op">(</span><span class="fu">rxExpandGrid</span><span class="op">(</span><span class="va">letters</span>, <span class="va">letters</span><span class="op">)</span>, <span class="fu">expand.grid.jc</span><span class="op">(</span><span class="va">letters</span>, <span class="va">letters</span><span class="op">)</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Unit: microseconds</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>                              expr   min      lq      mean  median     uq</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>    rxExpandGrid(letters, letters) 10.80 11.1305  45.96206 11.3510 11.602</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>  expand.grid.jc(letters, letters) 16.07 16.7410 193.97223 17.8285 18.640</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>       max neval</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>  3295.638   100</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>  4231.945   100</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>                              expr   min     lq      mean median     uq      max</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>    rxExpandGrid(letters, letters) 10.61 11.121  30.09082 11.412 11.772 1713.883</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>  expand.grid.jc(letters, letters) 16.37 17.312 210.03084 17.994 18.740 5885.897</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>  neval</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>    100</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>    100</span>
 <span class="r-in"><span><span class="co"># }</span></span></span>
 </code></pre></div>
     </div>
diff --git a/search.json b/search.json
index 1b0a79558..b821a5c43 100644
--- a/search.json
+++ b/search.json
@@ -1 +1 @@
-[{"path":"https://nlmixr2.github.io/rxode2/articles/Integrating-User-Defined-Functions-into-rxode2.html","id":"user-defined-functions","dir":"Articles","previous_headings":"","what":"User Defined Functions","title":"Integrating User Defined Functions into rxode2","text":"defining models may wished write small R function make function integrate rxode2 somehow. article discusses 2 ways : R-based user function can loaded simple function certain circumstances translated C run efficiently C function define integrate code","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()`"},{"path":"https://nlmixr2.github.io/rxode2/articles/Integrating-User-Defined-Functions-into-rxode2.html","id":"r-based-user-functions","dir":"Articles","previous_headings":"","what":"R based user functions","title":"Integrating User Defined Functions into rxode2","text":"R-based user function convenient include ODE, slower done written C , C++ compiled language. requested github appropriate example; However, use simple example simply illustrate concepts. Now ODE compiled R functions called solving ODE. Since calling R, forces parallization turned since R single-threaded. also takes time solve since shuttling back forth R C. Lets see simple function performs:  terribly bad, even though shuffling R C. can make better converting functions C:  C version almost twice fast R version. may noticed conversion also created C versions first derivative. done automatically gives just C versions function, C versions derivatives registers rxode2. allows C versions work rxode2 nlmixr2 models. function setup advance allow type conversion. general derivatives calculated return() statement user defined function. means simply let R return last value instead explictly calling return(). Many people prefer method coding. Even return function, function converted C. github issue, used function convert derivatives: still helpful functions early returns, nlmixr2 models requiring derivatives calculated non-optimized finite differences occurs. gets internals rxode2 nlmixr2 can see easily calculating derivatives: Whereas originally defined function newAbs() use new derivatives calculated well: circumstances, conversion C possible, though can still use R function. requirements R functions integrated rxode2 system: function must set number arguments, variable arguments like f(…) currently allowed. function given argument single number, function return single number requirements met can use R function rxode2. Additional requirements conversion C include: functions use within R function must understood available rxode2. Practically speaking fun2() refers fun1(), fun1() must changed C code available rxode2 changing function fun2() C. functions can include /else assignments simple return statements (either returning value value line ). Special R control structures functions (like lapply) present. function refer package functions mentioned, return() statement present, derivative C functions rxode2’s derivative table updated.","code":"newAbs <- function(x) {   if (x < 0) {     -x   } else {     x   } }  f <- rxode2({   a <- newAbs(time) }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ e <- et(-10, 10, length.out=40) mb1 <- microbenchmark::microbenchmark(withoutC=suppressWarnings(rxSolve(f,e)))  library(ggplot2) autoplot(mb1) + rxTheme() # Create C functions automatically with `rxFun()` rxFun(newAbs) #> → finding duplicate expressions in d(newAbs)/d(x)... #> [====|====|====|====|====|====|====|====|====|====] 0:00:00 #> → optimizing duplicate expressions in d(newAbs)/d(x)... #> [====|====|====|====|====|====|====|====|====|====] 0:00:00 #> converted R function 'newAbs' to C (will now use in rxode2) #> converted R function 'rx_newAbs_d_x' to C (will now use in rxode2) #> Added derivative table for 'newAbs' # Recompile to use the C functions # Note it would recompile anyway if you didn't do this step, # it just makes sure that it doesn't recompile every step in # the benchmark f <- rxode2({   a <- newAbs(time) }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ mb2 <- microbenchmark::microbenchmark(withC=rxSolve(f,e, cores=1))  mb <- rbind(mb1, mb2) autoplot(mb) + rxTheme() + xgxr::xgx_scale_y_log10() #> Scale for y is already present. #> Adding another scale for y, which will replace the existing scale. print(mb) #> Unit: milliseconds #>      expr      min       lq     mean   median       uq       max neval #>  withoutC 6.724595 6.871821 7.296725 6.924844 7.033558 12.172650   100 #>     withC 1.979874 2.033378 2.336496 2.132534 2.630452  6.166024   100 # Light f_R <- function(actRad, k_0, a_k) {   photfac <- a_k * actRad + k_0   if (photfac > 1) {     photfac = 1   }   return(photfac) }  rxFun(f_R) #> function contains return statement; derivatives not calculated #> converted R function 'f_R' to C (will now use in rxode2) rxFromSE(\"Derivative(f_R(actRad, k_0, a_k),k_0)\") #> [1] \"(f_R(actRad,(k_0)+6.05545445239334e-06,a_k)-f_R(actRad,k_0,a_k))/6.05545445239334e-06\" rxFromSE(\"Derivative(newAbs(x),x)\") #> [1] \"rx_newAbs_d_x(x)\""},{"path":"https://nlmixr2.github.io/rxode2/articles/Integrating-User-Defined-Functions-into-rxode2.html","id":"c-based-functions","dir":"Articles","previous_headings":"","what":"C based functions","title":"Integrating User Defined Functions into rxode2","text":"can add C functions directly rxode2 well using rxFun(): wanted also use C functions expressions derivatives using rxD() function: Removing function rxRmFun() also remove derivative table:","code":"fun <- \"  double fun(double a, double b, double c) {    return a*a+b*a+c;  } \" ## C-code for function  rxFun(\"fun\", c(\"a\", \"b\", \"c\"), fun) rxD(\"fun\", list(    function(a, b, c) { # derivative of arg1: a      paste0(\"2*\", a, \"+\", b)    },    function(a, b, c) { # derivative of arg2: b      return(a)    },    function(a, b, c) { # derivative of arg3: c      return(\"0.0\")    } )) rxRmFun(\"fun\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"modifying-model","dir":"Articles","previous_headings":"","what":"Modifying model","title":"Modifying Models","text":"tasks might want overall model: Change line model Add line model Rename parameters model Combine different models Create functions add certain model features model go model piping functions can use modify models even add functions modify models. cover model modification functions nlmixr2lib","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"modifying-a-model-line","dir":"Articles","previous_headings":"Modifying model","what":"Modifying a model line","title":"Modifying Models","text":"opinion, modifying lines model likely common task modifying model. may wish modify model subject variability add covariate effects. begin course need base model modify. Let’s start simple PK example, using single-dose theophylline dataset generously provided Dr. Robert . Upton University California, San Francisco: believed enough absorption support subject variability can change line drop subject modifying single line. simply type line want model piping expression: expected, line modified. Also can notice initial estimate subject variability dropped since longer part model. reason wanted add back model can modify model add back: modification, eta.ka automatically assumed subject variability parameter. Also since eta.ka mu-referenced rxode2 points . automatic detection eta.ka name follows convention. Parameters starting ending following names assumed subject variability parameters: eta (NONMEM convention) ppv (per patient variability) psv (per subject variability) iiv (inter-individual variability) bsv (subject variability) bpv (patient variability) functioning correctly can change covariate can add type initial estimate later: seen , eta.ka model assumed data-input parameter covariate instead estimated parameter. can also note WT automatically recognized covariate covWt automatically recognized covariate parameter. general covariates typical/population parameters automatically converted estimated parameters based parameter name starting (ending ): tv (typical value) t (also typical value) pop (population parameter) err (error parameter) eff (effect parameter) cov (covariate parameters) notable exceptions parameters like (wt, sex crcl) assumed covariates. don’t want automatic variable conversion, can also use auto=FALSE: case additional parameters (eta.ka, WT, covWt) assumed parameters dataset.","code":"one.compartment <- function() {   ini({     tka <- 0.45; label(\"Ka\")     tcl <- 1; label(\"Cl\")     tv <- 3.45; label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) } mod <- one.compartment |>   model(ka <- exp(tka)) #> ! remove between subject variability `eta.ka` print(mod) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tcl eta.cl    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } mod2 <- mod |>   model(ka <- tka * exp(eta.ka)) #> ℹ add between subject variability `eta.ka` and set estimate to 1 print(mod2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.ka #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.ka    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tcl eta.cl    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.ka ~ 1 #>     }) #>     model({ #>         ka <- tka * exp(eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } mod2 <- mod |>   model(ka <- tka * exp(eta.ka) + WT * covWt, cov=\"eta.ka\") #> ℹ add covariate `eta.ka` (as requested by cov option) #> ℹ add covariate `WT` #> ℹ add population parameter `covWt` and set estimate to 1 print(mod2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  covWt  #>   0.45   1.00   3.45   0.70   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tcl eta.cl    id            #> 2    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         covWt <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- tka * exp(eta.ka) + WT * covWt #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } mod3 <- mod |>   model(ka <- tka * exp(eta.ka) + WT * covWt, auto=FALSE) #> ℹ add covariate `eta.ka` #> ℹ add covariate `WT` #> ℹ add covariate `covWt` print(mod3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tcl eta.cl    id            #> 2    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- tka * exp(eta.ka) + WT * covWt #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"note-on-automatic-detection-of-variables","dir":"Articles","previous_headings":"Modifying model","what":"Note on automatic detection of variables","title":"Modifying Models","text":"automatic detection variables convenient many models may suit style; like can always change using options(): option disabled, variables assumed covariates promote population parameters ini block last example option enabled none variables starting t added model course use turn back : can use withr::with_options(list(rxode2.autoVarPiping=FALSE), ...) turn option temporarily. don’t like defaults changing variables change well rxSetPipingAuto() example wanted variables starting ending te can change : requested population parameters starting te added ini block. want reset defaults simply call rxSetPipingAuto() without arguments:","code":"options(rxode2.autoVarPiping=FALSE) mod7 <- mod3 |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE)  print(mod7) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tcl eta.cl    id            #> 2    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- tka * exp(eta.ka) + WT * covWt #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } options(rxode2.autoVarPiping=TRUE) mod8 <- mod |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 #> ℹ promote `tkin` to population parameter with initial estimate 1 #> ℹ promote `tkout` to population parameter with initial estimate 1 print(mod8) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  temax    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } rxSetPipingAuto(thetamodelVars = rex::rex(\"te\"))  mod9 <- mod |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 print(mod9) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  temax    te0  tec50  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } rxSetPipingAuto() mod10 <- mod |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 #> ℹ promote `tkin` to population parameter with initial estimate 1 #> ℹ promote `tkout` to population parameter with initial estimate 1"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"adding-model-lines","dir":"Articles","previous_headings":"Modifying model","what":"Adding model lines","title":"Modifying Models","text":"three ways insert lines rxode2/nlmixr2 model. can add lines end model, expression beginning model controlled append option. Let’s assume two different assays run compound noticed different variability. can modify model adding lines end model using append=TRUE: Perhaps instead may want add indirect response model addition concentrations, can choose add : append=lhsVar lhsVar left handed variable want insert new lines: last type insertion may wish add lines beginning model using append=FALSE:","code":"mod4 <- mod |>   model({     cp2 <- cp     cp2 ~ lnorm(lnorm.sd)   }, append=TRUE) #> ℹ add residual parameter `lnorm.sd` and set estimate to 1 print(mod4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>      tka      tcl       tv   add.sd lnorm.sd  #>     0.45     1.00     3.45     0.70     1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                cmt                dvid* #> 1   cp ~ …  cmt='cp' or cmt=3  dvid='cp' or dvid=1 #> 2  cp2 ~ … cmt='cp2' or cmt=4 dvid='cp2' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tcl eta.cl    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         lnorm.sd <- c(0, 1) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>         cp2 <- cp #>         cp2 ~ lnorm(lnorm.sd) #>     }) #> } mod5 <- mod |>   model({     PD <- 1-emax*cp/(ec50+cp)     ##     effect(0) <- e0     kin <- e0*kout     d/dt(effect) <- kin*PD -kout*effect   }, append=d/dt(center)) mod6 <- mod5 |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 #> ℹ promote `tkin` to population parameter with initial estimate 1 #> ℹ promote `tkout` to population parameter with initial estimate 1 print(mod6) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  temax    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3           effect #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         PD <- 1 - emax * cp/(ec50 + cp) #>         effect(0) <- e0 #>         kin <- e0 * kout #>         d/dt(effect) <- kin * PD - kout * effect #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"remove-lines-in-the-model","dir":"Articles","previous_headings":"Modifying model","what":"Remove lines in the model","title":"Modifying Models","text":"lines model can removed one 2 ways either use -param param <- NULL model piping:","code":"mod7 <- mod6 |>   model(-emax) #> ! remove population parameter `temax` print(mod7) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3           effect #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         PD <- 1 - emax * cp/(ec50 + cp) #>         effect(0) <- e0 #>         kin <- e0 * kout #>         d/dt(effect) <- kin * PD - kout * effect #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } # Equivalently  mod8 <- mod6 |>   model(emax <- NULL) #> ! remove population parameter `temax` print(mod8) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3           effect #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         PD <- 1 - emax * cp/(ec50 + cp) #>         effect(0) <- e0 #>         kin <- e0 * kout #>         d/dt(effect) <- kin * PD - kout * effect #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"rename-parameters-in-a-model","dir":"Articles","previous_headings":"Modifying model","what":"Rename parameters in a model","title":"Modifying Models","text":"may want rename parameters model, easy rxRename(). dplyr loaded can even replace rename(). semantics similar two functions, assigning newVar=oldVar. example: can see every instance variable named model renamed inside model ini block. completeness can see dplyr verb (since S3 method):","code":"mod11 <- mod10 |>   rxRename(drug1kout=kout, tv.drug1kout=tkout)  print(mod11) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>          tka          tcl           tv       add.sd        temax          te0  #>         0.45         1.00         3.45         0.70         1.00         1.00  #>        tec50         tkin tv.drug1kout  #>         1.00         1.00         1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tv.drug1kout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         drug1kout <- exp(tv.drug1kout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union mod12 <- mod10 |>   rename(drug1kout=kout, tv.drug1kout=tkout)  print(mod12) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>          tka          tcl           tv       add.sd        temax          te0  #>         0.45         1.00         3.45         0.70         1.00         1.00  #>        tec50         tkin tv.drug1kout  #>         1.00         1.00         1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tv.drug1kout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         drug1kout <- exp(tv.drug1kout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"combine-different-models","dir":"Articles","previous_headings":"Modifying model","what":"Combine different models","title":"Modifying Models","text":"can also combine different models rxAppendModel(). general need variables common combine. generally want models link . example pk pd model done renaming cp first model ceff second model: get error try combine models without variables common: want combine models without respecting variables common, can use common=FALSE:","code":"ocmt <- function() {   ini({     tka <- exp(0.45) # Ka     tcl <- exp(1) # Cl     tv <- exp(3.45); # log V     ## the label(\"Label name\") works with all models     add.sd <- 0.7   })   model({     ka <- tka     cl <- tcl     v <- tv     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl / v * center     cp <- center / v     cp ~ add(add.sd)   }) }  idr <- function() {   ini({     tkin <- log(1)     tkout <- log(1)     tic50 <- log(10)     gamma <- fix(1)     idr.sd <- 1   })   model({     kin <- exp(tkin)     kout <- exp(tkout)     ic50 <- exp(tic50)     d/dt(eff) <- kin - kout*(1-ceff^gamma/(ic50^gamma+ceff^gamma))     eff ~ add(idr.sd)   }) }  rxAppendModel(ocmt %>% rxRename(ceff=cp), idr) #> ℹ parameter labels from comments will be replaced by 'label()' #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd      tkin     tkout     tic50     gamma  #>  1.568312  2.718282 31.500392  0.700000  0.000000  0.000000  2.302585  1.000000  #>    idr.sd  #>  1.000000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3              eff #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                 cmt                 dvid* #> 1 ceff ~ … cmt='ceff' or cmt=4 dvid='ceff' or dvid=1 #> 2  eff ~ …  cmt='eff' or cmt=3  dvid='eff' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>         tkin <- 0 #>         tkout <- 0 #>         tic50 <- 2.30258509299405 #>         gamma <- fix(1) #>         idr.sd <- c(0, 1) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         ceff <- center/v #>         ceff ~ add(add.sd) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ic50 <- exp(tic50) #>         d/dt(eff) <- kin - kout * (1 - ceff^gamma/(ic50^gamma +  #>             ceff^gamma)) #>         eff ~ add(idr.sd) #>     }) #> } try(rxAppendModel(ocmt, idr)) #> ℹ parameter labels from comments will be replaced by 'label()' #> Error : not all the models have variables in common (use `common=FALSE` to allow this) mod2 <- rxAppendModel(ocmt, idr, common=FALSE) |>   model(ceff=cp, append=ic50) # here we add the translation after the #> ℹ parameter labels from comments will be replaced by 'label()' # ic50 line to make it reasonable  print(mod2) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd      tkin     tkout     tic50     gamma  #>  1.568312  2.718282 31.500392  0.700000  0.000000  0.000000  2.302585  1.000000  #>    idr.sd  #>  1.000000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3              eff #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                cmt                dvid* #> 1   cp ~ …  cmt='cp' or cmt=4  dvid='cp' or dvid=1 #> 2  eff ~ … cmt='eff' or cmt=3 dvid='eff' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>         tkin <- 0 #>         tkout <- 0 #>         tic50 <- 2.30258509299405 #>         gamma <- fix(1) #>         idr.sd <- c(0, 1) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ic50 <- exp(tic50) #>         ceff <- cp #>         d/dt(eff) <- kin - kout * (1 - ceff^gamma/(ic50^gamma +  #>             ceff^gamma)) #>         eff ~ add(idr.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"creating-more-complex-model-modification-functions","dir":"Articles","previous_headings":"Modifying model","what":"Creating more complex model modification functions","title":"Modifying Models","text":"pretty flexible, may want even , helper functions help create functions . discuss extract model function update . Lets start model: Lets assume moment want remove eta cl. First probably want get model lines. can modelExtract(): Now want worry cl line, can subset : Now wish change line “cl <- exp(tcl+eta.cl)” Note functions modify ini({}) block. may modify ini block first make valid rxode2/nlmixr2 model. particular case, using model piping easier, simply demonstrates two different way extract model information way add information final model. methods can tricky using model parsed correctly. means make sure parameters endpoints follow correct rules","code":"f <- function() {   ini({     tka <- 0.45     tcl <- 1     tv <- 3.45     eta.ka ~ 0.6     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl)     v <- exp(tv + eta.v)     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl/v * center     cp <- center/v   }) } totLines <- modelExtract(f, endpoint=NA) # endpoints should be included  print(totLines) #> [1] \"ka <- exp(tka + eta.ka)\"                    #> [2] \"cl <- exp(tcl)\"                             #> [3] \"v <- exp(tv + eta.v)\"                       #> [4] \"d/dt(depot) <- -ka * depot\"                 #> [5] \"d/dt(center) <- ka * depot - cl/v * center\" #> [6] \"cp <- center/v\" clLine <- modelExtract(f, cl, lines=TRUE) line <- attr(clLine, \"lines\") totLines[line] <- \"cl <- exp(tcl+eta.cl)\"  # For now lets remove the entire `ini` block (so you don't have to # worry about syncing parameters).  #  ini(f) <- NULL  model(f) <- totLines #> ℹ parameter labels from comments are typically ignored in non-interactive mode #> ℹ Need to run with the source intact to parse comments print(f) #> function ()  #> { #>     ini({ #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>     }) #> } #> <environment: 0x55b14af6e288>"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"modifying-initial-estimates","dir":"Articles","previous_headings":"","what":"Modifying initial estimates","title":"Modifying Models","text":"common items want initial estimates : Fix/Unfix parameter Change initial condition values bounds Change initial condition type Change labels transformations Reorder parameters may wish create functions; discuss .","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"fixing-or-unfixing-a-parameter","dir":"Articles","previous_headings":"Modifying initial estimates","what":"Fixing or unfixing a parameter","title":"Modifying Models","text":"can fix model estimates two ways. first fix value whatever model function, done piping model parameter name (like tka) setting equal fix (%>% ini(tka=fix)). full example: can also fix parameter different value wish; similar can specify value fix inside fix pseudo-function follows: %>% ini(tka=fix(0.1)). fully worked example :","code":"f <- function() {   ini({     tka <- 0.45     tcl <- 1     tv <- 3.45     add.sd <- c(0, 0.7)     eta.ka ~ 0.6     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl)     v <- exp(tv + eta.v)     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl/v * center     cp <- center/v     cp ~ add(add.sd)   }) }  f2 <- f |>   ini(tka=fix) #> ℹ fix `tka` to `0.45` print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.45) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f <- function() {   ini({     tka <- 0.45     tcl <- 1     tv <- 3.45     add.sd <- c(0, 0.7)     eta.ka ~ 0.6     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl)     v <- exp(tv + eta.v)     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl/v * center     cp <- center/v     cp ~ add(add.sd)   }) }  f2 <- f |>   ini(tka=fix(0.1)) #> ℹ fix `tka` to `0.45` #> ℹ change initial estimate of `tka` to `0.1` print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"unfixing-parameters","dir":"Articles","previous_headings":"Modifying initial estimates","what":"Unfixing parameters","title":"Modifying Models","text":"unfix parameters similarly fixing. Instead using fix keyword, use unfix keyword. unfix parameter (keeping value) pipe model using (|> ini(tka=unfix)). Starting fixed model fully worked example : can also unfix change initial estimate ini(parameter=unfix(newEst)):","code":"print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f2 |> ini(tka=unfix) #> ℹ unfix `tka` keeping initial estimate `0.1` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f2 |>   ini(tka=unfix(10)) #> ℹ unfix `tka` keeping initial estimate `0.1` #> ℹ change initial estimate of `tka` to `10` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>  10.00   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 10 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"multiple-parameter-assignment","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Multiple parameter assignment","title":"Modifying Models","text":"can also assign multiple parameters providing : vector/list multiple lines piped ini() block Using covariance matrix case vector can specify pipe model. example: can also added multiple lines commas separating estimates: also use matrix specify covariance:","code":"ini1 <- c(tka=0.1, tcl=1, tv=3)  f4 <- f |> ini(ini1) #> ℹ change initial estimate of `tka` to `0.1` #> ℹ change initial estimate of `tcl` to `1` #> ℹ change initial estimate of `tv` to `3` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.1    1.0    3.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # or equivalently  ini1 <- list(tka=0.1, tcl=1, tv=3)  f4a <- f |> ini(ini1) #> ℹ change initial estimate of `tka` to `0.1` #> ℹ change initial estimate of `tcl` to `1` #> ℹ change initial estimate of `tv` to `3` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.1    1.0    3.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # commas separating values: f4 <- f |> ini(tka=0.1, tcl=1, tv=3) #> ℹ change initial estimate of `tka` to `0.1` #> ℹ change initial estimate of `tcl` to `1` #> ℹ change initial estimate of `tv` to `3` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.1    1.0    3.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # multiple lines in {} f4 <- f |>   ini({     tka <- 0.2     tcl <- 2     tv <- 6   }) #> ℹ change initial estimate of `tka` to `0.2` #> ℹ change initial estimate of `tcl` to `2` #> ℹ change initial estimate of `tv` to `6` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.2    2.0    6.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.2 #>         tcl <- 2 #>         tv <- 6 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } ome <- lotri(eta.ka + eta.v ~ c(0.6,                                 0.01, 10.1))  f4 <- f |> ini(ome) #> ℹ change initial estimate of `eta.ka` to `0.6` #> ℹ add covariance between `eta.v` and `eta.ka` with initial estimate `0.01` #> ℹ change initial estimate of `eta.v` to `10.1` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka   0.60  0.01 #> eta.v    0.01 10.10 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka + eta.v ~ c(0.6, 0.01, 10.1) #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # or equavialtly use the lotri-type syntax for the omega:  f4 <- f |> ini(eta.ka + eta.v ~ c(0.6,                                   0.01, 0.2)) #> ℹ change initial estimate of `eta.ka` to `0.6` #> ℹ add covariance between `eta.v` and `eta.ka` with initial estimate `0.01` #> ℹ change initial estimate of `eta.v` to `0.2` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka   0.60  0.01 #> eta.v    0.01  0.20 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka + eta.v ~ c(0.6, 0.01, 0.2) #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"single-parameter-assignment","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Single parameter assignment","title":"Modifying Models","text":"simplest way change initial parameter estimates simply use ini(parameter=newValue). can also use <- ~ change value: fully worked example showing three types initial value modification : can change bounds like model specification using numeric vector c(low, estimate) c(low, estimate, hi). worked example: Note changing parameters default values might show parameter printout:","code":"f3 <- f |>   ini(tka <- 0.1) #> ℹ change initial estimate of `tka` to `0.1` f4 <- f |>   ini(tka=0.1) #> ℹ change initial estimate of `tka` to `0.1` f5 <- f |>   ini(tka ~ 0.1) #> ℹ change initial estimate of `tka` to `0.1` print(f5) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f |>   ini(tka <- c(0, 0.1, 0.2)) #> ℹ change initial estimate (0.1) and upper/lower bound (0 to 0.2) of `tka` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- c(0, 0.1, 0.2) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f |>   ini(tka <- c(0, 0.1)) #> ℹ change initial estimate (0.1) and lower bound (0) of `tka` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- c(0, 0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f |>   ini(tka <- c(0, 0.1, 0.2)) #> ℹ change initial estimate (0.1) and upper/lower bound (0 to 0.2) of `tka` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- c(0, 0.1, 0.2) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # Now reassign f4 <- f3 |>   ini(tka <- c(-Inf, 0.1, Inf)) #> ℹ change initial estimate (0.1) and upper/lower bound (-Inf to Inf) of `tka` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"changing-parameter-types","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Changing parameter types","title":"Modifying Models","text":"can change parameter type two operators either using -par convert parameter covariate ~par toggle population individual parameters. example 3:","code":"# Switch population parameter to between subject variability parameter: f4 <- f |>   ini( ~ tcl) #> ℹ convert 'tcl' from population parameter to between subject variability print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v tcl #> eta.ka    0.6   0.0   0 #> eta.v     0.0   0.1   0 #> tcl       0.0   0.0   1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>         tcl ~ 1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # Switch back to population parameter f5 <- f4 |>   ini( ~ tcl) #> ℹ convert 'tcl' from between subject variability to population parameter print(f5) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd    tcl  #>   0.45   3.45   0.70   1.00  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         tcl <- 1 #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # Change the variable to a covariate parameter (ie it doesn't have an # initial estimate so remove it with the `-` operator):  f6 <- f4 |>   ini(-tcl) #> ℹ changing between subject variability parameter 'tcl' to covariate parameter print(f6) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # You can change the covariate or remove the parameter estimate by # `tcl <- NULL`:  f6 <- f4 |>   ini(tcl <- NULL) #> ℹ changing between subject variability parameter 'tcl' to covariate parameter print(f6) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # to add it back as a between subject variability or population # parameter you can pipe it as follows:  f7 <- f6 |>   ini(tcl=4) #> ℹ promote `tcl` to population parameter with initial estimate 4 #> ℹ change initial estimate of `tcl` to `4` print(f7) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd    tcl  #>   0.45   3.45   0.70   4.00  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         tcl <- 4 #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f8 <- f6 |>   ini(tcl ~ 0.1) #> ℹ promote `tcl` to between subject variability with initial estimate 0.1 #> ℹ change initial estimate of `tcl` to `0.1` print(f8) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v tcl #> eta.ka    0.6   0.0 0.0 #> eta.v     0.0   0.1 0.0 #> tcl       0.0   0.0 0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>         tcl ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"changing-parameter-labels","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Changing parameter labels","title":"Modifying Models","text":"want change/add parameter label assign parameter label(\"label add\"). example: can also change order performing operations: want remove labels can remove ini(par=label(NULL)); example:","code":"f4 <- f |>   ini(tka=label(\"Typical Ka (1/hr)\"))  print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Typical Ka (1/hr)\") #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f5 <- f |>   ini(tka=label(\"Typical Ka (1/hr)\"), append=tcl)  print(f5) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tcl    tka     tv add.sd  #>   1.00   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tcl <- 1 #>         tka <- 0.45 #>         label(\"Typical Ka (1/hr)\") #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f6 <- f |>   ini(tka=label(NULL))  print(f6) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"changing-parameter-transformations","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Changing parameter transformations","title":"Modifying Models","text":"Back-transformations -ride back transformations nlmixr2 models. similar modification labels. use |> ini(tka=backTransform(exp)) add exponential back-transformation data: wish remove can also |> ini(tka=backTransform(NULL)):","code":"f7 <- f |>   ini(tka=backTransform(exp))  print(f7) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         backTransform(\"exp\") #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f8 <- f |>   ini(tka=backTransform(NULL))  print(f8) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"more-granular-access-of-initial-conditions","dir":"Articles","previous_headings":"Modifying initial estimates","what":"More granular access of initial conditions","title":"Modifying Models","text":"Just like model() can modify underlying data frame represents ini() block. case simply change initial estimate first parameter (tka):","code":"f <- rxode2(f)  ini <- f$iniDf  print(ini) #>   ntheta neta1 neta2   name lower  est upper   fix label backTransform #> 1      1    NA    NA    tka  -Inf 0.45   Inf FALSE  <NA>          <NA> #> 2      2    NA    NA    tcl  -Inf 1.00   Inf FALSE  <NA>          <NA> #> 3      3    NA    NA     tv  -Inf 3.45   Inf FALSE  <NA>          <NA> #> 4      4    NA    NA add.sd     0 0.70   Inf FALSE  <NA>          <NA> #> 5     NA     1     1 eta.ka  -Inf 0.60   Inf FALSE  <NA>          <NA> #> 6     NA     2     2  eta.v  -Inf 0.10   Inf FALSE  <NA>          <NA> #>   condition  err #> 1      <NA> <NA> #> 2      <NA> <NA> #> 3      <NA> <NA> #> 4        cp  add #> 5        id <NA> #> 6        id <NA> ini$est[1] <- 7  ini(f) <- ini  print(f) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   7.00   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 7 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-cmt.html","id":"how-rxode2-assigns-compartment-numbers","dir":"Articles","previous_headings":"","what":"How rxode2 assigns compartment numbers","title":"Changing rxode2 compartment numbers","text":"rxode2 automatically assigns compartment numbers parsing. example, Mavoglurant PBPK model following model may used: look printout, can see rxode2 assigned compartment number(s) can also see classic rxode2 model. case use summary() function: case, Venous_Blood assigned compartment 15. Figuring can inconvenient also lead re-numbering compartment simulation estimation datasets. easy probably clearer specify compartment name, tools support compartment numbers. Therefore, way number compartment easily can lead less data modification multiple tools.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` pbpk <- function() {   model({     KbBR = exp(lKbBR)     KbMU = exp(lKbMU)     KbAD = exp(lKbAD)     CLint= exp(lCLint + eta.LClint)     KbBO = exp(lKbBO)     KbRB = exp(lKbRB)      ## Regional blood flows     # Cardiac output (L/h) from White et al (1968)     CO  = (187.00*WT^0.81)*60/1000      QHT = 4.0 *CO/100     QBR = 12.0*CO/100     QMU = 17.0*CO/100     QAD = 5.0 *CO/100     QSK = 5.0 *CO/100     QSP = 3.0 *CO/100     QPA = 1.0 *CO/100     QLI = 25.5*CO/100     QST = 1.0 *CO/100     QGU = 14.0*CO/100     # Hepatic artery blood flow     QHA = QLI - (QSP + QPA + QST + QGU)      QBO = 5.0 *CO/100     QKI = 19.0*CO/100     QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI)     QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI + QRB      ## Organs' volumes = organs' weights / organs' density     VLU = (0.76 *WT/100)/1.051     VHT = (0.47 *WT/100)/1.030     VBR = (2.00 *WT/100)/1.036     VMU = (40.00*WT/100)/1.041     VAD = (21.42*WT/100)/0.916     VSK = (3.71 *WT/100)/1.116     VSP = (0.26 *WT/100)/1.054     VPA = (0.14 *WT/100)/1.045     VLI = (2.57 *WT/100)/1.040     VST = (0.21 *WT/100)/1.050     VGU = (1.44 *WT/100)/1.043     VBO = (14.29*WT/100)/1.990     VKI = (0.44 *WT/100)/1.050     VAB = (2.81 *WT/100)/1.040     VVB = (5.62 *WT/100)/1.040     VRB = (3.86 *WT/100)/1.040      ## Fixed parameters     BP = 0.61      # Blood:plasma partition coefficient     fup = 0.028    # Fraction unbound in plasma     fub = fup/BP   # Fraction unbound in blood      KbLU = exp(0.8334)     KbHT = exp(1.1205)     KbSK = exp(-.5238)     KbSP = exp(0.3224)     KbPA = exp(0.3224)     KbLI = exp(1.7604)     KbST = exp(0.3224)     KbGU = exp(1.2026)     KbKI = exp(1.3171)      ##-----------------------------------------     S15 = VVB*BP/1000     C15 = Venous_Blood/S15      ##-----------------------------------------     d/dt(Lungs) = QLU*(Venous_Blood/VVB - Lungs/KbLU/VLU)     d/dt(Heart) = QHT*(Arterial_Blood/VAB - Heart/KbHT/VHT)     d/dt(Brain) = QBR*(Arterial_Blood/VAB - Brain/KbBR/VBR)     d/dt(Muscles) = QMU*(Arterial_Blood/VAB - Muscles/KbMU/VMU)     d/dt(Adipose) = QAD*(Arterial_Blood/VAB - Adipose/KbAD/VAD)     d/dt(Skin) = QSK*(Arterial_Blood/VAB - Skin/KbSK/VSK)     d/dt(Spleen) = QSP*(Arterial_Blood/VAB - Spleen/KbSP/VSP)     d/dt(Pancreas) = QPA*(Arterial_Blood/VAB - Pancreas/KbPA/VPA)     d/dt(Liver) = QHA*Arterial_Blood/VAB + QSP*Spleen/KbSP/VSP +       QPA*Pancreas/KbPA/VPA + QST*Stomach/KbST/VST +       QGU*Gut/KbGU/VGU - CLint*fub*Liver/KbLI/VLI - QLI*Liver/KbLI/VLI     d/dt(Stomach) = QST*(Arterial_Blood/VAB - Stomach/KbST/VST)     d/dt(Gut) = QGU*(Arterial_Blood/VAB - Gut/KbGU/VGU)     d/dt(Bones) = QBO*(Arterial_Blood/VAB - Bones/KbBO/VBO)     d/dt(Kidneys) = QKI*(Arterial_Blood/VAB - Kidneys/KbKI/VKI)     d/dt(Arterial_Blood) = QLU*(Lungs/KbLU/VLU - Arterial_Blood/VAB)     d/dt(Venous_Blood) = QHT*Heart/KbHT/VHT + QBR*Brain/KbBR/VBR +       QMU*Muscles/KbMU/VMU + QAD*Adipose/KbAD/VAD + QSK*Skin/KbSK/VSK +       QLI*Liver/KbLI/VLI + QBO*Bones/KbBO/VBO + QKI*Kidneys/KbKI/VKI +       QRB*Rest_of_Body/KbRB/VRB - QLU*Venous_Blood/VVB     d/dt(Rest_of_Body) = QRB*(Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB)   }) } pbpk <- pbpk() print(pbpk) #>  -- rxode2-based free-form 16-cmt ODE model -------------------------------------  #>  #> States ($state or $stateDf):  #>    Compartment Number Compartment Name #> 1                   1            Lungs #> 2                   2            Heart #> 3                   3            Brain #> 4                   4          Muscles #> 5                   5          Adipose #> 6                   6             Skin #> 7                   7           Spleen #> 8                   8         Pancreas #> 9                   9            Liver #> 10                 10          Stomach #> 11                 11              Gut #> 12                 12            Bones #> 13                 13          Kidneys #> 14                 14   Arterial_Blood #> 15                 15     Venous_Blood #> 16                 16     Rest_of_Body #>  -- Model (Normalized Syntax): --  #> function() { #>     model({ #>         KbBR = exp(lKbBR) #>         KbMU = exp(lKbMU) #>         KbAD = exp(lKbAD) #>         CLint = exp(lCLint + eta.LClint) #>         KbBO = exp(lKbBO) #>         KbRB = exp(lKbRB) #>         CO = (187 * WT^0.81) * 60/1000 #>         QHT = 4 * CO/100 #>         QBR = 12 * CO/100 #>         QMU = 17 * CO/100 #>         QAD = 5 * CO/100 #>         QSK = 5 * CO/100 #>         QSP = 3 * CO/100 #>         QPA = 1 * CO/100 #>         QLI = 25.5 * CO/100 #>         QST = 1 * CO/100 #>         QGU = 14 * CO/100 #>         QHA = QLI - (QSP + QPA + QST + QGU) #>         QBO = 5 * CO/100 #>         QKI = 19 * CO/100 #>         QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO +  #>             QKI) #>         QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI +  #>             QRB #>         VLU = (0.76 * WT/100)/1.051 #>         VHT = (0.47 * WT/100)/1.03 #>         VBR = (2 * WT/100)/1.036 #>         VMU = (40 * WT/100)/1.041 #>         VAD = (21.42 * WT/100)/0.916 #>         VSK = (3.71 * WT/100)/1.116 #>         VSP = (0.26 * WT/100)/1.054 #>         VPA = (0.14 * WT/100)/1.045 #>         VLI = (2.57 * WT/100)/1.04 #>         VST = (0.21 * WT/100)/1.05 #>         VGU = (1.44 * WT/100)/1.043 #>         VBO = (14.29 * WT/100)/1.99 #>         VKI = (0.44 * WT/100)/1.05 #>         VAB = (2.81 * WT/100)/1.04 #>         VVB = (5.62 * WT/100)/1.04 #>         VRB = (3.86 * WT/100)/1.04 #>         BP = 0.61 #>         fup = 0.028 #>         fub = fup/BP #>         KbLU = exp(0.8334) #>         KbHT = exp(1.1205) #>         KbSK = exp(-0.5238) #>         KbSP = exp(0.3224) #>         KbPA = exp(0.3224) #>         KbLI = exp(1.7604) #>         KbST = exp(0.3224) #>         KbGU = exp(1.2026) #>         KbKI = exp(1.3171) #>         S15 = VVB * BP/1000 #>         C15 = Venous_Blood/S15 #>         d/dt(Lungs) = QLU * (Venous_Blood/VVB - Lungs/KbLU/VLU) #>         d/dt(Heart) = QHT * (Arterial_Blood/VAB - Heart/KbHT/VHT) #>         d/dt(Brain) = QBR * (Arterial_Blood/VAB - Brain/KbBR/VBR) #>         d/dt(Muscles) = QMU * (Arterial_Blood/VAB - Muscles/KbMU/VMU) #>         d/dt(Adipose) = QAD * (Arterial_Blood/VAB - Adipose/KbAD/VAD) #>         d/dt(Skin) = QSK * (Arterial_Blood/VAB - Skin/KbSK/VSK) #>         d/dt(Spleen) = QSP * (Arterial_Blood/VAB - Spleen/KbSP/VSP) #>         d/dt(Pancreas) = QPA * (Arterial_Blood/VAB - Pancreas/KbPA/VPA) #>         d/dt(Liver) = QHA * Arterial_Blood/VAB + QSP * Spleen/KbSP/VSP +  #>             QPA * Pancreas/KbPA/VPA + QST * Stomach/KbST/VST +  #>             QGU * Gut/KbGU/VGU - CLint * fub * Liver/KbLI/VLI -  #>             QLI * Liver/KbLI/VLI #>         d/dt(Stomach) = QST * (Arterial_Blood/VAB - Stomach/KbST/VST) #>         d/dt(Gut) = QGU * (Arterial_Blood/VAB - Gut/KbGU/VGU) #>         d/dt(Bones) = QBO * (Arterial_Blood/VAB - Bones/KbBO/VBO) #>         d/dt(Kidneys) = QKI * (Arterial_Blood/VAB - Kidneys/KbKI/VKI) #>         d/dt(Arterial_Blood) = QLU * (Lungs/KbLU/VLU - Arterial_Blood/VAB) #>         d/dt(Venous_Blood) = QHT * Heart/KbHT/VHT + QBR * Brain/KbBR/VBR +  #>             QMU * Muscles/KbMU/VMU + QAD * Adipose/KbAD/VAD +  #>             QSK * Skin/KbSK/VSK + QLI * Liver/KbLI/VLI + QBO *  #>             Bones/KbBO/VBO + QKI * Kidneys/KbKI/VKI + QRB * Rest_of_Body/KbRB/VRB -  #>             QLU * Venous_Blood/VVB #>         d/dt(Rest_of_Body) = QRB * (Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB) #>     }) #> } pbpk <- pbpk$simulationModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ summary(pbpk) #> rxode2 2.1.3 model named rx_a688fd3d12f8698d15feb47597e1364f model (ready).  #> DLL: /tmp/RtmpSPRppR/rxode2/rx_a688fd3d12f8698d15feb47597e1364f__.rxd/rx_a688fd3d12f8698d15feb47597e1364f_.so #> NULL #>  #> Calculated Variables: #>  [1] \"KbBR\"  \"KbMU\"  \"KbAD\"  \"CLint\" \"KbBO\"  \"KbRB\"  \"CO\"    \"QHT\"   \"QBR\"   #> [10] \"QMU\"   \"QAD\"   \"QSK\"   \"QSP\"   \"QPA\"   \"QLI\"   \"QST\"   \"QGU\"   \"QHA\"   #> [19] \"QBO\"   \"QKI\"   \"QRB\"   \"QLU\"   \"VLU\"   \"VHT\"   \"VBR\"   \"VMU\"   \"VAD\"   #> [28] \"VSK\"   \"VSP\"   \"VPA\"   \"VLI\"   \"VST\"   \"VGU\"   \"VBO\"   \"VKI\"   \"VAB\"   #> [37] \"VVB\"   \"VRB\"   \"fub\"   \"KbLU\"  \"KbHT\"  \"KbSK\"  \"KbSP\"  \"KbPA\"  \"KbLI\"  #> [46] \"KbST\"  \"KbGU\"  \"KbKI\"  \"S15\"   \"C15\"   #> -- rxode2 Model Syntax -- #> rxode2({ #>     param(lKbBR, lKbMU, lKbAD, lCLint, eta.LClint, lKbBO, lKbRB,  #>         WT) #>     KbBR = exp(lKbBR) #>     KbMU = exp(lKbMU) #>     KbAD = exp(lKbAD) #>     CLint = exp(lCLint + eta.LClint) #>     KbBO = exp(lKbBO) #>     KbRB = exp(lKbRB) #>     CO = (187 * WT^0.81) * 60/1000 #>     QHT = 4 * CO/100 #>     QBR = 12 * CO/100 #>     QMU = 17 * CO/100 #>     QAD = 5 * CO/100 #>     QSK = 5 * CO/100 #>     QSP = 3 * CO/100 #>     QPA = 1 * CO/100 #>     QLI = 25.5 * CO/100 #>     QST = 1 * CO/100 #>     QGU = 14 * CO/100 #>     QHA = QLI - (QSP + QPA + QST + QGU) #>     QBO = 5 * CO/100 #>     QKI = 19 * CO/100 #>     QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI) #>     QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI + QRB #>     VLU = (0.76 * WT/100)/1.051 #>     VHT = (0.47 * WT/100)/1.03 #>     VBR = (2 * WT/100)/1.036 #>     VMU = (40 * WT/100)/1.041 #>     VAD = (21.42 * WT/100)/0.916 #>     VSK = (3.71 * WT/100)/1.116 #>     VSP = (0.26 * WT/100)/1.054 #>     VPA = (0.14 * WT/100)/1.045 #>     VLI = (2.57 * WT/100)/1.04 #>     VST = (0.21 * WT/100)/1.05 #>     VGU = (1.44 * WT/100)/1.043 #>     VBO = (14.29 * WT/100)/1.99 #>     VKI = (0.44 * WT/100)/1.05 #>     VAB = (2.81 * WT/100)/1.04 #>     VVB = (5.62 * WT/100)/1.04 #>     VRB = (3.86 * WT/100)/1.04 #>     BP = 0.61 #>     fup = 0.028 #>     fub = fup/BP #>     KbLU = exp(0.8334) #>     KbHT = exp(1.1205) #>     KbSK = exp(-0.5238) #>     KbSP = exp(0.3224) #>     KbPA = exp(0.3224) #>     KbLI = exp(1.7604) #>     KbST = exp(0.3224) #>     KbGU = exp(1.2026) #>     KbKI = exp(1.3171) #>     S15 = VVB * BP/1000 #>     C15 = Venous_Blood/S15 #>     d/dt(Lungs) = QLU * (Venous_Blood/VVB - Lungs/KbLU/VLU) #>     d/dt(Heart) = QHT * (Arterial_Blood/VAB - Heart/KbHT/VHT) #>     d/dt(Brain) = QBR * (Arterial_Blood/VAB - Brain/KbBR/VBR) #>     d/dt(Muscles) = QMU * (Arterial_Blood/VAB - Muscles/KbMU/VMU) #>     d/dt(Adipose) = QAD * (Arterial_Blood/VAB - Adipose/KbAD/VAD) #>     d/dt(Skin) = QSK * (Arterial_Blood/VAB - Skin/KbSK/VSK) #>     d/dt(Spleen) = QSP * (Arterial_Blood/VAB - Spleen/KbSP/VSP) #>     d/dt(Pancreas) = QPA * (Arterial_Blood/VAB - Pancreas/KbPA/VPA) #>     d/dt(Liver) = QHA * Arterial_Blood/VAB + QSP * Spleen/KbSP/VSP +  #>         QPA * Pancreas/KbPA/VPA + QST * Stomach/KbST/VST + QGU *  #>         Gut/KbGU/VGU - CLint * fub * Liver/KbLI/VLI - QLI * Liver/KbLI/VLI #>     d/dt(Stomach) = QST * (Arterial_Blood/VAB - Stomach/KbST/VST) #>     d/dt(Gut) = QGU * (Arterial_Blood/VAB - Gut/KbGU/VGU) #>     d/dt(Bones) = QBO * (Arterial_Blood/VAB - Bones/KbBO/VBO) #>     d/dt(Kidneys) = QKI * (Arterial_Blood/VAB - Kidneys/KbKI/VKI) #>     d/dt(Arterial_Blood) = QLU * (Lungs/KbLU/VLU - Arterial_Blood/VAB) #>     d/dt(Venous_Blood) = QHT * Heart/KbHT/VHT + QBR * Brain/KbBR/VBR +  #>         QMU * Muscles/KbMU/VMU + QAD * Adipose/KbAD/VAD + QSK *  #>         Skin/KbSK/VSK + QLI * Liver/KbLI/VLI + QBO * Bones/KbBO/VBO +  #>         QKI * Kidneys/KbKI/VKI + QRB * Rest_of_Body/KbRB/VRB -  #>         QLU * Venous_Blood/VVB #>     d/dt(Rest_of_Body) = QRB * (Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB) #> })"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-cmt.html","id":"changing-compartment-numbers-by-pre-declaring-the-compartments","dir":"Articles","previous_headings":"","what":"Changing compartment numbers by pre-declaring the compartments","title":"Changing rxode2 compartment numbers","text":"add compartments rxode2 model order desire simply need pre-declare compartments cmt. example specifying Venous_Blood Skin 1st 2nd compartments, respectively, simple: can see change simple printout first two compartments Venous_Blood followed Skin.","code":"pbpk2 <- function() {   model({     ## Now this is the first compartment, ie cmt=1     cmt(Venous_Blood)     ## Skin may be a compartment you wish to dose to as well,     ##  so it is now cmt=2     cmt(Skin)     KbBR = exp(lKbBR)     KbMU = exp(lKbMU)     KbAD = exp(lKbAD)     CLint= exp(lCLint + eta.LClint)     KbBO = exp(lKbBO)     KbRB = exp(lKbRB)      ## Regional blood flows     # Cardiac output (L/h) from White et al (1968)m     CO  = (187.00*WT^0.81)*60/1000;     QHT = 4.0 *CO/100;     QBR = 12.0*CO/100;     QMU = 17.0*CO/100;     QAD = 5.0 *CO/100;     QSK = 5.0 *CO/100;     QSP = 3.0 *CO/100;     QPA = 1.0 *CO/100;     QLI = 25.5*CO/100;     QST = 1.0 *CO/100;     QGU = 14.0*CO/100;     QHA = QLI - (QSP + QPA + QST + QGU); # Hepatic artery blood flow     QBO = 5.0 *CO/100;     QKI = 19.0*CO/100;     QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI);     QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI + QRB;      ## Organs' volumes = organs' weights / organs' density     VLU = (0.76 *WT/100)/1.051;     VHT = (0.47 *WT/100)/1.030;     VBR = (2.00 *WT/100)/1.036;     VMU = (40.00*WT/100)/1.041;     VAD = (21.42*WT/100)/0.916;     VSK = (3.71 *WT/100)/1.116;     VSP = (0.26 *WT/100)/1.054;     VPA = (0.14 *WT/100)/1.045;     VLI = (2.57 *WT/100)/1.040;     VST = (0.21 *WT/100)/1.050;     VGU = (1.44 *WT/100)/1.043;     VBO = (14.29*WT/100)/1.990;     VKI = (0.44 *WT/100)/1.050;     VAB = (2.81 *WT/100)/1.040;     VVB = (5.62 *WT/100)/1.040;     VRB = (3.86 *WT/100)/1.040;      ## Fixed parameters     BP = 0.61;      # Blood:plasma partition coefficient     fup = 0.028;    # Fraction unbound in plasma     fub = fup/BP;   # Fraction unbound in blood      KbLU = exp(0.8334);     KbHT = exp(1.1205);     KbSK = exp(-.5238);     KbSP = exp(0.3224);     KbPA = exp(0.3224);     KbLI = exp(1.7604);     KbST = exp(0.3224);     KbGU = exp(1.2026);     KbKI = exp(1.3171);      ##-----------------------------------------     S15 = VVB*BP/1000;     C15 = Venous_Blood/S15      ##-----------------------------------------     d/dt(Lungs) = QLU*(Venous_Blood/VVB - Lungs/KbLU/VLU);     d/dt(Heart) = QHT*(Arterial_Blood/VAB - Heart/KbHT/VHT);     d/dt(Brain) = QBR*(Arterial_Blood/VAB - Brain/KbBR/VBR);     d/dt(Muscles) = QMU*(Arterial_Blood/VAB - Muscles/KbMU/VMU);     d/dt(Adipose) = QAD*(Arterial_Blood/VAB - Adipose/KbAD/VAD);     d/dt(Skin) = QSK*(Arterial_Blood/VAB - Skin/KbSK/VSK);     d/dt(Spleen) = QSP*(Arterial_Blood/VAB - Spleen/KbSP/VSP);     d/dt(Pancreas) = QPA*(Arterial_Blood/VAB - Pancreas/KbPA/VPA);     d/dt(Liver) = QHA*Arterial_Blood/VAB + QSP*Spleen/KbSP/VSP +       QPA*Pancreas/KbPA/VPA + QST*Stomach/KbST/VST + QGU*Gut/KbGU/VGU -       CLint*fub*Liver/KbLI/VLI - QLI*Liver/KbLI/VLI;       d/dt(Stomach) = QST*(Arterial_Blood/VAB - Stomach/KbST/VST);       d/dt(Gut) = QGU*(Arterial_Blood/VAB - Gut/KbGU/VGU);       d/dt(Bones) = QBO*(Arterial_Blood/VAB - Bones/KbBO/VBO);       d/dt(Kidneys) = QKI*(Arterial_Blood/VAB - Kidneys/KbKI/VKI);       d/dt(Arterial_Blood) = QLU*(Lungs/KbLU/VLU - Arterial_Blood/VAB);       d/dt(Venous_Blood) = QHT*Heart/KbHT/VHT + QBR*Brain/KbBR/VBR +         QMU*Muscles/KbMU/VMU + QAD*Adipose/KbAD/VAD + QSK*Skin/KbSK/VSK +         QLI*Liver/KbLI/VLI + QBO*Bones/KbBO/VBO + QKI*Kidneys/KbKI/VKI +         QRB*Rest_of_Body/KbRB/VRB - QLU*Venous_Blood/VVB;         d/dt(Rest_of_Body) = QRB*(Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB);   }) } pbpk2 <- pbpk2() pbpk2 #>  -- rxode2-based free-form 16-cmt ODE model -------------------------------------  #>  #> States ($state or $stateDf):  #>    Compartment Number Compartment Name #> 1                   1     Venous_Blood #> 2                   2             Skin #> 3                   3            Lungs #> 4                   4            Heart #> 5                   5            Brain #> 6                   6          Muscles #> 7                   7          Adipose #> 8                   8           Spleen #> 9                   9         Pancreas #> 10                 10            Liver #> 11                 11          Stomach #> 12                 12              Gut #> 13                 13            Bones #> 14                 14          Kidneys #> 15                 15   Arterial_Blood #> 16                 16     Rest_of_Body #>  -- Model (Normalized Syntax): --  #> function() { #>     model({ #>         cmt(Venous_Blood) #>         cmt(Skin) #>         KbBR = exp(lKbBR) #>         KbMU = exp(lKbMU) #>         KbAD = exp(lKbAD) #>         CLint = exp(lCLint + eta.LClint) #>         KbBO = exp(lKbBO) #>         KbRB = exp(lKbRB) #>         CO = (187 * WT^0.81) * 60/1000 #>         QHT = 4 * CO/100 #>         QBR = 12 * CO/100 #>         QMU = 17 * CO/100 #>         QAD = 5 * CO/100 #>         QSK = 5 * CO/100 #>         QSP = 3 * CO/100 #>         QPA = 1 * CO/100 #>         QLI = 25.5 * CO/100 #>         QST = 1 * CO/100 #>         QGU = 14 * CO/100 #>         QHA = QLI - (QSP + QPA + QST + QGU) #>         QBO = 5 * CO/100 #>         QKI = 19 * CO/100 #>         QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO +  #>             QKI) #>         QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI +  #>             QRB #>         VLU = (0.76 * WT/100)/1.051 #>         VHT = (0.47 * WT/100)/1.03 #>         VBR = (2 * WT/100)/1.036 #>         VMU = (40 * WT/100)/1.041 #>         VAD = (21.42 * WT/100)/0.916 #>         VSK = (3.71 * WT/100)/1.116 #>         VSP = (0.26 * WT/100)/1.054 #>         VPA = (0.14 * WT/100)/1.045 #>         VLI = (2.57 * WT/100)/1.04 #>         VST = (0.21 * WT/100)/1.05 #>         VGU = (1.44 * WT/100)/1.043 #>         VBO = (14.29 * WT/100)/1.99 #>         VKI = (0.44 * WT/100)/1.05 #>         VAB = (2.81 * WT/100)/1.04 #>         VVB = (5.62 * WT/100)/1.04 #>         VRB = (3.86 * WT/100)/1.04 #>         BP = 0.61 #>         fup = 0.028 #>         fub = fup/BP #>         KbLU = exp(0.8334) #>         KbHT = exp(1.1205) #>         KbSK = exp(-0.5238) #>         KbSP = exp(0.3224) #>         KbPA = exp(0.3224) #>         KbLI = exp(1.7604) #>         KbST = exp(0.3224) #>         KbGU = exp(1.2026) #>         KbKI = exp(1.3171) #>         S15 = VVB * BP/1000 #>         C15 = Venous_Blood/S15 #>         d/dt(Lungs) = QLU * (Venous_Blood/VVB - Lungs/KbLU/VLU) #>         d/dt(Heart) = QHT * (Arterial_Blood/VAB - Heart/KbHT/VHT) #>         d/dt(Brain) = QBR * (Arterial_Blood/VAB - Brain/KbBR/VBR) #>         d/dt(Muscles) = QMU * (Arterial_Blood/VAB - Muscles/KbMU/VMU) #>         d/dt(Adipose) = QAD * (Arterial_Blood/VAB - Adipose/KbAD/VAD) #>         d/dt(Skin) = QSK * (Arterial_Blood/VAB - Skin/KbSK/VSK) #>         d/dt(Spleen) = QSP * (Arterial_Blood/VAB - Spleen/KbSP/VSP) #>         d/dt(Pancreas) = QPA * (Arterial_Blood/VAB - Pancreas/KbPA/VPA) #>         d/dt(Liver) = QHA * Arterial_Blood/VAB + QSP * Spleen/KbSP/VSP +  #>             QPA * Pancreas/KbPA/VPA + QST * Stomach/KbST/VST +  #>             QGU * Gut/KbGU/VGU - CLint * fub * Liver/KbLI/VLI -  #>             QLI * Liver/KbLI/VLI #>         d/dt(Stomach) = QST * (Arterial_Blood/VAB - Stomach/KbST/VST) #>         d/dt(Gut) = QGU * (Arterial_Blood/VAB - Gut/KbGU/VGU) #>         d/dt(Bones) = QBO * (Arterial_Blood/VAB - Bones/KbBO/VBO) #>         d/dt(Kidneys) = QKI * (Arterial_Blood/VAB - Kidneys/KbKI/VKI) #>         d/dt(Arterial_Blood) = QLU * (Lungs/KbLU/VLU - Arterial_Blood/VAB) #>         d/dt(Venous_Blood) = QHT * Heart/KbHT/VHT + QBR * Brain/KbBR/VBR +  #>             QMU * Muscles/KbMU/VMU + QAD * Adipose/KbAD/VAD +  #>             QSK * Skin/KbSK/VSK + QLI * Liver/KbLI/VLI + QBO *  #>             Bones/KbBO/VBO + QKI * Kidneys/KbKI/VKI + QRB * Rest_of_Body/KbRB/VRB -  #>             QLU * Venous_Blood/VVB #>         d/dt(Rest_of_Body) = QRB * (Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-cmt.html","id":"appending-compartments-to-the-model","dir":"Articles","previous_headings":"","what":"Appending compartments to the model","title":"Changing rxode2 compartment numbers","text":"can also append “compartments” model. ODE solving internals, add fake compartments model differential equations defined. example legal: can see clearly underlying classic rxode2 model: compartments defined differential equations supported; model : give error:","code":"ode.1c.ka <- function(){   model({     C2 = center/V     d / dt(depot) = -KA * depot     d/dt(center) = KA * depot - CL*C2     cmt(eff)   }) }  ode.1c.ka <- ode.1c.ka() print(ode.1c.ka) #>  -- rxode2-based free-form 2-cmt ODE model --------------------------------------  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  -- Model (Normalized Syntax): --  #> function() { #>     model({ #>         C2 = center/V #>         d/dt(depot) = -KA * depot #>         d/dt(center) = KA * depot - CL * C2 #>         cmt(eff) #>     }) #> } ode.1c.ka$simulationModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_31957e383dd61d7e82e22d46f409ad8a model (ready).  #> x$state: depot, center #> x$stateExtra: eff #> x$params: V, KA, CL #> x$lhs: C2 ode.1c.ka <- rxode2({     cmt(eff)     C2 = center/V;     d / dt(depot) = -KA * depot     d/dt(center) = KA * depot - CL*C2 }) Error in rxModelVars_(obj) :    Evaluation error: Compartment 'eff' needs differential equations defined."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-covariates.html","id":"individual-covariates","dir":"Articles","previous_headings":"","what":"Individual Covariates","title":"rxode2 Covariates","text":"individual covariate wish solve may specify iCov dataset:","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` library(units) ## udunits database from /usr/share/xml/udunits/udunits2.xml library(xgxr)  mod3 <- function() {   ini({     TKA   <- 2.94E-01     ## Clearance with individuals     TCL   <- 1.86E+01      TV2   <-4.02E+01     TQ    <-1.05E+01     TV3   <-2.97E+02     TKin  <- 1     TKout <- 1     TEC50 <-200   })   model({     KA            <- TKA     CL            <- TCL * (WT / 70) ^ 0.75     V2            <- TV2     Q             <- TQ     V3            <- TV3     Kin           <- TKin     Kout          <- TKout     EC50          <- TEC50     Tz            <- 8     amp           <- 0.1     C2            <- central/V2     C3            <- peri/V3     d/dt(depot)   <- -KA*depot     d/dt(central) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)    <-                    Q*C2 - Q*C3     d/dt(eff)     <-  Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0)        <- 1  ## This specifies that the effect compartment starts at 1.   }) }  ev <- et(amount.units=\"mg\", time.units=\"hours\") %>%   et(amt=10000, cmt=1) %>%   et(0,48,length.out=100) %>%   et(id=1:4)  set.seed(10) rxSetSeed(10) ## Now use iCov to simulate a 4-id sample r1 <- solve(mod3, ev,             # Create individual covariate data-frame             iCov=data.frame(id=1:4, WT=rnorm(4, 70, 10))) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(r1) ## ── Solved rxode2 object ── ## ── Parameters ($params): ── ##     TKA     TCL     TV2      TQ     TV3    TKin   TKout   TEC50      Tz     amp  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000   8.000   0.100  ## ── Initial Conditions ($inits): ── ##   depot central    peri     eff  ##       0       0       0       1  ## ── First part of data (object): ── ## # A tibble: 400 × 17 ##      id  time    KA    CL    V2     Q    V3   Kin  Kout  EC50    C2    C3  depot ##   <int>   [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> ## 1     1 0     0.294  18.6  40.2  10.5   297     1     1   200   0   0     10000  ## 2     1 0.485 0.294  18.6  40.2  10.5   297     1     1   200  27.8 0.257  8671. ## 3     1 0.970 0.294  18.6  40.2  10.5   297     1     1   200  43.7 0.873  7519. ## 4     1 1.45  0.294  18.6  40.2  10.5   297     1     1   200  51.7 1.68   6520. ## 5     1 1.94  0.294  18.6  40.2  10.5   297     1     1   200  54.7 2.56   5654. ## 6     1 2.42  0.294  18.6  40.2  10.5   297     1     1   200  54.5 3.45   4903. ## # ℹ 394 more rows ## # ℹ 4 more variables: central <dbl>, peri <dbl>, eff <dbl>, WT <dbl> plot(r1, C2, log=\"y\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-covariates.html","id":"time-varying-covariates","dir":"Articles","previous_headings":"","what":"Time Varying Covariates","title":"rxode2 Covariates","text":"Covariates easy specify rxode2, can specify variable. Time-varying covariates, like clock time circadian rhythm model, can also used. Extending indirect response model already discussed, : Now covariate present event dataset, system can solved combining dataset model: solving ODE equations, solver may sample times outside data. happens, ODE solver can use linear interpolation covariate values. equivalent R’s approxfun method=\"linear\".   Note linear approximation case leads kinks solved system 24-hours covariate linear interpolation near 24 near 0. linear seems reasonable, cases like clock time make interpolation methods attractive. rxode2 default covariate interpolation last observation carried forward (locf), constant approximation. equivalent R’s approxfun method=\"constant\". gives following plots:   case, plots seem smoother. can also use NONMEM’s preferred interpolation style next observation carried backward (NOCB): gives following plots:","code":"library(rxode2) library(units)  mod4 <- mod3 %>%   model(d/dt(eff) <-  Kin - Kout*(1-C2/(EC50+C2))*eff) %>%   model(-Kin) %>%   model(Kin <- TKin + amp *cos(2*pi*(ctime-Tz)/24), append=C2, cov=\"ctime\")   ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=10000, cmt=1) %>%     et(0,48,length.out=100)   ## Create data frame of 8 am dosing for the first dose This is done ## with base R but it can be done with dplyr or data.table ev$ctime <- (ev$time+set_units(8,hr)) %% 24 ev$WT <- 70 r1 <- solve(mod4, ev, covsInterpolation=\"linear\") print(r1) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>        TKA        TCL        TV2         TQ        TV3      TKout      TEC50  #>   0.294000  18.600000  40.200000  10.500000 297.000000   1.000000 200.000000  #>       TKin         Tz        amp         pi  #>   1.000000   8.000000   0.100000   3.141593  #> ── Initial Conditions ($inits): ── #>   depot central    peri     eff  #>       0       0       0       1  #> ── First part of data (object): ── #> # A tibble: 100 × 17 #>    time    KA    CL    V2     Q    V3  Kout  EC50    C2   Kin    C3  depot #>     [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> #> 1 0     0.294  18.6  40.2  10.5   297     1   200   0    1.1  0     10000  #> 2 0.485 0.294  18.6  40.2  10.5   297     1   200  27.8  1.10 0.257  8671. #> 3 0.970 0.294  18.6  40.2  10.5   297     1   200  43.7  1.10 0.874  7519. #> 4 1.45  0.294  18.6  40.2  10.5   297     1   200  51.8  1.09 1.68   6520. #> 5 1.94  0.294  18.6  40.2  10.5   297     1   200  54.8  1.09 2.56   5654. #> 6 2.42  0.294  18.6  40.2  10.5   297     1   200  54.6  1.08 3.45   4903. #> # ℹ 94 more rows #> # ℹ 5 more variables: central <dbl>, peri <dbl>, eff <dbl>, ctime [h], WT <dbl> plot(r1,C2, ylab=\"Central Concentration\") plot(r1,eff) + ylab(\"Effect\") + xlab(\"Time\") r1 <- solve(mod4, ev,covsInterpolation=\"locf\") print(r1) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>        TKA        TCL        TV2         TQ        TV3      TKout      TEC50  #>   0.294000  18.600000  40.200000  10.500000 297.000000   1.000000 200.000000  #>       TKin         Tz        amp         pi  #>   1.000000   8.000000   0.100000   3.141593  #> ── Initial Conditions ($inits): ── #>   depot central    peri     eff  #>       0       0       0       1  #> ── First part of data (object): ── #> # A tibble: 100 × 17 #>    time    KA    CL    V2     Q    V3  Kout  EC50    C2   Kin    C3  depot #>     [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> #> 1 0     0.294  18.6  40.2  10.5   297     1   200   0    1.1  0     10000  #> 2 0.485 0.294  18.6  40.2  10.5   297     1   200  27.8  1.10 0.257  8671. #> 3 0.970 0.294  18.6  40.2  10.5   297     1   200  43.7  1.10 0.874  7519. #> 4 1.45  0.294  18.6  40.2  10.5   297     1   200  51.8  1.09 1.68   6520. #> 5 1.94  0.294  18.6  40.2  10.5   297     1   200  54.8  1.09 2.56   5654. #> 6 2.42  0.294  18.6  40.2  10.5   297     1   200  54.6  1.08 3.45   4903. #> # ℹ 94 more rows #> # ℹ 5 more variables: central <dbl>, peri <dbl>, eff <dbl>, ctime [h], WT <dbl> plot(r1,C2, ylab=\"Central Concentration\", xlab=\"Time\") plot(r1,eff, ylab=\"Effect\", xlab=\"Time\") r1 <- solve(mod4, ev,covsInterpolation=\"nocb\") print(r1) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>        TKA        TCL        TV2         TQ        TV3      TKout      TEC50  #>   0.294000  18.600000  40.200000  10.500000 297.000000   1.000000 200.000000  #>       TKin         Tz        amp         pi  #>   1.000000   8.000000   0.100000   3.141593  #> ── Initial Conditions ($inits): ── #>   depot central    peri     eff  #>       0       0       0       1  #> ── First part of data (object): ── #> # A tibble: 100 × 17 #>    time    KA    CL    V2     Q    V3  Kout  EC50    C2   Kin    C3  depot #>     [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> #> 1 0     0.294  18.6  40.2  10.5   297     1   200   0    1.1  0     10000  #> 2 0.485 0.294  18.6  40.2  10.5   297     1   200  27.8  1.10 0.257  8671. #> 3 0.970 0.294  18.6  40.2  10.5   297     1   200  43.7  1.10 0.874  7519. #> 4 1.45  0.294  18.6  40.2  10.5   297     1   200  51.8  1.09 1.68   6520. #> 5 1.94  0.294  18.6  40.2  10.5   297     1   200  54.8  1.09 2.56   5654. #> 6 2.42  0.294  18.6  40.2  10.5   297     1   200  54.6  1.08 3.45   4903. #> # ℹ 94 more rows #> # ℹ 5 more variables: central <dbl>, peri <dbl>, eff <dbl>, ctime [h], WT <dbl> plot(r1,C2, ylab=\"Central Concentration\", xlab=\"Time\") plot(r1,eff, ylab=\"Effect\", xlab=\"Time\")"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"creating-an-interactive-data-frame","dir":"Articles","previous_headings":"Using rxode2 data frames","what":"Creating an interactive data frame","title":"rxode2 Data Frames","text":"rxode2 supports returning solved object modified data-frame. done predict(), solve(), rxSolve() methods.","code":"library(rxode2) library(units)  ## Setup example model mod1 <- function() {   ini({     # central      KA <- 2.94E-01     CL <- 1.86E+01     # peripheral     V2 <- 4.02E+01      Q  <- 1.05E+01     V3 <- 2.97E+02     # effects     Kin  <- 1     Kout <- 1     EC50 <- 200   })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <-  KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                     Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) <- 1   }) }  ## Seup parameters and initial conditions   ## Setup dosing event information ev <- et(amountUnits=\"mg\", timeUnits = \"hours\") %>%   et(amt=10000, addl=9, ii=12) %>%   et(amt=20000, addl=4, time=120, ii=24) %>%   et(0:240)   ## Now solve x <- predict(mod1, ev)  x #> -- Solved rxode2 object -- #> -- Parameters (x$params): -- #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> -- Initial Conditions (x$inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 241 x 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # i 235 more rows"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"using-the-solved-object-as-a-simple-data-frame","dir":"Articles","previous_headings":"Using rxode2 data frames","what":"Using the solved object as a simple data frame","title":"rxode2 Data Frames","text":"solved object acts data.frame tbl can filtered dpylr. example filter easily.","code":"library(dplyr) ## You can  drop units for comparisons and filtering x <- mod1 %>% solve(ev) %>%   drop_units() %>% filter(time <= 3) %>% as_tibble()  ## or keep them and compare with the proper units. x <- mod1 %>% solve(ev) %>%     filter(time <= set_units(3, hr)) %>% as_tibble() x #> # A tibble: 4 x 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"updating-the-data-set-interactively","dir":"Articles","previous_headings":"","what":"Updating the data-set interactively","title":"rxode2 Data Frames","text":"However isn’t just simple data object. can use solved object update parameters fly, even change sampling time. First need recreate original solved system:","code":"x <- mod1 %>% solve(ev) print(x) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 241 x 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # i 235 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"modifying-observation-times-for-rxode2","dir":"Articles","previous_headings":"Updating the data-set interactively","what":"Modifying observation times for rxode2","title":"rxode2 Data Frames","text":"Notice initial effect now 2. can also change sampling times easily method changing t time. example:","code":"x$t <- seq(0,5,length.out=20) print(x) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 20 x 7 #>    time    C2     C3  depot centr  peri   eff #>     [h] <dbl>  <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1 0       0   0      10000     0    0    1    #> 2 0.263  16.8 0.0817  9255.  677.  24.3  1.01 #> 3 0.526  29.5 0.299   8566. 1187.  88.7  1.03 #> 4 0.789  38.9 0.615   7929. 1562. 183.   1.06 #> 5 1.05   45.5 1.00    7338. 1830. 298.   1.09 #> 6 1.32   50.1 1.44    6792. 2013. 427.   1.12 #> # i 14 more rows plot(x)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"modifying-simulation-parameters","dir":"Articles","previous_headings":"Updating the data-set interactively","what":"Modifying simulation parameters","title":"rxode2 Data Frames","text":"can also access change parameters $ operator. example, accessing KA can done : may change assigning new value.  can access/change parameters, initialization(s) events $params, $inits, $events accessor syntax, similar used . syntax makes easy update explore effect various parameters solved object.","code":"x$KA #> [1] 0.294 x$KA <- 1 print(x) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #>    KA    CL    V2     Q    V3   Kin  Kout  EC50  #>   1.0  18.6  40.2  10.5 297.0   1.0   1.0 200.0  #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 20 x 7 #>    time    C2    C3  depot centr   peri   eff #>     [h] <dbl> <dbl>  <dbl> <dbl>  <dbl> <dbl> #> 1 0       0   0     10000     0     0    1    #> 2 0.263  52.2 0.261  7686. 2098.   77.6  1.03 #> 3 0.526  83.3 0.900  5908. 3348.  267.   1.09 #> 4 0.789  99.8 1.75   4541. 4010.  519.   1.15 #> 5 1.05  106.  2.69   3490. 4273.  800.   1.21 #> 6 1.32  106.  3.66   2683. 4272. 1086.   1.26 #> # i 14 more rows plot(x)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"datasets-for-rxode2-nlmixr","dir":"Articles","previous_headings":"","what":"Datasets for rxode2 & nlmixr","title":"rxode2 Event Types","text":"Data input nlmixr type data input rxode2, similar data NONMEM (NONMEM-ready datasets can used directly nlmixr).","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"subject-identification-columns","dir":"Articles","previous_headings":"Columns Described by Type of Use","what":"Subject Identification Columns","title":"rxode2 Event Types","text":"subject identification column separates subjects identification random effects. ID: subject identifier may integer, character, factor.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"observation-columns","dir":"Articles","previous_headings":"Columns Described by Type of Use","what":"Observation Columns","title":"rxode2 Event Types","text":"Observation columns used indicate dependent variable use measure . DV: numeric column measurement CENS: numeric column indication censoring, limit quantification assay. LIMIT: numeric column helping indicate type censoring, limit quantification assay. MDV: indicator missing DV values CMT: name number compartment DVID: dependent variable identifier EVID event identifier","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dosing-columns","dir":"Articles","previous_headings":"Columns Described by Type of Use","what":"Dosing Columns","title":"rxode2 Event Types","text":"AMT: amount dose CMT: name number compartment EVID: event identifier ADDL: number additional doses RATE DUR: rate duration dose","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"details-for-specific-dataset-columns","dir":"Articles","previous_headings":"","what":"Details for Specific Dataset Columns","title":"rxode2 Event Types","text":"details sorted alphabetically column name. grouping use, see documentation .","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"amt-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"AMT Column","title":"rxode2 Event Types","text":"AMT column defines amount dose. observation rows, 0 NA. dosing rows, amount dose administered CMT. dose zero-order rate (constant infusion), infusion may setup using RATE DUR column.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"censlimit-columns","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"CENS/LIMIT Columns","title":"rxode2 Event Types","text":"CENS column indicator column indicating censoring occurred. pharmacokinetic modeling, censoring typically sample limit quantification. Internally rxode2 saves values nlmixr can use likelihood calculations. CENS = 0 indicates value DV measured without censoring. CENS = 1 indicates value left censored (limit quantitation) value DV censoring/quantitation limit. CENS = -1 indicates value right censored (limit quantitation) value DV censoring/quantitation limit. LIMIT additional information censoring handled nlmixr stored rxode2’s data structure well. value left censored, like limit 1 may also believe value certain threshold, like zero. case, limit 0 indicates censored value 0 1. short : CENS = 0 LIMIT ignored observation censored CENS = 1 value censored (LIMIT, DV) CENS = -1 value censored (DV, LIMIT)","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"cmt-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"CMT Column","title":"rxode2 Event Types","text":"CMT column indicates compartment event occurs. given character string factor (preferred method), matched name model. given integer, matched order compartments appear model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dur-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"DUR Column","title":"rxode2 Event Types","text":"DUR column defines duration infusion. used set duration zero-order rate infusion.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dv-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"DV Column","title":"rxode2 Event Types","text":"DV column indicates current measurement current compartment (see CMT) current measurement identifier (see DVID) may missing (see MDV) censored (see CENS).","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dvid-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"DVID Column","title":"rxode2 Event Types","text":"TODO","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"evid-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"EVID Column","title":"rxode2 Event Types","text":"EVID column event identifier row data. observation records, 0. normal dosing records, 1. Many EVID values detailed rxode2 Event Types Classic rxode2 Events vignettes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"id-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"ID Column","title":"rxode2 Event Types","text":"ID column subject identifier. column used separate one individual (usually single person animal) another. model, ID column used separate individuals. numerical integrator re-initializes new individual, new values random effects selected.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"rate-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"RATE Column","title":"rxode2 Event Types","text":"TODO","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"creating-rxode2s-event-tables","dir":"Articles","previous_headings":"","what":"Creating rxode2’s event tables","title":"Easy rxode2 Event Tables","text":"event table rxode2 specialized data frame acts container rxode2’s events observation times. create rxode2 event table may use code eventTable(), et(), even create data frame right event information contained . closely related types events rxode2 supports. event table can add sampling/observations doses piping direct access. short table two main functions create dosing Sampling times can added add.sampling( sampling times ) et( sampling times ). Dosing intervals sampling windows also supported. documentation using et() syntax, though supported. models, can illustrate using model shared rxode2 tutorial:","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` library(units) #> udunits database from /usr/share/xml/udunits/udunits2.xml (ev <- et()) #> -- EventTable with 0 records -- #> 0 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) (ev <- et()) #> -- EventTable with 0 records -- #> 0 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) ## Model from rxode2 tutorial m1 <-function() {   ini({     KA   <- 2.94E-01     CL   <- 1.86E+01     V2   <- 4.02E+01     Q    <- 1.05E+01     V3   <- 2.97E+02     Kin  <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250   })   model({     C2          <- centr/V2     C3          <- peri/V3     d/dt(depot) <- -KA*depot     f(depot)    <- fdepot     dur(depot)  <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff;     eff(0)      <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"adding-doses-to-the-event-table","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Adding doses to the event table","title":"Easy rxode2 Event Tables","text":"created can add dosing event table add.dosing(), et() functions. Using add.dosing() function : Notice units specified table. specified, units use units package keep track units convert needed. Additionally, ggforce uses label ggplot axes. set_units drop_units useful set drop rxode2 event table units. example, can see time axes labeled:  familiar NONMEM/rxode2 event records, can also specify dosing using et dose elements directly: gives:  shows easy creating event tables can .","code":"ev <- eventTable(amount.units=\"mg\", time.units=\"hr\")  ## The methods ar attached to the event table, so you can use  ## them directly ev$add.dosing(dose=10000, nbr.doses = 3)# loading doses ## Starts at time 0; Default dosing interval is 24  ## You can also pipe the event tables to these methods. ev <- ev %>%   add.dosing(dose=5000, nbr.doses=14,              dosing.interval=12)# maintenance  ev #> -- EventTable with 2 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 2 x 5 #>   time   amt  ii  addl evid         #>    [h]  [mg] [h] <int> <evid>       #> 1    0 10000  24     2 1:Dose (Add) #> 2    0  5000  12    13 1:Dose (Add) rxSolve(m1, ev) %>% plot(C2) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ev <- et(timeUnits=\"hr\") %>%   et(amt=10000, until = set_units(3, days),      ii=12) # loading doses  ev #> -- EventTable with 1 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 1 x 5 #>   time   amt  ii  addl evid         #>    [h] <dbl> [h] <int> <evid>       #> 1    0 10000  12     6 1:Dose (Add) rxSolve(m1, ev) %>% plot(C2) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"adding-sampling-to-an-event-table","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Adding sampling to an event table","title":"Easy rxode2 Event Tables","text":"notice examples, rxode2 generated default sampling times since sampling times. wish control sampling time, add samples rxode2 event table add.sampling et gives:  use et can simply add similar way add.sampling: gives following rxode2 solve:  Note jagged nature plots since sample times.","code":"ev <- eventTable(amount.units=\"mg\", time.units=\"hr\")  ## The methods ar attached to the event table, so you can use them ## directly ev$add.dosing(dose=10000, nbr.doses = 3)# loading doses  ev$add.sampling(seq(0,24,by=4))  ev #> -- EventTable with 8 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 7 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 8 x 5 #>   time   amt  ii  addl evid          #>    [h]  [mg] [h] <int> <evid>        #> 1    0    NA  NA    NA 0:Observation #> 2    0 10000  24     2 1:Dose (Add)  #> 3    4    NA  NA    NA 0:Observation #> 4    8    NA  NA    NA 0:Observation #> 5   12    NA  NA    NA 0:Observation #> 6   16    NA  NA    NA 0:Observation #> 7   20    NA  NA    NA 0:Observation #> 8   24    NA  NA    NA 0:Observation solve(m1, ev) %>% plot(C2) ev <- et(timeUnits=\"hr\") %>%   et(amt=10000, until = set_units(3, days),      ii=12) %>% # loading doses   et(seq(0,24,by=4))  ev #> -- EventTable with 8 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 7 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 8 x 5 #>   time   amt  ii  addl evid          #>    [h] <dbl> [h] <int> <evid>        #> 1    0    NA  NA    NA 0:Observation #> 2    0 10000  12     6 1:Dose (Add)  #> 3    4    NA  NA    NA 0:Observation #> 4    8    NA  NA    NA 0:Observation #> 5   12    NA  NA    NA 0:Observation #> 6   16    NA  NA    NA 0:Observation #> 7   20    NA  NA    NA 0:Observation #> 8   24    NA  NA    NA 0:Observation solve(m1, ev) %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"expand-the-event-table-to-a-multi-subject-event-table-","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Expand the event table to a multi-subject event table.","title":"Easy rxode2 Event Tables","text":"thing needed expand event table list IDs want expand; can see following simulation 4 individuals solved :","code":"ev <- et(timeUnits=\"hr\") %>%   et(amt=10000, until = set_units(3, days),      ii=12) %>% # loading doses   et(seq(0,48,length.out=200)) %>%   et(id=1:4)  ev #> -- EventTable with 804 records -- #> 4 dosing records (see x$get.dosing(); add with add.dosing or et) #> 800 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 804 x 6 #>       id  time   amt  ii  addl evid          #>    <int>   [h] <dbl> [h] <int> <evid>        #>  1     1 0        NA  NA    NA 0:Observation #>  2     1 0     10000  12     6 1:Dose (Add)  #>  3     1 0.241    NA  NA    NA 0:Observation #>  4     1 0.482    NA  NA    NA 0:Observation #>  5     1 0.724    NA  NA    NA 0:Observation #>  6     1 0.965    NA  NA    NA 0:Observation #>  7     1 1.21     NA  NA    NA 0:Observation #>  8     1 1.45     NA  NA    NA 0:Observation #>  9     1 1.69     NA  NA    NA 0:Observation #> 10     1 1.93     NA  NA    NA 0:Observation #> # i 794 more rows set.seed(42) rxSetSeed(42) solve(m1, ev,       params=data.frame(KA=0.294*exp(rnorm(4)),                         CL=18.6*exp(rnorm(4)))) %>%     plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"add-doses-and-samples-within-a-sampling-window","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Add doses and samples within a sampling window","title":"Easy rxode2 Event Tables","text":"addition adding fixed doses fixed sampling times, can windows sample draw doses . dosing windows specify time ordered numerical vector lowest dosing time highest dosing time inside list. example, start dosing time 6 hour dosing window: can clearly see different dosing times following simulation:  course reality dosing interval may 2 hours:  sort thing can specified sampling times. specify sampling times terms sampling window, can create list sampling times. sampling time two element ordered numeric vector.  shows flexibility dosing sampling rxode2 event tables allow.","code":"set.seed(42) rxSetSeed(42) ev <- et(timeUnits=\"hr\") %>%   et(time=list(c(0,6)), amt=10000, until = set_units(2, days),      ii=12) %>% # loading doses   et(id=1:4)  ev #> -- EventTable with 16 records -- #> 16 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 16 x 6 #>       id low   time high   amt evid         #>    <int> [h]    [h]  [h] <dbl> <evid>       #>  1     1   0  5.49     6 10000 1:Dose (Add) #>  2     1  12 17.0     18 10000 1:Dose (Add) #>  3     1  24 25.7     30 10000 1:Dose (Add) #>  4     1  36 41.6     42 10000 1:Dose (Add) #>  5     2   0  4.31     6 10000 1:Dose (Add) #>  6     2  12 14.7     18 10000 1:Dose (Add) #>  7     2  24 28.2     30 10000 1:Dose (Add) #>  8     2  36 39.9     42 10000 1:Dose (Add) #>  9     3   0  0.808    6 10000 1:Dose (Add) #> 10     3  12 16.4     18 10000 1:Dose (Add) #> 11     3  24 27.1     30 10000 1:Dose (Add) #> 12     3  36 39.9     42 10000 1:Dose (Add) #> 13     4   0  4.98     6 10000 1:Dose (Add) #> 14     4  12 13.7     18 10000 1:Dose (Add) #> 15     4  24 29.6     30 10000 1:Dose (Add) #> 16     4  36 41.5     42 10000 1:Dose (Add) ev <- ev %>% et(seq(0,48,length.out=200))  solve(m1, ev,       params=data.frame(KA=0.294*exp(rnorm(4)),                         CL=18.6*exp(rnorm(4)))) %>%   plot(C2) set.seed(42) rxSetSeed(42) ev <- et(timeUnits=\"hr\") %>%   et(time=list(c(0,2)), amt=10000, until = set_units(2, days),      ii=12) %>% # loading doses   et(id=1:4) %>%   et(seq(0,48,length.out=200))  solve(m1, ev,       params=data.frame(KA=0.294*exp(rnorm(4)),                         CL=18.6*exp(rnorm(4)))) %>%   plot(C2) rxSetSeed(42) set.seed(42) ev <- et(timeUnits=\"hr\") %>%   et(time=list(c(0,2)), amt=10000, until = set_units(2, days),      ii=12) %>% # loading doses   et(id=1:4)  ## Create 20 samples in the first 24 hours and 20 samples in the ## second 24 hours samples <- c(lapply(1:20, function(...){c(0,24)}),              lapply(1:20, function(...){c(20,48)}))  ## Add the random collection to the event table ev <- ev %>% et(samples)  library(ggplot2) solve(m1, ev, params=data.frame(KA=0.294*exp(rnorm(4)),                                 CL=18.6*exp(rnorm(4)))) %>%   plot(C2) + geom_point()"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"combining-event-tables","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Combining event tables","title":"Easy rxode2 Event Tables","text":"Since can create dosing records sampling records, can create complex dosing regimen wish. addition, rxode2 allows combine event tables c, seq, rep, rbind.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"sequencing-event-tables","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Sequencing event tables","title":"Easy rxode2 Event Tables","text":"One way combine event table sequence c, seq etSeq. takes two dosing groups adds least one inter-dose interval :  sequencing events, can also separate sequence period time; example wanted separate week, easily following sequence event tables:  Note example time bid qd event tables exactly one week, 1 week plus 24 hours inter-dose interval. want behavior, can sequence using wait=\"+ii\".  Also note, rxode2 assumes dosing want space event tables , clears sampling records combine event tables. true, can also use option samples=\"use\"","code":"## bid for 5 days bid <- et(timeUnits=\"hr\") %>%        et(amt=10000,ii=12,until=set_units(5, \"days\"))  ## qd for 5 days qd <- et(timeUnits=\"hr\") %>%       et(amt=20000,ii=24,until=set_units(5, \"days\"))  ## bid for 5 days followed by qd for 5 days et <- seq(bid,qd) %>% et(seq(0,11*24,length.out=100));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days followed by qd for 5 days et <- seq(bid,set_units(1, \"week\"), qd) %>%     et(seq(0,18*24,length.out=100));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days followed by qd for 5 days et <- seq(bid,set_units(1, \"week\"), qd,wait=\"+ii\") %>%     et(seq(0,18*24,length.out=100));  rxSolve(m1, et) %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"repeating-event-tables","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Repeating event tables","title":"Easy rxode2 Event Tables","text":"can event table can repeat etRep rep. example 4 rounds 2 weeks QD therapy 1 week therapy can simply specified:  simplified way use sequence event tables. Therefore, many options still apply; samples cleared unless use samples=\"use\", time event tables least inter-dose interval. can adjust timing wait option.","code":"qd <-et(timeUnits = \"hr\") %>%   et(amt=10000, ii=24, until=set_units(2, \"weeks\"), cmt=\"depot\")  et <- rep(qd, times=4, wait=set_units(1,\"weeks\")) %>%       add.sampling(set_units(seq(0, 12.5,by=0.005),weeks))  rxSolve(m1, et)  %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"combining-event-tables-with-rbind","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Combining event tables with rbind","title":"Easy rxode2 Event Tables","text":"may combine event tables rbind. consider event times combining event tables, keeps times. space event tables waiting period, also consider inter-dose interval. Using previous seq can clearly see difference. sequence:  bind together rbind  Still waiting period applies (consider inter-dose interval)  can also bind tables together make ID event table unique; can good combine cohorts different expected dosing sampling times. requires id=\"unique\" option; Using first example shows different case:","code":"## bid for 5 days bid <- et(timeUnits=\"hr\") %>%        et(amt=10000,ii=12,until=set_units(5, \"days\"))  ## qd for 5 days qd <- et(timeUnits=\"hr\") %>%       et(amt=20000,ii=24,until=set_units(5, \"days\"))  et <- seq(bid,qd) %>%     et(seq(0,18*24,length.out=500));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days et <- rbind(bid,qd) %>%     et(seq(0,18*24,length.out=500));  rxSolve(m1, et) %>% plot(C2) et <- rbind(bid,wait=set_units(10,days),qd) %>%     et(seq(0,18*24,length.out=500));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days et <- etRbind(bid,qd, id=\"unique\") %>%     et(seq(0,150,length.out=500));  library(ggplot2) rxSolve(m1, et) %>% plot(C2) + facet_wrap( ~ id)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"event-tables-in-rstudio-notebooks","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Event tables in Rstudio Notebooks","title":"Easy rxode2 Event Tables","text":"addition output console shown examples, Rstudio notebook output different can seen following screenshots; first screenshot shows event table looks evaluating Rstduio notebook  simple dataframe allows page contents. click first box Rstudio notebook output, notes event table: ## Expanding events Event tables can expanded contain addl data item, like following example: can expand events addl items $expand() etExpand(ev): first, etExpand(ev) expands event table without modifying original data frame: can see addl events expanded, however original data frame remained intact: use ev$expand() modify ev object. similar object-oriented method:","code":"ev <- et() %>%   et(dose=50, ii=8, until=48)  ev #> -- EventTable with 1 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 1 x 5 #>    time   amt    ii  addl evid         #>   <dbl> <dbl> <dbl> <int> <evid>       #> 1     0    50     8     6 1:Dose (Add) etExpand(ev) #> -- EventTable with 7 records -- #> 7 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 7 x 4 #>    time   amt    ii evid         #>   <dbl> <dbl> <dbl> <evid>       #> 1     0    50     0 1:Dose (Add) #> 2     8    50     0 1:Dose (Add) #> 3    16    50     0 1:Dose (Add) #> 4    24    50     0 1:Dose (Add) #> 5    32    50     0 1:Dose (Add) #> 6    40    50     0 1:Dose (Add) #> 7    48    50     0 1:Dose (Add) print(ev) #> -- EventTable with 1 records -- #> 1 dosing records (see $get.dosing(); add with add.dosing or et) #> 0 observation times (see $get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with $expand(); or etExpand() #> -- First part of : -- #> # A tibble: 1 x 5 #>    time   amt    ii  addl evid         #>   <dbl> <dbl> <dbl> <int> <evid>       #> 1     0    50     8     6 1:Dose (Add) ev$expand() ev #> -- EventTable with 7 records -- #> 7 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 7 x 4 #>    time   amt    ii evid         #>   <dbl> <dbl> <dbl> <evid>       #> 1     0    50     0 1:Dose (Add) #> 2     8    50     0 1:Dose (Add) #> 3    16    50     0 1:Dose (Add) #> 4    24    50     0 1:Dose (Add) #> 5    32    50     0 1:Dose (Add) #> 6    40    50     0 1:Dose (Add) #> 7    48    50     0 1:Dose (Add)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"rxode2-event-tables","dir":"Articles","previous_headings":"","what":"rxode2 event tables","title":"rxode2 Event Types","text":"general, rxode2 event tables follow NONMEM dataset convention exceptions: may turn compartment negative compartment number “-cmt” cmt compartment name. compartment data item (cmt) can still number, number compartment defined appearance compartment name model. can tedious count, can specify compartment numbers easier using cmt(cmtName) beginning model. Bioavailability changes change rate infusion since dur/amt fixed input data. Similarly, specifying rate/amt infusion, bioavailability change infusion duration since rate/amt fixed input data. infrequent NONMEM columns supported: pcmt, call. evid=5 replace event; replaces value compartment value specified amt column. equivalent deSolve=replace. evid=6 multiply event; multiplies value compartment value specified amt column. equivalent deSolve=multiply. evid=7 transit compartment model/phantom event. puts dose dose() function calculates time since last dose tad() doesn’t actually put dose compartment. allows transit() function easily apply compartment. legal entries data table: notes: evid can classic RxODE (described ) NONMEM-style evid described . NONMEM’s DV required; rxode2 ODE solving framework. NONMEM’s MDV required, since captured EVID. Instead NONMEM-compatible data, can accept deSolve compatible data-frames. returning rxode2 solved data-set additional event ids (EVID) may see depending solving options: EVID = -1 modeled rate ends (corresponds rate = -1) EVID = -2 modeled duration ends (corresponds rate=-2) EVID = -10 rate specified zero-order infusion ends (corresponds rate > 0) EVID = -20 duration specified zero-order infusion ends (corresponds dur > 0) EVID = 101, 102, 103,... correspond 1, 2, 3, … modeled time (mtime). can accessed solving option combination addDosing=TRUE subsetNonmem=FALSE. want see classic EVID equivalents can use addDosing=NA. illustrate event types use model original rxode2 tutorial.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` ## Model from rxode2 tutorial  m1 <- function() {   ini({     KA   <- 2.94E-01     CL   <- 1.86E+01     V2   <- 4.02E+01     Q    <- 1.05E+01     V3   <- 2.97E+02     Kin  <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250   })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     f(depot) <- fdepot     dur(depot) <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)  <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"bolusadditive-doses","dir":"Articles","previous_headings":"","what":"Bolus/Additive Doses","title":"rxode2 Event Types","text":"bolus dose default type dose rxode2 requires amt/dose. Note uses convenience function et() described rxode2 event tables","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 5 #>     time   amt  ii  addl evid          #>      [h] <dbl> [h] <int> <evid>        #>  1 0        NA  NA    NA 0:Observation #>  2 0     10000  12     2 1:Dose (Add)  #>  3 0.242    NA  NA    NA 0:Observation #>  4 0.485    NA  NA    NA 0:Observation #>  5 0.727    NA  NA    NA 0:Observation #>  6 0.970    NA  NA    NA 0:Observation #>  7 1.21     NA  NA    NA 0:Observation #>  8 1.45     NA  NA    NA 0:Observation #>  9 1.70     NA  NA    NA 0:Observation #> 10 1.94     NA  NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev) %>% plot(C2) +     xlab(\"Time\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"infusion-doses","dir":"Articles","previous_headings":"","what":"Infusion Doses","title":"rxode2 Event Types","text":"different type infusions rxode2 supports: Constant Rate Infusion (rate) Constant Duration Infusion (dur) Estimated Rate Infusion Estimated Duration Infusion","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"constant-infusion-in-terms-of-duration-and-rate","dir":"Articles","previous_headings":"Infusion Doses","what":"Constant Infusion (in terms of duration and rate)","title":"rxode2 Event Types","text":"next type event infusion; two ways specify infusion; first dur keyword. example :  can also specified rate component:  exception bioavailability changes infusion. case modeling rate, bioavailability decrease, decreases infusion duration, NONMEM. example:  Similarly increasing bioavailability increases infusion duration.  rationale behavior rate amt specified event table, thing can change bioavailability increase duration infusion. specify amt dur components event table, bioavailability changes affect rate infusion. can see side--side comparison bioavailability changes affecting rate instead duration records following plots:","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, dur=8) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt  ii  addl evid          dur #>      [h] <dbl> [h] <int> <evid>        [h] #>  1 0        NA  NA    NA 0:Observation  NA #>  2 0     10000  12     2 1:Dose (Add)    8 #>  3 0.242    NA  NA    NA 0:Observation  NA #>  4 0.485    NA  NA    NA 0:Observation  NA #>  5 0.727    NA  NA    NA 0:Observation  NA #>  6 0.970    NA  NA    NA 0:Observation  NA #>  7 1.21     NA  NA    NA 0:Observation  NA #>  8 1.45     NA  NA    NA 0:Observation  NA #>  9 1.70     NA  NA    NA 0:Observation  NA #> 10 1.94     NA  NA    NA 0:Observation  NA #> # i 91 more rows rxSolve(m1, ev) %>% plot(depot, C2) +     xlab(\"Time\") ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, rate=10000/8) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt rate        ii  addl evid          #>      [h] <dbl> <rate/dur> [h] <int> <evid>        #>  1 0        NA NA          NA    NA 0:Observation #>  2 0     10000  1250       12     2 1:Dose (Add)  #>  3 0.242    NA NA          NA    NA 0:Observation #>  4 0.485    NA NA          NA    NA 0:Observation #>  5 0.727    NA NA          NA    NA 0:Observation #>  6 0.970    NA NA          NA    NA 0:Observation #>  7 1.21     NA NA          NA    NA 0:Observation #>  8 1.45     NA NA          NA    NA 0:Observation #>  9 1.70     NA NA          NA    NA 0:Observation #> 10 1.94     NA NA          NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev) %>% plot(depot, C2) +     xlab(\"Time\") rxSolve(m1, ev, c(fdepot=0.25)) %>% plot(depot, C2) +     xlab(\"Time\") rxSolve(m1, ev, c(fdepot=1.25)) %>% plot(depot, C2) +     xlab(\"Time\") ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, dur=8) %>%     et(seq(0, 24, length.out=100)) library(ggplot2) library(patchwork)  p1 <- rxSolve(m1, ev, c(fdepot=1.25)) %>% plot(depot) +     xlab(\"Time\") + ylim(0,5000)  p2 <- rxSolve(m1, ev, c(fdepot=0.25)) %>% plot(depot) +     xlab(\"Time\")+ ylim(0,5000)  ## Use patchwork syntax to combine plots p1 * p2"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"modeled-rate-and-duration-of-infusion","dir":"Articles","previous_headings":"Infusion Doses","what":"Modeled Rate and Duration of Infusion","title":"rxode2 Event Types","text":"can model duration, equivalent NONMEM’s rate=-2.  Similarly, may also model rate. equivalent NONMEM’s rate=-1 rxode2’s event table specifies data item well.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, rate=-2) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt rate        ii  addl evid          #>      [h] <dbl> <rate/dur> [h] <int> <evid>        #>  1 0        NA NA          NA    NA 0:Observation #>  2 0     10000 -2:dur      12     2 1:Dose (Add)  #>  3 0.242    NA NA          NA    NA 0:Observation #>  4 0.485    NA NA          NA    NA 0:Observation #>  5 0.727    NA NA          NA    NA 0:Observation #>  6 0.970    NA NA          NA    NA 0:Observation #>  7 1.21     NA NA          NA    NA 0:Observation #>  8 1.45     NA NA          NA    NA 0:Observation #>  9 1.70     NA NA          NA    NA 0:Observation #> 10 1.94     NA NA          NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev, c(durDepot=7)) %>% plot(depot, C2) +     xlab(\"Time\") ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, rate=-1) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt rate        ii  addl evid          #>      [h] <dbl> <rate/dur> [h] <int> <evid>        #>  1 0        NA NA          NA    NA 0:Observation #>  2 0     10000 -1:rate     12     2 1:Dose (Add)  #>  3 0.242    NA NA          NA    NA 0:Observation #>  4 0.485    NA NA          NA    NA 0:Observation #>  5 0.727    NA NA          NA    NA 0:Observation #>  6 0.970    NA NA          NA    NA 0:Observation #>  7 1.21     NA NA          NA    NA 0:Observation #>  8 1.45     NA NA          NA    NA 0:Observation #>  9 1.70     NA NA          NA    NA 0:Observation #> 10 1.94     NA NA          NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev, c(rateDepot=10000/3)) %>% plot(depot, C2) +     xlab(\"Time\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"steady-state","dir":"Articles","previous_headings":"","what":"Steady State","title":"rxode2 Event Types","text":"doses solved steady state reached constant inter-dose interval.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, ss=1) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 101 x 5 #>     time   amt  ii evid             ss #>      [h] <dbl> [h] <evid>        <int> #>  1 0        NA  NA 0:Observation    NA #>  2 0     10000  12 1:Dose (Add)      1 #>  3 0.242    NA  NA 0:Observation    NA #>  4 0.485    NA  NA 0:Observation    NA #>  5 0.727    NA  NA 0:Observation    NA #>  6 0.970    NA  NA 0:Observation    NA #>  7 1.21     NA  NA 0:Observation    NA #>  8 1.45     NA  NA 0:Observation    NA #>  9 1.70     NA  NA 0:Observation    NA #> 10 1.94     NA  NA 0:Observation    NA #> # i 91 more rows rxSolve(m1, ev) %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"steady-state-for-complex-dosing","dir":"Articles","previous_headings":"Steady State","what":"Steady state for complex dosing","title":"rxode2 Event Types","text":"using ss=2 flag, can use super-positioning principle linear kinetics get steady state nonstandard dosing (.e. morning 100 mg vs evening 150 mg). done : Saving state values Resetting states solving system steady state Adding back prior state values  can see takes full dose cycle reach true complex steady state dosing.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=24, ss=1) %>%     et(time=12, amt=15000, ii=24, ss=2) %>%     et(time=24, amt=10000, ii=24, addl=3) %>%     et(time=36, amt=15000, ii=24, addl=3) %>%     et(seq(0, 64, length.out=500))  library(ggplot2)  rxSolve(m1, ev,maxsteps=10000) %>% plot(C2) +   annotate(\"rect\", xmin=0, xmax=24, ymin=-Inf, ymax=Inf,            alpha=0.2) +   annotate(\"text\", x=12.5, y=7,            label=\"Initial Steady State Period\") +   annotate(\"text\", x=44,   y=7,            label=\"Steady State AM/PM dosing\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"steady-state-for-constant-infusion-or-zero-order-processes","dir":"Articles","previous_headings":"Steady State","what":"Steady state for constant infusion or zero order processes","title":"rxode2 Event Types","text":"last type steady state rxode2 supports steady-state constant infusion rate. can specified way NONMEM, : inter-dose interval ii=0 steady state dose, ie ss=1 Either positive rate (rate>0) estimated rate rate=-1. zero dose, ie amt=0 steady-state constant infusion achieved, infusion turned using record, just like NONMEM. Note rate=-2 model duration infusion doesn’t make much sense since solving infusion steady state. duration specified steady state solution. Also note bioavailability changes steady state infusion also make sense neither change rate duration steady state infusion. Hence modeled bioavailability type dosing event ignored. example:  can used PK, can used steady-state disease processes.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=0, ss=1,rate=10000/8)  p1 <- rxSolve(m1, ev) %>% plot(C2, eff)   ev <- et(timeUnits=\"hr\") %>%     et(amt=200000, rate=10000/8) %>%     et(0, 250, length.out=1000)  p2 <- rxSolve(m1, ev) %>% plot(C2, eff)  library(patchwork)  p1 / p2"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"reset-events","dir":"Articles","previous_headings":"","what":"Reset Events","title":"rxode2 Event Types","text":"Reset events implemented evid=3 evid=reset, reset evid=4 reset dose. solving show happens system system reset 6 hours post-dose.  can see compartments reset initial values. next dose start dosing cycle . case, whole system reset dose given","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, evid=reset) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 102 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 102 x 5 #>     time   amt  ii  addl evid          #>      [h] <dbl> [h] <int> <evid>        #>  1 0        NA  NA    NA 0:Observation #>  2 0     10000  12     3 1:Dose (Add)  #>  3 0.242    NA  NA    NA 0:Observation #>  4 0.485    NA  NA    NA 0:Observation #>  5 0.727    NA  NA    NA 0:Observation #>  6 0.970    NA  NA    NA 0:Observation #>  7 1.21     NA  NA    NA 0:Observation #>  8 1.45     NA  NA    NA 0:Observation #>  9 1.70     NA  NA    NA 0:Observation #> 10 1.94     NA  NA    NA 0:Observation #> # i 92 more rows rxSolve(m1, ev) %>% plot(depot,C2, eff) ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, amt=10000, evid=4) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 102 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 102 x 5 #>     time   amt  ii  addl evid          #>      [h] <dbl> [h] <int> <evid>        #>  1 0        NA  NA    NA 0:Observation #>  2 0     10000  12     3 1:Dose (Add)  #>  3 0.242    NA  NA    NA 0:Observation #>  4 0.485    NA  NA    NA 0:Observation #>  5 0.727    NA  NA    NA 0:Observation #>  6 0.970    NA  NA    NA 0:Observation #>  7 1.21     NA  NA    NA 0:Observation #>  8 1.45     NA  NA    NA 0:Observation #>  9 1.70     NA  NA    NA 0:Observation #> 10 1.94     NA  NA    NA 0:Observation #> # i 92 more rows rxSolve(m1, ev) %>% plot(depot,C2, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"turning-off-compartments","dir":"Articles","previous_headings":"","what":"Turning off compartments","title":"rxode2 Event Types","text":"may also turn compartment, similar reset event. Solving shows system:  case, depot turned , depot compartment concentrations set initial values compartment concentrations/levels reset. another dose depot administered depot compartment turned back . Note dose compartment turns back compartment dosed. Hence turn effect compartment, continues another dose depot.  turn back compartment, zero-dose compartment evid=2 compartment needed.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, cmt=\"-depot\", evid=2) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 102 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 102 x 6 #>     time cmt         amt  ii  addl evid          #>      [h] <chr>     <dbl> [h] <int> <evid>        #>  1 0     (obs)        NA  NA    NA 0:Observation #>  2 0     (default) 10000  12     3 1:Dose (Add)  #>  3 0.242 (obs)        NA  NA    NA 0:Observation #>  4 0.485 (obs)        NA  NA    NA 0:Observation #>  5 0.727 (obs)        NA  NA    NA 0:Observation #>  6 0.970 (obs)        NA  NA    NA 0:Observation #>  7 1.21  (obs)        NA  NA    NA 0:Observation #>  8 1.45  (obs)        NA  NA    NA 0:Observation #>  9 1.70  (obs)        NA  NA    NA 0:Observation #> 10 1.94  (obs)        NA  NA    NA 0:Observation #> # i 92 more rows rxSolve(m1, ev) %>% plot(depot,C2, eff) ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, cmt=\"-eff\", evid=2) %>%     et(seq(0, 24, length.out=100))  rxSolve(m1, ev) %>% plot(depot,C2, eff) ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, cmt=\"-eff\", evid=2) %>%     et(time=12,cmt=\"eff\",evid=2) %>%     et(seq(0, 24, length.out=100))  rxSolve(m1, ev) %>% plot(depot,C2, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"classic-rxode-evid-values","dir":"Articles","previous_headings":"","what":"Classic RxODE evid values","title":"Classic rxode2 Events","text":"Originally RxODE supported compound event IDs; rxode2 still supports parameters, often useful use normal NONMEM dataset standard used many modeling tools like NONMEM, Monolix nlmixr, described rxode2 types article. Classically, RxODE supported event coding single event id evid described following table. classic EVID concatenate numbers table, infusion compartment 1 10101 infusion compartment 199 119901. EVID = 0 (observations), EVID=2 (type event) EVID=3 supported. Internally EVID=9 non-observation event makes sure system initialized zero; EVID=9 manually set. EVID 10-99 represents modeled time interventions, similar NONMEM’s MTIME. along amount (amt) time columns specify events ODE system. infusions specified EVIDs > 100 amt column represents rate value. Infusion flags 1 2 +amt turn infusion specific compartment -amt turn infusion specific compartment. specify dose/duration place dosing records time duration starts stops. modeled rate/duration infusion flags infusion flag must followed infusion record. number concatenated together form full RxODE event ID, shown following examples:","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"bolus-dose-examples","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Bolus Dose Examples","title":"Classic rxode2 Events","text":"100 bolus dose compartment #1 time 0 100 bolus dose compartment #99 time 0 100 bolus dose compartment #199 time 0","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"infusion-event-examples","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Infusion Event Examples","title":"Classic rxode2 Events","text":"Bolus infusion rate 50 compartment 1 1.5 hr, (modeled bioavailability changes duration infusion) Bolus infusion rate 50 compartment 1 1.5 hr (modeled bioavailability changes rate infusion) Modeled rate amount 50 Modeled duration amount 50","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"steady-state-for-classic-rxode-evid-example","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Steady State for classic RxODE EVID example","title":"Classic rxode2 Events","text":"Steady state dose cmt 1 Steady State super-positioning principle 50 pm 100 dose","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"steady-state-with-lagged-dose-for-classic-rxode-evid","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Steady state with lagged dose for classic RxODE evid","title":"Classic rxode2 Events","text":"Steady state lagged dose bolus dose: time | evid | amt | ii | 0 | 109 | 100 | 24 | 0 | 101 | 100 | 0 | event 109 calculates trough amount steady state passed (ii-lag_time) event 101 applies next steady state dose. Steady state (=2) lagged bolus dose Steady state infusion lag time time | evid | amt | ii | 0 | 10109 | 5 | 24 | 0 | 10108 | -5 | 24 | 0 | 10101 | 5 | | 20 | 10101 | -5 | | case bolus, 10109 event calculates trough concentration ii=24, may (may ) still infusion running. infusion running 10108 event turn infusion appropriate time. infusion completed, 10108 event ignored. next 2 events 10101 represent event times (assuming bioavailability applied).","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"turning-off-a-compartment-with-classic-rxode-evid","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Turning off a compartment with classic RxODE EVID","title":"Classic rxode2 Events","text":"Turn first compartment time 12 Event coding rxode2 encoded single event number evid. compartments 100, coded : event 0 observation events. 100*(Compartment Number) + 1 dose captured amt 10000 + 100*(Compartment Number) + 1 infusion rate captured amt column infusion turned subtracting amt evid stop infusion. compartments greater equal 100, 100s place digits transferred 100,000th place digit. doses 99th compartment evid bolus dose 9901 evid infusion 19901. bolus dose 199th compartment evid bolus dose 109901. infusion dosing record 199th compartment 119901.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"introduction","dir":"Articles","previous_headings":"","what":"Introduction","title":"Introduction to rxode2","text":"rxode2 R package facilitates simulation ODE models R. designed pharmacometrics models mind, can applied generally ODE model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"description-of-rxode2-illustrated-through-an-example","dir":"Articles","previous_headings":"","what":"Description of rxode2 illustrated through an example","title":"Introduction to rxode2","text":"model equations can specified text string, model file R expression. differential algebraic equations permitted. Differential equations specified d/dt(var_name) =. equation can separated semicolon. load rxode2 package compile model: Model parameters may specified ini({}) model block, initial conditions can specified within model cmt(0)= X, like model eff(0) <- 1. may also specify subject variability initial conditions residual error components just like nlmixr2. allows single interface nlmixr2/rxode2 models. Also note, classic rxode2 interface still works just like past (don’t worry breaking code time). fact, can get classic rxode2 model $simulationModel ui object:","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod1 <- function() {   ini({     # central      KA=2.94E-01     CL=1.86E+01     V2=4.02E+01     # peripheral     Q=1.05E+01     V3=2.97E+02     # effects     Kin=1     Kout=1     EC50=200    })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <- Q*C2 - Q*C3     eff(0) <- 1     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff   }) } mod1 <- mod1() # create the ui object (can also use `rxode2(mod1)`) mod1  summary(mod1$simulationModel)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"specify-dosing-and-sampling-in-rxode2","dir":"Articles","previous_headings":"Description of rxode2 illustrated through an example","what":"Specify Dosing and sampling in rxode2","title":"Introduction to rxode2","text":"rxode2 provides simple flexible way specify dosing sampling functions generate event table. First, empty event table generated “et()” function. interface similar NONMEM event tables: can see code, can dose compartment named rxode2 model. slight deviation NONMEM can reduce need compartment renumbering. events can also combined expanded (multi-subject events complex regimens) rbind, c, seq, rep. information creating complex dosing regimens using rxode2 see rxode2 events vignette.","code":"ev  <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(0:240) # Add sampling"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"solving-odes","dir":"Articles","previous_headings":"Description of rxode2 illustrated through an example","what":"Solving ODEs","title":"Introduction to rxode2","text":"ODE can now solved using rxSolve: returns modified data frame. can see compartment values plot :  ,  Note labels automatically labeled units initial event table. rxode2 extracts units label plot (present).","code":"x <- mod1 %>% rxSolve(ev) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ x #> ── Solved rxode2 object ── #> ── Parameters (x$params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions (x$inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows library(ggplot2) plot(x,C2) + ylab(\"Central Concentration\") plot(x,eff)  + ylab(\"Effect\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-model-types.html","id":"prediction-only-models","dir":"Articles","previous_headings":"","what":"Prediction only models","title":"rxode2 additional model types","text":"Prediction models simple create. use rxode2 syntax without ODE systems . simple example one-compartment model. Solving rxode2 models saving simple ODE system, faster course.","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` mod <- function(){   model({     ipre <- 10 * exp(-ke * t)   }) } et  <- et(seq(0,24,length.out=50)) cmt1 <- rxSolve(mod,et,params=c(ke=0.5)) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ cmt1 ## -- Solved rxode2 object -- ## -- Parameters (x$params): -- ##  ke  ## 0.5  ## -- Initial Conditions (x$inits): -- ## named numeric(0) ## -- First part of data (object): -- ## # A tibble: 50 x 2 ##    time  ipre ##   <dbl> <dbl> ## 1 0     10    ## 2 0.490  7.83 ## 3 0.980  6.13 ## 4 1.47   4.80 ## 5 1.96   3.75 ## 6 2.45   2.94 ## # i 44 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-model-types.html","id":"solved-compartment-models","dir":"Articles","previous_headings":"","what":"Solved compartment models","title":"rxode2 additional model types","text":"Solved models also simple create. simply place linCmt() psuedo-function code. linCmt() function figures type model use based parameter names specified. often, pharmacometric models parameterized terms volume clearances. Clearances specified NONMEM-style names CL, Q, Q1, Q2, etc. distributional clearances CLD, CLD2. Volumes specified Central (VC V), Peripheral/Tissue (VP, VT). translations available, example translations : Another popular parameterization terms micro-constants. rxode2 assumes compartment 1 central compartment. elimination constant specified K, Ke Kel. example translations : last parameterization possible using alpha V //B/C. example translations : linCmt() sleuthing complete, 1, 2 3 compartment model solution used value linCmt(). compartments can dose linear solved system depot central linear absorption constant model ka. Without additional ODEs, compartments numbered depot=1 central=2. absorption constant ka missing, may dose central compartment. Without additional ODEs compartment number central=1. compartments take sort events ODE model can take, discussed rxode2 events vignette. acts ODE model; specify dose depot compartment solve system:","code":"mod <- function() {   ini({     kel <- 0.5     V <- 1   })   model({     ipre <- linCmt(V, kel)   }) } et  <- et(amt=10,time=0,cmt=depot) %>%   et(seq(0,24,length.out=50)) cmt1 <- rxSolve(mod,et) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ cmt1 ## -- Solved rxode2 object -- ## -- Parameters (x$params): -- ## kel   V  ## 0.5 1.0  ## -- Initial Conditions (x$inits): -- ## named numeric(0) ## -- First part of data (object): -- ## # A tibble: 50 x 2 ##    time  ipre ##   <dbl> <dbl> ## 1 0     10    ## 2 0.490  7.83 ## 3 0.980  6.13 ## 4 1.47   4.80 ## 5 1.96   3.75 ## 6 2.45   2.94 ## # i 44 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-model-types.html","id":"mixing-solved-systems-and-odes","dir":"Articles","previous_headings":"","what":"Mixing Solved Systems and ODEs","title":"rxode2 additional model types","text":"addition pure ODEs, may mix solved systems ODEs. prior 2-compartment indirect response model can simplified linCmt() function: allows indirect response model assign 2-compartment model C2 variable used indirect response model. mixing solved systems ODEs, solved system’s compartment always last compartment. solved system technically isn’t compartment solved. Adding dosing compartment end interfere actual ODE solved. Therefore,two-compartment indirect response model, effect compartment compartment #1 PK dosing compartment depot compartment #2. compartment model requires new event table since compartment number changed: can solved following command: Note solving require specifying effect compartment initial condition 1. Rather, already pre-specified eff(0)=1. can solved different initial conditions easily: rxode2 detective also require specify variables linCmt() function already defined block. Therefore, following function also work solve system. Note specify parameters solving system since built model, can override parameters:","code":"library(rxode2) ## Setup example model mod1 <-function() {   model({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) =-KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  =                    Q*C2 - Q*C3     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff   }) }  ## Seup parameters and initial conditions  theta <-     c(KA=2.94E-01, CL=1.86E+01, V2=4.02E+01, # central       Q=1.05E+01,  V3=2.97E+02,              # peripheral       Kin=1, Kout=1, EC50=200)               # effects  inits <- c(eff=1)  ## Setup dosing event information ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=10000, addl=9, ii=12) %>%     et(amt=20000, addl=4, time=120, ii=24) %>%     add.sampling(0:240)  ## Setup a mixed solved/ode system: mod2 <- function() {   model({     ## the order of variables do not matter, the type of compartmental     ## model is determined by the parameters specified.     C2   = linCmt(KA, CL, V2, Q, V3);     eff(0) = 1  ## This specifies that the effect compartment starts at 1.     d/dt(eff) =  Kin - Kout*(1-C2/(EC50+C2))*eff;   }) } ev <- et(amountUnits='mg', timeUnits='hours') %>%   et(amt=10000, addl=9, ii=12, cmt=2) %>%   et(amt=20000, addl=4, time=120, ii=24, cmt=2) %>%   et(0:240) x <- mod2 %>%  solve(theta, ev) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##      KA      CL      V2       Q      V3     Kin    Kout    EC50  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  ## -- Initial Conditions ($inits): -- ## eff  ##   1  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   1    ## 2    1 121.   1.35 ## 3    2  60.3  1.38 ## 4    3  31.0  1.28 ## 5    4  17.0  1.18 ## 6    5  10.2  1.11 ## # i 235 more rows x <- mod2 %>%  solve(theta, ev,c(eff=2)) print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##      KA      CL      V2       Q      V3     Kin    Kout    EC50  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  ## -- Initial Conditions ($inits): -- ## eff  ##   2  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   2    ## 2    1 121.   1.93 ## 3    2  60.3  1.67 ## 4    3  31.0  1.41 ## 5    4  17.0  1.23 ## 6    5  10.2  1.13 ## # i 235 more rows mod3 <- function() {   ini({     KA <- 2.94E-01     CL <- 1.86E+01     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200   })   model({     # Since the parameters are in the ini block, put them in linCmt so     # that the model is detected correctly     C2   <- linCmt(KA, CL, V2, Q, V3)     eff(0) <- 1  ## This specifies that the effect compartment starts at 1.     d/dt(eff) <-  Kin - Kout*(1-C2/(EC50+C2))*eff;   }) }  x <- mod3 %>%  solve(ev)  print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##      KA      CL      V2       Q      V3     Kin    Kout    EC50  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  ## -- Initial Conditions ($inits): -- ## eff  ##   1  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   1    ## 2    1 121.   1.35 ## 3    2  60.3  1.38 ## 4    3  31.0  1.28 ## 5    4  17.0  1.18 ## 6    5  10.2  1.11 ## # i 235 more rows x <- mod3 %>%  solve(c(KA=10),ev) print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##    KA    CL    V2     Q    V3   Kin  Kout  EC50  ##  10.0  18.6  40.2  10.5 297.0   1.0   1.0 200.0  ## -- Initial Conditions ($inits): -- ## eff  ##   1  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   1    ## 2    1 121.   1.35 ## 3    2  60.3  1.38 ## 4    3  31.0  1.28 ## 5    4  17.0  1.18 ## 6    5  10.2  1.11 ## # i 235 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"nesting-in-rxode2","dir":"Articles","previous_headings":"","what":"Nesting in rxode2","title":"Nesting levels in rxode2","text":"one level nesting possible rxode2; example using following uncertainties sources variability:","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"event-table","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Event table","title":"Nesting levels in rxode2","text":"event table contains nesting variables: inv: investigator id id: subject id eye: eye id (left right) occ: occasion","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(seq(0, 240, by=4)) %>% # Assumes sampling when there is no dosing information   et(seq(0, 240, by=4) + 0.1) %>% ## adds 0.1 for separate eye   et(id=1:20) %>%   ## Add an occasion per dose   mutate(occ=cumsum(!is.na(amt))) %>%   mutate(occ=ifelse(occ == 0, 1, occ)) %>%   mutate(occ=2- occ %% 2) %>%   mutate(eye=ifelse(round(time) == time, 1, 2)) %>%   mutate(inv=ifelse(id < 10, 1, 2)) %>% as_tibble ->   ev"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"rxode2-model","dir":"Articles","previous_headings":"Nesting in rxode2","what":"rxode2 model","title":"Nesting levels in rxode2","text":"creates rxode2 model multi-level nesting. Note variables inv.Cl, inv.Ka, eta.Cl etc; need one variable level nesting.","code":"mod <- rxode2({   ## Clearance with individuals   eff(0) = 1   C2 = centr/V2*(1+prop.sd)   C3 = peri/V3   CL =  TCl*exp(eta.Cl + eye.Cl + iov.Cl + inv.Cl)   KA = TKA * exp(eta.Ka + eye.Ka + iov.Cl + inv.Ka)   d/dt(depot) =-KA*depot   d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3   d/dt(peri)  =                    Q*C2 - Q*C3   d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff   ef0 = eff + add.sd }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"uncertainty-in-model-parameters","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Uncertainty in Model parameters","title":"Nesting levels in rxode2","text":"","code":"theta <- c(\"TKA\"=0.294, \"TCl\"=18.6, \"V2\"=40.2,            \"Q\"=10.5, \"V3\"=297, \"Kin\"=1, \"Kout\"=1, \"EC50\"=200)  ## Creating covariance matrix tmp <- matrix(rnorm(8^2), 8, 8) tMat <- tcrossprod(tmp, tmp) / (8 ^ 2) dimnames(tMat) <- list(names(theta), names(theta))  tMat #>               TKA           TCl           V2            Q           V3 #> TKA   0.173571236 -0.1003204607  0.038185010 -0.004108928 -0.095032973 #> TCl  -0.100320461  0.2195710868 -0.043849095  0.013295549 -0.007477895 #> V2    0.038185010 -0.0438490948  0.129784612 -0.017270432 -0.038004762 #> Q    -0.004108928  0.0132955493 -0.017270432  0.022145634  0.020376451 #> V3   -0.095032973 -0.0074778948 -0.038004762  0.020376451  0.165568340 #> Kin  -0.040867119 -0.0492597458 -0.003056722 -0.033468634 -0.003021883 #> Kout  0.035469225  0.0275087955  0.033725901  0.027668205  0.005497301 #> EC50  0.026158042  0.0009434711  0.039426946 -0.036283167 -0.093134292 #>               Kin         Kout          EC50 #> TKA  -0.040867119  0.035469225  0.0261580416 #> TCl  -0.049259746  0.027508796  0.0009434711 #> V2   -0.003056722  0.033725901  0.0394269465 #> Q    -0.033468634  0.027668205 -0.0362831667 #> V3   -0.003021883  0.005497301 -0.0931342922 #> Kin   0.226735493 -0.083447793  0.0659884544 #> Kout -0.083447793  0.117195570 -0.0291684598 #> EC50  0.065988454 -0.029168460  0.0928611407"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"nesting-variability","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Nesting Variability","title":"Nesting levels in rxode2","text":"specify multiple levels nesting, can specify nested lotri matrix; using approach use condition operator | specify variable nesting occurs ; Bayesian simulation need specify much information parameter; rxode2 nu parameter. case: - id, nu=100 model came 100 subjects - eye, nu=200 model came 200 eyes - occ, nu=200 model came 200 occasions - inv, nu=10 model came 10 investigators specify lotri can use | var(nu=X), :","code":"omega <- lotri(lotri(eta.Cl ~ 0.1,                      eta.Ka ~ 0.1) | id(nu=100),                lotri(eye.Cl ~ 0.05,                      eye.Ka ~ 0.05) | eye(nu=200),                lotri(iov.Cl ~ 0.01,                      iov.Ka ~ 0.01) | occ(nu=200),                lotri(inv.Cl ~ 0.02,                      inv.Ka ~ 0.02) | inv(nu=10)) omega #> $id #>        eta.Cl eta.Ka #> eta.Cl    0.1    0.0 #> eta.Ka    0.0    0.1 #>  #> $eye #>        eye.Cl eye.Ka #> eye.Cl   0.05   0.00 #> eye.Ka   0.00   0.05 #>  #> $occ #>        iov.Cl iov.Ka #> iov.Cl   0.01   0.00 #> iov.Ka   0.00   0.01 #>  #> $inv #>        inv.Cl inv.Ka #> inv.Cl   0.02   0.00 #> inv.Ka   0.00   0.02 #>  #> Properties: nu"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"unexplained-variability","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Unexplained variability","title":"Nesting levels in rxode2","text":"last piece variability specify unexplained variability","code":"sigma <- lotri(prop.sd ~ .25,                add.sd~ 0.125)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"solving-the-problem","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Solving the problem","title":"Nesting levels in rxode2","text":"multiple investigators study; investigator number individuals enrolled site. rxode2 automatically determines number investigators simulate effect investigator. output, inv.Cl(inv==1) inv.Cl investigator 1, inv.Cl(inv==2) inv.Cl investigator 2, etc. inv.Cl(inv==1), inv.Cl(inv==2), etc simulated study combined form investigator variability. equation form represent following: look simulated parameters can see inv.Cl(inv==1) inv.Cl(inv==2) s$params; study: eye variability occasion variability individual simulates number variables become eye occasion variability; case eye: look simulation variables (ie eye.Cl(eye==1), eye.Cl(eye==2), etc) change individual combined make eye variability occasion variability can seen pharamcometric models.","code":"s <- rxSolve(mod, theta, ev,              thetaMat=tMat, omega=omega,              sigma=sigma, sigmaDf=400,              nStud=400) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> unhandled error message: EE:[lsoda] 70000 steps taken before reaching tout #>  @(lsoda.c:754 #> Warning: some ID(s) could not solve the ODEs correctly; These values are #> replaced with 'NA' print(s) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #> # A tibble: 8,000 x 24 #>    sim.id id    `inv.Cl(inv==1)` `inv.Cl(inv==2)` `inv.Ka(inv==1)` #>     <int> <fct>            <dbl>            <dbl>            <dbl> #>  1      1 1              0.00900           -0.117            0.312 #>  2      1 2              0.00900           -0.117            0.312 #>  3      1 3              0.00900           -0.117            0.312 #>  4      1 4              0.00900           -0.117            0.312 #>  5      1 5              0.00900           -0.117            0.312 #>  6      1 6              0.00900           -0.117            0.312 #>  7      1 7              0.00900           -0.117            0.312 #>  8      1 8              0.00900           -0.117            0.312 #>  9      1 9              0.00900           -0.117            0.312 #> 10      1 10             0.00900           -0.117            0.312 #> # i 7,990 more rows #> # i 19 more variables: `inv.Ka(inv==2)` <dbl>, `eye.Cl(eye==1)` <dbl>, #> #   `eye.Cl(eye==2)` <dbl>, `eye.Ka(eye==1)` <dbl>, `eye.Ka(eye==2)` <dbl>, #> #   `iov.Cl(occ==1)` <dbl>, `iov.Cl(occ==2)` <dbl>, `iov.Ka(occ==1)` <dbl>, #> #   `iov.Ka(occ==2)` <dbl>, V2 <dbl>, V3 <dbl>, TCl <dbl>, eta.Cl <dbl>, #> #   TKA <dbl>, eta.Ka <dbl>, Q <dbl>, Kin <dbl>, Kout <dbl>, EC50 <dbl> #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #>  #> Simulation with uncertainty in: #> * parameters ($thetaMat for changes) #> * omega matrix ($omegaList) #>  #> -- First part of data (object): -- #> # A tibble: 976,000 x 21 #>   sim.id    id time  inv.Cl inv.Ka  eye.Cl  eye.Ka iov.Cl iov.Ka    C2     C3 #>    <int> <int>  [h]   <dbl>  <dbl>   <dbl>   <dbl>  <dbl>  <dbl> <dbl>  <dbl> #> 1      1     1  0   0.00900  0.312 -0.330  -0.0547 0.0370  0.139  0    0      #> 2      1     1  0.1 0.00900  0.312  0.0792  0.185  0.0370  0.139  8.04 0.0108 #> 3      1     1  4   0.00900  0.312 -0.330  -0.0547 0.0370  0.139 55.4  5.58   #> 4      1     1  4.1 0.00900  0.312  0.0792  0.185  0.0370  0.139 23.2  5.69   #> 5      1     1  8   0.00900  0.312 -0.330  -0.0547 0.0370  0.139 19.6  7.95   #> 6      1     1  8.1 0.00900  0.312  0.0792  0.185  0.0370  0.139 20.0  7.97   #> # i 975,994 more rows #> # i 10 more variables: CL <dbl>, KA <dbl>, ef0 <dbl>, depot <dbl>, centr <dbl>, #> #   peri <dbl>, eff <dbl>, occ <fct>, eye <fct>, inv <fct> inv.Cl = (inv == 1) * `inv.Cl(inv==1)` + (inv == 2) * `inv.Cl(inv==2)` print(head(s$params)) #>   sim.id id inv.Cl(inv==1) inv.Cl(inv==2) inv.Ka(inv==1) inv.Ka(inv==2) #> 1      1  1    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 2      1  2    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 3      1  3    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 4      1  4    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 5      1  5    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 6      1  6    0.008995575     -0.1165083      0.3118205   -0.001926179 #>   eye.Cl(eye==1) eye.Cl(eye==2) eye.Ka(eye==1) eye.Ka(eye==2) iov.Cl(occ==1) #> 1    -0.33023666    0.079190835    -0.05471051     0.18487157    0.037021761 #> 2    -0.42211135   -0.006773841    -0.15399939     0.22463048   -0.007762485 #> 3    -0.05128596    0.123504458     0.18243012     0.13361259   -0.168921159 #> 4    -0.06360621    0.340443016     0.65606292     0.04807945    0.157027939 #> 5     0.08009126   -0.027627640     0.06662396    -0.05838949   -0.226341946 #> 6    -0.11937190    0.090540884    -0.17320340     0.02039840   -0.211648032 #>   iov.Cl(occ==2) iov.Ka(occ==1) iov.Ka(occ==2)       V2       V3      TCl #> 1    -0.19726803     0.13932112    0.088763868 40.26476 296.4748 19.25043 #> 2     0.04555116     0.07716277    0.049148937 40.26476 296.4748 19.25043 #> 3     0.17916513     0.08347085   -0.002990168 40.26476 296.4748 19.25043 #> 4    -0.13444417     0.24736740   -0.077647794 40.26476 296.4748 19.25043 #> 5     0.02049190    -0.09568733    0.054658457 40.26476 296.4748 19.25043 #> 6     0.12619350    -0.01547841    0.040206608 40.26476 296.4748 19.25043 #>        eta.Cl       TKA      eta.Ka        Q       Kin     Kout     EC50 #> 1  0.09788109 0.2354877 -0.21526790 10.43157 0.9256189 1.044937 200.3679 #> 2  0.33302598 0.2354877 -0.01028112 10.43157 0.9256189 1.044937 200.3679 #> 3  0.08504325 0.2354877  0.37253705 10.43157 0.9256189 1.044937 200.3679 #> 4  0.16521678 0.2354877 -0.01859974 10.43157 0.9256189 1.044937 200.3679 #> 5  0.53131380 0.2354877 -0.37109644 10.43157 0.9256189 1.044937 200.3679 #> 6 -0.30786713 0.2354877  0.14897588 10.43157 0.9256189 1.044937 200.3679 print(head(s$params %>% filter(sim.id == 2))) #>   sim.id id inv.Cl(inv==1) inv.Cl(inv==2) inv.Ka(inv==1) inv.Ka(inv==2) #> 1      2  1    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 2      2  2    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 3      2  3    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 4      2  4    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 5      2  5    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 6      2  6    -0.07227613    -0.01884171     0.07620864    -0.01962499 #>   eye.Cl(eye==1) eye.Cl(eye==2) eye.Ka(eye==1) eye.Ka(eye==2) iov.Cl(occ==1) #> 1    -0.25102634      0.1354374   -0.135230741    -0.01605664    0.081212198 #> 2    -0.24832622      0.2233348   -0.380770273    -0.37101991    0.207985292 #> 3     0.24081384     -0.3826770    0.133361609    -0.32496203   -0.099983807 #> 4    -0.11021791      0.3655308    0.009423897     0.08736599   -0.002580804 #> 5     0.21155058      0.3626918    0.194977611    -0.19452347    0.058392202 #> 6     0.03663028     -0.1003754    0.068047203     0.32368023    0.167658084 #>   iov.Cl(occ==2) iov.Ka(occ==1) iov.Ka(occ==2)       V2       V3      TCl #> 1    -0.06556217   -0.032678168   -0.061277262 40.21091 296.6424 18.86433 #> 2    -0.05016896    0.130258486   -0.103721118 40.21091 296.6424 18.86433 #> 3     0.06144127   -0.262726137   -0.104186144 40.21091 296.6424 18.86433 #> 4     0.03954265   -0.037289852   -0.021344468 40.21091 296.6424 18.86433 #> 5     0.12713682   -0.007530921    0.081238197 40.21091 296.6424 18.86433 #> 6    -0.02811353    0.026801132   -0.002702762 40.21091 296.6424 18.86433 #>        eta.Cl       TKA      eta.Ka        Q       Kin     Kout   EC50 #> 1  0.19368101 0.5686675  0.26184170 10.54673 0.5434136 1.323825 200.09 #> 2 -0.16262369 0.5686675  0.12969487 10.54673 0.5434136 1.323825 200.09 #> 3  0.22164455 0.5686675  0.24501105 10.54673 0.5434136 1.323825 200.09 #> 4  0.05997662 0.5686675  0.03059399 10.54673 0.5434136 1.323825 200.09 #> 5  0.45691421 0.5686675 -0.20098716 10.54673 0.5434136 1.323825 200.09 #> 6 -0.19943300 0.5686675 -0.36846082 10.54673 0.5434136 1.323825 200.09 eye.Cl = (eye == 1) * `eye.Cl(eye==1)` + (eye == 2) * `eye.Cl(eye==2)`"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"setting-up-the-rxode2-model-for-the-pipeline","dir":"Articles","previous_headings":"","what":"Setting up the rxode2 model for the pipeline","title":"rxode2 in a pipeline","text":"example show use rxode2 simple pipeline. can start model can used different simulation workflows rxode2 can handle: tumor growth model described Ribba 2012. case, compiled model R object Ribba2012, though rxode2 simulation pipeline, assign compiled model object, though think makes sense.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` Ribba2012 <- function() {   ini({     k = 100     tkde = 0.24     eta.tkde = 0     tkpq = 0.0295     eta.kpq = 0     tkqpp = 0.0031     eta.kqpp = 0     tlambdap = 0.121     eta.lambdap = 0     tgamma = 0.729     eta.gamma = 0     tdeltaqp = 0.00867     eta.deltaqp = 0     prop.sd <- 0     tpt0 = 7.13     eta.pt0 = 0     tq0 = 41.2     eta.q0 = 0   })   model({     kde ~ tkde*exp(eta.tkde)     kpq ~ tkpq * exp(eta.kpq)     kqpp ~ tkqpp * exp(eta.kqpp)     lambdap ~ tlambdap*exp(eta.lambdap)     gamma ~ tgamma*exp(eta.gamma)     deltaqp ~ tdeltaqp*exp(eta.deltaqp)     d/dt(c) = -kde * c     d/dt(pt) = lambdap * pt *(1-pstar/k) + kqpp*qp -       kpq*pt - gamma*c*kde*pt     d/dt(q) = kpq*pt -gamma*c*kde*q     d/dt(qp) = gamma*c*kde*q - kqpp*qp - deltaqp*qp     ## initial conditions     pt0 ~ tpt0*exp(eta.pt0)     q0 ~ tq0*exp(eta.q0)     pt(0) = pt0     q(0) = q0     pstar <- (pt+q+qp)     pstar ~  prop(prop.sd)   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulating-one-event-table","dir":"Articles","previous_headings":"","what":"Simulating one event table","title":"rxode2 in a pipeline","text":"Simulating single event table quite simple: pipe rxode2 simulation object event table object et(). events completely specified, simply solve ODE system rxSolve(). case can pipe output plot() conveniently view results. pt (Proliferative Tissue), q (quiescent tissue) qp (DNA-Damaged quiescent tissue) pstar (total tumor tissue)","code":"Ribba2012 %>% # Use rxode2     et(time.units=\"months\") %>% # Pipe to a new event table     et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months     et(0, 250, by=0.5) %>% # Add some sampling times (not required)     rxSolve() %>% # Solve the simulation     plot(pt, q, qp, pstar) # Plot it, plotting the variables of interest #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulating-with-between-subject-variability","dir":"Articles","previous_headings":"Simulating multiple subjects from a single event table","what":"Simulating with between subject variability","title":"rxode2 in a pipeline","text":"next sort simulation may useful simulating multiple patients treatments. case, use omega matrix specified paper: information, easy simulate 3 subjects model-based parameters:  Note two different things added simulation: - nSub specify many subjects model - omega specify subject variability.","code":"## Add CVs from paper for individual simulation ## Uses exact formula:  lognCv = function(x){log((x/100)^2+1)}  library(lotri) ## Now create omega matrix ## I'm using lotri to quickly specify names/diagonals omega <- lotri(eta.pt0 ~ lognCv(94),                eta.q0 ~ lognCv(54),                eta.lambdap ~ lognCv(72),                eta.kqp ~ lognCv(76),                eta.kqpp ~ lognCv(97),                eta.deltaqp ~ lognCv(115),                eta.tkde ~ lognCv(70))  omega #>               eta.pt0    eta.q0 eta.lambdap   eta.kqp  eta.kqpp eta.deltaqp #> eta.pt0     0.6331848 0.0000000   0.0000000 0.0000000 0.0000000   0.0000000 #> eta.q0      0.0000000 0.2558818   0.0000000 0.0000000 0.0000000   0.0000000 #> eta.lambdap 0.0000000 0.0000000   0.4176571 0.0000000 0.0000000   0.0000000 #> eta.kqp     0.0000000 0.0000000   0.0000000 0.4559047 0.0000000   0.0000000 #> eta.kqpp    0.0000000 0.0000000   0.0000000 0.0000000 0.6631518   0.0000000 #> eta.deltaqp 0.0000000 0.0000000   0.0000000 0.0000000 0.0000000   0.8426442 #> eta.tkde    0.0000000 0.0000000   0.0000000 0.0000000 0.0000000   0.0000000 #>              eta.tkde #> eta.pt0     0.0000000 #> eta.q0      0.0000000 #> eta.lambdap 0.0000000 #> eta.kqp     0.0000000 #> eta.kqpp    0.0000000 #> eta.deltaqp 0.0000000 #> eta.tkde    0.3987761 set.seed(1089) rxSetSeed(1089) Ribba2012 %>% # Use rxode2     et(time.units=\"months\") %>% # Pipe to a new event table     et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months     et(0, 250, by=0.5) %>% # Add some sampling times (not required)     rxSolve(nSub=3, omega=omega) %>% # Solve the simulation     plot(pt, q, qp, pstar) # Plot it, plotting the variables of interest #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> Warning: multi-subject simulation without without 'omega'"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulation-with-unexplained-variability","dir":"Articles","previous_headings":"Simulating multiple subjects from a single event table","what":"Simulation with unexplained variability","title":"rxode2 in a pipeline","text":"can even add unexplained variability quite easily:  case added sigma matrix unexplained variability pstar total tumor tissue. can even simulate uncertainty theta omega sigma values wish.","code":"Ribba2012 %>% # Use rxode2   ini(prop.sd=0.05) %>% # change variability   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=3, omega=omega) %>%   plot(pt, q, qp, sim) # Plot it, plotting the variables of interest #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega' # note that sim is the simulated pstar since this is simulated from the # model with a nlmixr2 endpoint"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulation-with-uncertainty-in-all-the-parameters-by-matrices","dir":"Articles","previous_headings":"Simulating multiple subjects from a single event table","what":"Simulation with uncertainty in all the parameters (by matrices)","title":"rxode2 in a pipeline","text":"assume parameters came 95 subjects 8 observations apiece, degrees freedom omega matrix 95, degrees freedom sigma matrix 95*8=760 95 items informed omega matrix, 760 items informed sigma matrix.  Often simulations full covariance matrix fixed effect parameters. case, matrix, specified thetaMat. full covariance matrix, can information diagonal elements covariance matrix model paper. can converted follows: Now thetaMat represent uncertainty theta matrix, well pieces simulation. Typically can put information simulation thetaMat matrix. large variability theta easy sample negative rate constant, make sense. example: correct problems simply need use truncated multivariate normal specify reasonable ranges parameters. theta specified thetaLower thetaUpper. Similar parameters matrices: omegaLower, omegaUpper, sigmaLower sigmaUpper. may named vectors, one numeric value, numeric vector matching number parameters specified thetaMat matrix. case simulation simply modified thetaLower=0 make sure rates positive:","code":"Ribba2012 %>% # Use rxode2   ini(prop.sd = 0.05) %>%   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=3, nStud=3, omega=omega,           dfSub=760, dfObs=95) %>% # Solve the simulation   plot(pt, q, qp, sim) # Plot it, plotting the variables of interest #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega' rseVar <- function(est, rse){   return(est*rse/100)^2 }  thetaMat <- lotri(tpt0 ~ rseVar(7.13,25),                   tq0 ~ rseVar(41.2,7),                   tlambdap ~ rseVar(0.121, 16),                   tkqpp ~ rseVar(0.0031, 35),                   tdeltaqp ~ rseVar(0.00867, 21),                   tgamma ~ rseVar(0.729, 37),                   tkde ~ rseVar(0.24, 33)                   )  thetaMat #>            tpt0   tq0 tlambdap    tkqpp  tdeltaqp  tgamma   tkde #> tpt0     1.7825 0.000  0.00000 0.000000 0.0000000 0.00000 0.0000 #> tq0      0.0000 2.884  0.00000 0.000000 0.0000000 0.00000 0.0000 #> tlambdap 0.0000 0.000  0.01936 0.000000 0.0000000 0.00000 0.0000 #> tkqpp    0.0000 0.000  0.00000 0.001085 0.0000000 0.00000 0.0000 #> tdeltaqp 0.0000 0.000  0.00000 0.000000 0.0018207 0.00000 0.0000 #> tgamma   0.0000 0.000  0.00000 0.000000 0.0000000 0.26973 0.0000 #> tkde     0.0000 0.000  0.00000 0.000000 0.0000000 0.00000 0.0792 Ribba2012 %>% # Use rxode2 ini(prop.sd = 0.05) %>% et(time.units=\"months\") %>% # Pipe to a new event table et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months et(0, 250, by=0.5) %>% # Add some sampling times (not required) rxSolve(nSub=2, nStud=2, omega=omega, thetaMat=thetaMat, dfSub=760, dfObs=95) %>% # Solve the simulation plot(pt, q, qp, pstar) # Plot it, plotting the variables of interest  #> ℹ change initial estimate of `prop.sd` to `0.05` #> unhandled error message: EE:[lsoda] 70000 steps taken before reaching tout #> @(lsoda.c:750 #> Warning message: #> In rxSolve_(object, .ctl, .nms, .xtra, params, events, inits, setupOnly = .setupOnly) : #>  Some ID(s) could not solve the ODEs correctly; These values are replaced with NA. Ribba2012 %>% # Use rxode2   ini(prop.sd = 0.05) %>%   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=2, nStud=2, omega=omega,           thetaMat=thetaMat,           thetaLower=0, # Make sure the rates are reasonable           dfSub=760, dfObs=95) %>% # Solve the simulation   plot(pt, q, qp, sim) # Plot it, plotting the variables of interest #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega'"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"summarizing-the-simulation-output","dir":"Articles","previous_headings":"","what":"Summarizing the simulation output","title":"rxode2 in a pipeline","text":"easy use dplyr data.table perform summary simulations, rxode2 also provides ability confint function.","code":"## This takes a little more time; Most of the time is the summary ## time.  sim0 <- Ribba2012 %>% # Use rxode2   ini(prop.sd=0.05) %>%   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=10, nStud=10, omega=omega,           thetaMat=thetaMat,           thetaLower=0, # Make sure the rates are reasonable           dfSub=760, dfObs=95) %>% # Solve the simulation   confint(c(\"pt\",\"q\",\"qp\",\"sim\"),level=0.90); # Create Simulation intervals #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega' #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done sim0 %>% plot() # Plot the simulation intervals"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulating-from-a-data-frame-of-parameters","dir":"Articles","previous_headings":"Summarizing the simulation output","what":"Simulating from a data-frame of parameters","title":"rxode2 in a pipeline","text":"simulation matrices can useful fast way simulate information, sometimes may want simulate complex scenarios. instance, may reason believe tkde needs tlambdap, therefore need simulated carefully. can generate data frame whatever way want. internal method simulating new parameters exported .","code":"library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union Ribba2012 <- Ribba2012() # Convert to classic rxode2 model with ini attached  r <- Ribba2012$simulationIniModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ pars <- rxInits(r) pars <- pars[regexpr(\"(prop|eta)\",names(pars)) == -1] print(pars) #>           k        tkde        tkpq       tkqpp    tlambdap      tgamma  #>    1.00e+02    2.40e-01    2.95e-02    3.10e-03    1.21e-01    7.29e-01  #>    tdeltaqp        tpt0         tq0 rxerr.pstar  #>    8.67e-03    7.13e+00    4.12e+01    1.00e+00 ## This is the exported method for simulation of Theta/Omega internally in rxode2 df <- rxSimThetaOmega(params=pars, omega=omega,dfSub=760,                       thetaMat=thetaMat, thetaLower=0, nSub=60,nStud=60) %>%     filter(tkde > tlambdap) %>% as_tibble() ## You could also simulate more and bind them together to a data frame. print(df) #> # A tibble: 2,040 × 17 #>        k  tkde   tkpq tkqpp tlambdap tgamma tdeltaqp  tpt0   tq0 rxerr.pstar #>    <dbl> <dbl>  <dbl> <dbl>    <dbl>  <dbl>    <dbl> <dbl> <dbl>       <dbl> #>  1   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  2   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  3   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  4   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  5   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  6   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  7   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  8   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  9   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #> 10   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #> # ℹ 2,030 more rows #> # ℹ 7 more variables: eta.pt0 <dbl>, eta.q0 <dbl>, eta.lambdap <dbl>, #> #   eta.kqp <dbl>, eta.kqpp <dbl>, eta.deltaqp <dbl>, eta.tkde <dbl> ## Quick check to make sure that all the parameters are OK. all(df$tkde>df$tlambdap) #> [1] TRUE sim1 <- r %>% # Use rxode2     et(time.units=\"months\") %>% # Pipe to a new event table     et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months     et(0, 250, by=0.5) %>% # Add some sampling times (not required)     rxSolve(df) ## Note this information looses information about which ID is in a ## \"study\", so it summarizes the confidence intervals by dividing the ## subjects into sqrt(#subjects) subjects and then summarizes the ## confidence intervals sim2 <- sim1 %>% confint(c(\"pt\",\"q\",\"qp\",\"sim\"),level=0.90); # Create Simulation intervals #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done save(sim2, file = file.path(system.file(package = \"rxode2\"), \"pipeline-sim2.rds\"), version = 2) sim2 %>% plot()"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"setting-up-model-for-plotting-explanation","dir":"Articles","previous_headings":"Plotting in rxode2","what":"Setting up model for plotting explanation","title":"Plotting rxode2 models","text":"first step explain rxode2 model plots setup: rxode2 model event table solved object","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` ## Model from rxode2 tutorial m1 <- function() {   ini({     KA <- 2.94E-01     CL <- 1.86E+01     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250   })   model({     C2 <- centr / V2     C3 <- peri / V3     d/dt(depot) <- -KA * depot     f(depot) <- fdepot     dur(depot) <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3     d/dt(peri) <- Q * C2 - Q * C3     d/dt(eff) <- Kin - Kout * (1 - C2 / (EC50 + C2)) * eff     eff(0) <- 1   }) }  ev <- et(timeUnits = \"hr\") %>%   et(amt = 10000, ii = 12, until = 24) %>%   et(seq(0, 24, length.out = 100))  s <- rxSolve(m1, ev)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"default-rxode2-plot","dir":"Articles","previous_headings":"Plotting in rxode2","what":"Default rxode2 plot","title":"Plotting rxode2 models","text":"default plot rxode2 plot time versus state calculated lhs values:  Things note: First, plots actually ggplot2 plots can modified standard ggplot2 grammar graphics Also note plot method rxSolve objects can currently use arguments default plot method: log character string contains \"x\" x axis logarithmic, \"y\" y axis logarithmic \"xy\" \"yx\" axes logarithmic. available, use xgxr xgx_scale_x_log10() xgx_scale_y_log10() functions. xlab bit different simply supplying x-label ggplot2, retain unit information originally supplied event table. Also note, available, unit information used used meaningful ticks xgx_scale_x_time_units() ylab supplies y label following currently unsupported: . type b. xlim c. ylim d. main e. sub f. ann g. axes","code":"plot(s)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"subsetting-plots","dir":"Articles","previous_headings":"","what":"Subsetting plots","title":"Plotting rxode2 models","text":"interesting, often useful subset plots based values interested ; example, interesting plot eff C2. , simply specify interesting items plot function. instance:","code":"plot(s, C2, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"semi-log-plots","dir":"Articles","previous_headings":"","what":"Semi-log plots","title":"Plotting rxode2 models","text":"Semi-log plots PK concentrations common; simply need use log=\"y\"  uses xgx_scale_y_log10() available clearly show semi-log nature plot.","code":"plot(s, C2, log=\"y\", ylab=\"Concentration\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"plots-with-multi-subject-plots","dir":"Articles","previous_headings":"","what":"Plots with multi-subject plots","title":"Plotting rxode2 models","text":"multi-subject plots can easily plot applying principles . illustrate plot, lets expand event table include plot 4 subjects lognormal random variability subjects; complete, may plot method:  Notice colored individual labeled legend. interested concentration, produces similar plot:  Notice plot individual labeled attached id let know individual comes . done ggrepel available, otherwise legend retained.","code":"# Setup the new problem m2 <- function() {   ini({     KA <- 2.94E-01     TCL <- 1.86E+01     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250     eta.Cl ~ 0.4^2       })   model({     CL <- TCL * exp(eta.Cl)     C2 <- centr / V2     C3 <- peri / V3     d/dt(depot) <- -KA * depot     f(depot) <- fdepot     dur(depot) <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3     d/dt(peri) <- Q * C2 - Q * C3     d/dt(eff) <- Kin - Kout * (1 - C2 / (EC50 + C2)) * eff     eff(0) <- 1   }) }  # Create the event table ev <- et(timeUnits = \"hr\") %>%   et(amt = 10000, until = units::set_units(3, days), ii = 12) %>% # loading doses   et(seq(0, 48, length.out = 200)) %>%   et(id = 1:4)  s <- rxSolve(m2, ev) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ plot(s, C2, eff) plot(s, C2, log=\"y\", ylab=\"Concentration\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"multi-subject-plots-with-large-number-of-subjects","dir":"Articles","previous_headings":"Plots with multi-subject plots","what":"Multi-subject plots with large number of subjects","title":"Plotting rxode2 models","text":"can get bit much many subjects solved (example lets simulate 100)  case, individuals put plot transparent grayscale plot pane. allows places subjects present darker. number individuals plots switch legend grayscale controlled changing options rxode2.spaghetti. ie. options(rxode2.spaghetti=7). 7 individuals default value plotting changes individual grayscale spaghetti plots. can also create confidence interval simulations confint: plot plot","code":"ev <- et(timeUnits = \"hr\") %>%   et(amt = 10000, until = units::set_units(3, days), ii = 12) %>% # loading doses   et(seq(0, 48, length.out = 200)) %>%   et(id = 1:100) # 100 subjects  s <- rxSolve(m2, ev)  plot(s, C2, log=\"y\", ylab=\"Concentration\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values. s2 <- confint(s, parm=\"C2\") ## ! in order to put confidence bands around the intervals, you need at least 2500 simulations ## summarizing data...done plot(s2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-prior-data.html","id":"using-prior-data-for-solving","dir":"Articles","previous_headings":"","what":"Using prior data for solving","title":"Using Prior Data for ODE solving","text":"rxode2 can use single subject multiple subjects single event table solve ODEs. Additionally, rxode2 can use arbitrary data frame individualized events. example using nlmixr, use theo_sd data frame","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` library(nlmixr2data)  ## Load data from nlmixr d <- theo_sd  ## Create rxode2 model theo <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v   })  }  ## Create parameter dataset library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union parsDf <- tribble(   ~ eta.ka, ~ eta.cl, ~ eta.v,    0.105, -0.487, -0.080,   0.221, 0.144, 0.021,   0.368, 0.031, 0.058,  -0.277, -0.015, -0.007,  -0.046, -0.155, -0.142,  -0.382, 0.367, 0.203,  -0.791, 0.160, 0.047,  -0.181, 0.168, 0.096,   1.420, 0.042, 0.012,  -0.738, -0.391, -0.170,   0.790, 0.281, 0.146,  -0.527, -0.126, -0.198) %>%     mutate(tka = 0.451, tcl = 1.017, tv = 3.449)  ## Now solve the dataset solveData <- rxSolve(theo, parsDf, d) #> ℹ parameter labels from comments will be replaced by 'label()' #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ plot(solveData, cp) print(solveData) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 12 × 7 #>    id      tka   tcl    tv eta.ka eta.cl  eta.v #>    <fct> <dbl> <dbl> <dbl>  <dbl>  <dbl>  <dbl> #>  1 1     0.451  1.02  3.45  0.105 -0.487 -0.08  #>  2 2     0.451  1.02  3.45  0.221  0.144  0.021 #>  3 3     0.451  1.02  3.45  0.368  0.031  0.058 #>  4 4     0.451  1.02  3.45 -0.277 -0.015 -0.007 #>  5 5     0.451  1.02  3.45 -0.046 -0.155 -0.142 #>  6 6     0.451  1.02  3.45 -0.382  0.367  0.203 #>  7 7     0.451  1.02  3.45 -0.791  0.16   0.047 #>  8 8     0.451  1.02  3.45 -0.181  0.168  0.096 #>  9 9     0.451  1.02  3.45  1.42   0.042  0.012 #> 10 10    0.451  1.02  3.45 -0.738 -0.391 -0.17  #> 11 11    0.451  1.02  3.45  0.79   0.281  0.146 #> 12 12    0.451  1.02  3.45 -0.527 -0.126 -0.198 #> ── Initial Conditions ($inits): ── #>  depot center  #>      0      0  #> ── First part of data (object): ── #> # A tibble: 132 × 8 #>      id  time    ka    cl     v    cp   depot center #>   <int> <dbl> <dbl> <dbl> <dbl> <dbl>   <dbl>  <dbl> #> 1     1  0     1.74  1.70  29.0  0    320.        0  #> 2     1  0.25  1.74  1.70  29.0  3.86 207.      112. #> 3     1  0.57  1.74  1.70  29.0  6.81 118.      198. #> 4     1  1.12  1.74  1.70  29.0  9.06  45.4     263. #> 5     1  2.02  1.74  1.70  29.0  9.79   9.45    284. #> 6     1  3.82  1.74  1.70  29.0  9.10   0.410   264. #> # ℹ 126 more rows ## Of course the fasest way to solve if you don't care about the rxode2 extra parameters is  solveData <- rxSolve(theo, parsDf, d, returnType=\"data.frame\") #> ℹ parameter labels from comments will be replaced by 'label()' #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## solved data dplyr::as_tibble(solveData) #> # A tibble: 132 × 8 #>       id  time    ka    cl     v    cp         depot center #>    <int> <dbl> <dbl> <dbl> <dbl> <dbl>         <dbl>  <dbl> #>  1     1  0     1.74  1.70  29.0  0    320.              0  #>  2     1  0.25  1.74  1.70  29.0  3.86 207.            112. #>  3     1  0.57  1.74  1.70  29.0  6.81 118.            198. #>  4     1  1.12  1.74  1.70  29.0  9.06  45.4           263. #>  5     1  2.02  1.74  1.70  29.0  9.79   9.45          284. #>  6     1  3.82  1.74  1.70  29.0  9.10   0.410         264. #>  7     1  5.1   1.74  1.70  29.0  8.46   0.0440        246. #>  8     1  7.03  1.74  1.70  29.0  7.56   0.00152       219. #>  9     1  9.05  1.74  1.70  29.0  6.71   0.0000449     195. #> 10     1 12.1   1.74  1.70  29.0  5.61   0.000000212   163. #> # ℹ 122 more rows data.table::data.table(solveData) #>         id  time        ka       cl        v       cp        depot    center #>      <int> <num>     <num>    <num>    <num>    <num>        <num>     <num> #>   1:     1  0.00 1.7436838 1.698932 29.04946 0.000000 3.199920e+02   0.00000 #>   2:     1  0.25 1.7436838 1.698932 29.04946 3.861730 2.069289e+02 112.18117 #>   3:     1  0.57 1.7436838 1.698932 29.04946 6.805372 1.184389e+02 197.69240 #>   4:     1  1.12 1.7436838 1.698932 29.04946 9.058196 4.539354e+01 263.13572 #>   5:     1  2.02 1.7436838 1.698932 29.04946 9.791088 9.450361e+00 284.42585 #>  ---                                                                         #> 128:    12  5.07 0.9268162 2.437566 25.81614 8.442535 2.919432e+00 217.95370 #> 129:    12  7.07 0.9268162 2.437566 25.81614 7.074251 4.573778e-01 182.62989 #> 130:    12  9.03 0.9268162 2.437566 25.81614 5.892253 7.436222e-02 152.11524 #> 131:    12 12.05 0.9268162 2.437566 25.81614 4.432614 4.526550e-03 114.43300 #> 132:    12 24.15 0.9268162 2.437566 25.81614 1.414179 6.154773e-08  36.50865"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-rxUse.html","id":"using-pre-compiled-models-in-your-packages","dir":"Articles","previous_headings":"","what":"Using Pre-compiled models in your packages","title":"Compiled rxode2 models in Packages","text":"package like include pre-compiled rxode2 models package easy create package. simple make package rxPkg() command. : Add model package; can use package data idr package loads Add right package requirements DESCRIPTION file. want update describe package modify authors, license etc. Create skeleton model documentation files can add package documentation. case file idr-doc.R R directory Create configure configure.win script removes regenerates src directory based whatever version rxode2 compiled . modified plan compiled code, though suggested. can write R code package interacts rxode2 object can distribute shiny apps similar things package context. present can add models package rxUse(). Simply compile rxode2 model package add model rxUse() Now model idr model library. also create model-doc.R R directory can document model. can use devtools methods install/test model","code":"library(rxode2); ## Now Create a model idr <- rxode2({     C2 = centr/V2;     C3 = peri/V3;     d/dt(depot) =-KA*depot;     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  =                    Q*C2 - Q*C3;     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff; })  ## You can specify as many models as you want to add  rxPkg(idr, package=\"myPackage\"); ## Add the idr model to your package rxUse(model) devtools::load_all() # Load all the functions in the package devtools::document() # Create package documentation devtools::install() # Install package devtools::check() # Check the package devtools::build() # build the package so you can submit it to places like CRAN"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-rxUse.html","id":"using-models-in-a-already-present-package","dir":"Articles","previous_headings":"","what":"Using Models in a already present package","title":"Compiled rxode2 models in Packages","text":"illustrate, lets start blank package rxUse() : - Create rxode2 sources move package’s src/ directory. R source package, also finish directory library-init.c registers rxode2 models package use R. - Create stub R documentation models including package. able see R documentation loading package standard ? interface. still need : - Export least one function. function wish export, can add re-export rxode2 using roxygen follows: want use Suggests instead Depends package, way want export rxode2’s normal routines also need instruct R load model library models included model’s dll. done : R package rxode2 models intend add compiled sources (recommended), can add following configure scripts Depending check may need dummy autoconf script, want integrate sources Rcpp C/Fortan based packages, need include rxModels-compiled.h : - Add define macro compiledModelCall list registered .Call functions. - Register C interface allow model solving R_init0_rxModels_rxode2_models() (rxModels replaced package name). complete, can compile/document standard methods: load package new version rxode2, models recompiled used. However, want models recompiled recent version rxode2, simply need call rxUse() project directory followed standard methods install/create package. Note include rxode2 code required generate model regenerate rxode2 c-code src directory. rxode2 objects, summary show one way recreate model. example compiled models package can found rxModels repository.","code":"library(rxode2) library(usethis) pkgPath  <- file.path(rxTempDir(),\"MyRxModel\") create_package(pkgPath); use_gpl3_license(\"Matt\") use_package(\"rxode2\", \"LinkingTo\") use_package(\"rxode2\", \"Depends\") ##  library(rxode2) on load; Can use imports instead. use_roxygen_md() ##use_readme_md() library(rxode2); ## Now Create a model idr <- rxode2({     C2 = centr/V2;     C3 = peri/V3;     d/dt(depot) =-KA*depot;     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  =                    Q*C2 - Q*C3;     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff; });  rxUse(idr); ## Add the idr model to your package rxUse(); # Update the compiled rxode2 sources for all of your packages ##' @importFrom rxode2 rxode2 ##' @export rxode2::rxode2 ##' @importFrom rxode2 rxode2 ##' @export rxode2::rxode2  ##' @importFrom rxode2 et ##' @export rxode2::et  ##' @importFrom rxode2 etRep ##' @export rxode2::etRep  ##' @importFrom rxode2 etSeq ##' @export rxode2::etSeq  ##' @importFrom rxode2 as.et ##' @export rxode2::as.et  ##' @importFrom rxode2 eventTable ##' @export rxode2::eventTable  ##' @importFrom rxode2 add.dosing ##' @export rxode2::add.dosing  ##' @importFrom rxode2 add.sampling ##' @export rxode2::add.sampling  ##' @importFrom rxode2 rxSolve ##' @export rxode2::rxSolve  ##' @importFrom rxode2 rxControl ##' @export rxode2::rxControl  ##' @importFrom rxode2 rxClean ##' @export rxode2::rxClean  ##' @importFrom rxode2 rxUse ##' @export rxode2::rxUse  ##' @importFrom rxode2 rxShiny ##' @export rxode2::rxShiny  ##' @importFrom rxode2 genShinyApp.template ##' @export rxode2::genShinyApp.template  ##' @importFrom rxode2 cvPost ##' @export rxode2::cvPost  # This is actually from `magrittr` but allows less imports ##' @importFrom rxode2 %>% ##' @export rxode2::`%>%` # In this case `rxModels` is the package name ##' @useDynLib rxModels, .registration=TRUE #!/bin/sh # This should be used for both configure and configure.win echo \"unlink('src', recursive=TRUE);rxode2::rxUse()\" > build.R ${R_HOME}/bin/Rscript build.R rm build.R ## dummy autoconf script ## It is saved to configure.ac devtools::load_all() devtools::document() devtools::install() devtools::load_all() devtools::document() devtools::install()"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-shiny.html","id":"facilities-for-generating-r-shiny-applications","dir":"Articles","previous_headings":"","what":"Facilities for generating R shiny applications","title":"rxode2 and Shiny","text":"example creating R shiny application interactively explore responses various complex dosing regimens available http://qsp.engr.uga.edu:3838/rxode2/RegimenSimulator. Shiny applications like one may programmatically created experimental function genShinyApp.template(). application includes widgets varying dose, dosing regimen, dose cycle, number cycles. Click go Shiny App","code":"genShinyApp.template(appDir = \"shinyExample\", verbose=TRUE)  library(shiny) runApp(\"shinyExample\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-shiny.html","id":"exploring-parameter-fits-graphically-using-shiny","dir":"Articles","previous_headings":"","what":"Exploring parameter fits graphically using shiny","title":"rxode2 and Shiny","text":"rxode2 object can explored rxShiny(obj). rxShiny() also allow try new models see behave.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-variability-with-rxode2","dir":"Articles","previous_headings":"Population Simulations with rxode2","what":"Simulation of Variability with rxode2","title":"rxode2 Simulation","text":"pharmacometrics nonlinear-mixed effect modeling software (like nlmixr) characterizes -subject variability. subject variability can simulate new subjects. Assuming 2-compartment, indirect response model, can set create rxode2 model describing system :","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"setting-up-the-rxode2-model","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Setting up the rxode2 model","title":"rxode2 Simulation","text":"","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` set.seed(32) rxSetSeed(32)  mod <- function() {   ini({     KA <- 2.94E-01     TCl <- 1.86E+01     # between subject variability     eta.Cl ~ 0.4^2     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200   })   model({     C2 <- centr/V2     C3 <- peri/V3     CL <-  TCl*exp(eta.Cl) ## This is coded as a variable in the model     d/dt(depot) <- -KA*depot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulating","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Simulating","title":"rxode2 Simulation","text":"next step simulate create dosing regimen overall simulation: wish, can also add sampling times (though rxode2 can fill ): Note et takes similar arguments seq adding sampling times. methods adding sampling times events make complex dosing regimens (See event vignette). includes ways add variability sampling dosing times). complete can simulate using rxSolve routine: quickly look customize simulation use default plot routine. Since rxode2 object, create ggplot2 object can modify wish. extra parameter plot tells rxode2/R piece information interested plotting. case, interested looking derived parameter C2:","code":"ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, cmt=\"centr\") ev <- ev %>% et(0,48, length.out=100) sim  <- rxSolve(mod, ev, nSub=100) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"checking-the-simulation-with-plot","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Checking the simulation with plot","title":"rxode2 Simulation","text":"course additional parameter also state value, like eff:  even look two side--side:  stack patchwork","code":"library(ggplot2) ## The plots from rxode2 are ggplots so they can be modified with ## standard ggplot commands. plot(sim, C2, log=\"y\") +     ylab(\"Central Compartment\") ## They also takes many of the standard plot arguments; See ?plot plot(sim, eff, ylab=\"Effect\") plot(sim, C2, eff) library(patchwork) plot(sim, C2, log=\"y\") / plot(sim, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"processing-the-data-to-create-summary-plots","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Processing the data to create summary plots","title":"rxode2 Simulation","text":"Usually pharmacometric simulations enough simply simulate system. something easier digest, like look central extreme tendencies simulation. Since rxode2 solve object type data frame now straightforward perform calculations generate plots simulated data. can , 5th, 50th, 95th percentiles simulated data plotted.   Note can see parameters simulated example","code":"confint(sim, \"C2\", level=0.95) %>%     plot(ylab=\"Central Concentration\", log=\"y\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done confint(sim, \"eff\", level=0.95) %>%     plot(ylab=\"Effect\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done head(sim$param) #>   sim.id    KA  TCl   V2    Q  V3 Kin Kout EC50     eta.Cl #> 1      1 0.294 18.6 40.2 10.5 297   1    1  200  0.8067570 #> 2      2 0.294 18.6 40.2 10.5 297   1    1  200  0.0236961 #> 3      3 0.294 18.6 40.2 10.5 297   1    1  200 -0.3340697 #> 4      4 0.294 18.6 40.2 10.5 297   1    1  200 -0.1098197 #> 5      5 0.294 18.6 40.2 10.5 297   1    1  200 -0.2803752 #> 6      6 0.294 18.6 40.2 10.5 297   1    1  200 -0.2114090"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-unexplained-variability-sigma","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Simulation of unexplained variability (sigma)","title":"rxode2 Simulation","text":"addition conveniently simulating subject variability, can also easily simulate unexplained variability. One way create rxode2 model endpoints defined. Model piping can keeping model intact: can see dataset defined $multipleEndpoint: see endpoints defined eff C2: allows solve system: Since simulated model residual specification included multiple endpoint model, can summarize endpoint simply taking confidence interval \"sim\":  want subset specific endpoint, like eff can create confidence interval endpoint using specification sim.eff, endpoint name separated sim dot:","code":"mod2 <- mod %>%   model(eff ~ add(eff.sd), append=TRUE) %>%   model(C2 ~ prop(prop.sd), append=TRUE) %>%   ini(eff.sd=sqrt(0.1), prop.sd=sqrt(0.1)) #> ℹ add residual parameter `eff.sd` and set estimate to 1 #> ℹ add residual parameter `prop.sd` and set estimate to 1 #> ℹ change initial estimate of `eff.sd` to `0.316227766016838` #> ℹ change initial estimate of `prop.sd` to `0.316227766016838` mod2$multipleEndpoint #>   variable                cmt                dvid* #> 1  eff ~ … cmt='eff' or cmt=4 dvid='eff' or dvid=1 #> 2   C2 ~ …  cmt='C2' or cmt=5  dvid='C2' or dvid=2 ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, cmt=\"centr\") %>%   et(seq(0,48, length.out=100), cmt=\"eff\") %>%   et(seq(0,48, length.out=100), cmt=\"C2\") sim  <- rxSolve(mod2, ev, nSub=100) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ s <- confint(sim, \"sim\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(s) seff <- confint(sim, \"sim.eff\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(seff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-individuals","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Simulation of Individuals","title":"rxode2 Simulation","text":"Sometimes may want match dosing observations individuals clinical trial. create data.frame using rxode2 event specification well ID column indicate individual. rxode2 event vignette talks datasets created.","code":"library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union ev1 <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=10000, cmt=2) %>%     et(0,48,length.out=10)  ev2 <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=5000, cmt=2) %>%     et(0,48,length.out=8)  dat <- rbind(data.frame(ID=1, ev1$get.EventTable()),              data.frame(ID=2, ev2$get.EventTable()))   ## Note the number of subject is not needed since it is determined by the data sim  <- rxSolve(mod, dat) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ sim %>% select(id, time, eff, C2) #>    id          time      eff          C2 #> 1   1  0.000000 [h] 1.000000 248.7562189 #> 2   1  5.333333 [h] 1.168348  19.1658271 #> 3   1 10.666667 [h] 1.040652   7.1764936 #> 4   1 16.000000 [h] 1.030849   6.0167596 #> 5   1 21.333333 [h] 1.027742   5.4446478 #> 6   1 26.666667 [h] 1.025204   4.9484199 #> 7   1 32.000000 [h] 1.022911   4.4984868 #> 8   1 37.333333 [h] 1.020828   4.0895169 #> 9   1 42.666667 [h] 1.018933   3.7177314 #> 10  1 48.000000 [h] 1.017211   3.3797461 #> 11  2  0.000000 [h] 1.000000 124.3781095 #> 12  2  6.857143 [h] 1.016301   1.9290018 #> 13  2 13.714286 [h] 1.006648   1.2918078 #> 14  2 20.571429 [h] 1.005617   1.0964981 #> 15  2 27.428571 [h] 1.004772   0.9317005 #> 16  2 34.285714 [h] 1.004055   0.7916739 #> 17  2 41.142857 [h] 1.003446   0.6726918 #> 18  2 48.000000 [h] 1.002928   0.5715923"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-clinical-trials","dir":"Articles","previous_headings":"","what":"Simulation of Clinical Trials","title":"rxode2 Simulation","text":"either using simple single event table, data clinical trial described , complete clinical trial simulation can performed. Typically clinical trial simulations want account uncertainty fixed parameter estimates, even uncertainty subject variability well unexplained variability. rxode2 allows account uncertainties simulating multiple virtual “studies,” specified parameter nStud. studies samples realization fixed effect parameters covariance matrices subject variability(omega) unexplained variabilities (sigma). Depending information models, strategies simulating realization omega sigma matrices. first strategy occurs either standard errors standard deviations (related parameters), modeled correlation model simulating . case suggested strategy use inverse Wishart (parameterized scale conjugate prior)/scaled inverse chi distribution. approach uses single parameter inform variability covariance matrix sampled (degrees freedom). second strategy occurs standard errors variance/standard deviation modeled correlations covariance matrix. approach perform separate simulations standard deviations correlation matrix. First simulate variance/standard deviation components thetaMat multivariate normal simulation. simulation transformation standard deviations, correlation matrix simulated using degrees freedom covariance matrix. Combining simulated standard deviation simulated correlation matrix give simulated covariance matrix. smaller dimension covariance matrices (dimension < 10x10) recommended use lkj distribution simulate correlation matrix. higher dimension covariance matrices suggested use inverse wishart distribution (transformed correlation matrix) simulations. covariance/variance prior simulated rxode2s cvPost() function.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-from-inverse-wishart-correlations","dir":"Articles","previous_headings":"Simulation of Clinical Trials","what":"Simulation from inverse Wishart correlations","title":"rxode2 Simulation","text":"example simulation :  wish can see omega sigma used virtual study accessing solved data object $omega.list $sigma.list: can also see parameter realizations $params data frame.","code":"## Creating covariance matrix tmp <- matrix(rnorm(8^2), 8, 8) tMat <- tcrossprod(tmp, tmp) / (8 ^ 2) dimnames(tMat) <- list(NULL, names(mod2$theta)[1:8])  sim  <- rxSolve(mod2, ev, nSub=100, thetaMat=tMat, nStud=10,                 dfSub=10, dfObs=100)  s <-sim %>% confint(\"sim\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(s) head(sim$omegaList) #> [[1]] #>           eta.Cl #> eta.Cl 0.1676778 #>  #> [[2]] #>           eta.Cl #> eta.Cl 0.2917085 #>  #> [[3]] #>           eta.Cl #> eta.Cl 0.1776813 #>  #> [[4]] #>           eta.Cl #> eta.Cl 0.1578682 #>  #> [[5]] #>           eta.Cl #> eta.Cl 0.1845614 #>  #> [[6]] #>           eta.Cl #> eta.Cl 0.3282268 head(sim$sigmaList) #> [[1]] #>            rxerr.eff   rxerr.C2 #> rxerr.eff 1.12416983 0.04197039 #> rxerr.C2  0.04197039 0.97293971 #>  #> [[2]] #>             rxerr.eff    rxerr.C2 #> rxerr.eff  0.84311199 -0.06277998 #> rxerr.C2  -0.06277998  1.22140938 #>  #> [[3]] #>           rxerr.eff  rxerr.C2 #> rxerr.eff 0.9834771 0.1060251 #> rxerr.C2  0.1060251 1.0024751 #>  #> [[4]] #>            rxerr.eff   rxerr.C2 #> rxerr.eff 1.25556975 0.07690868 #> rxerr.C2  0.07690868 0.90991261 #>  #> [[5]] #>           rxerr.eff  rxerr.C2 #> rxerr.eff  1.116261 -0.184748 #> rxerr.C2  -0.184748  1.320288 #>  #> [[6]] #>            rxerr.eff   rxerr.C2 #> rxerr.eff 0.93539238 0.07270049 #> rxerr.C2  0.07270049 0.98648424"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulate-using-variancestandard-deviation-standard-errors","dir":"Articles","previous_headings":"Simulation of Clinical Trials","what":"Simulate using variance/standard deviation standard errors","title":"rxode2 Simulation","text":"Lets assume wish simulate nonmem run included xpose First setup model; Since taking nonmem like use free-form style classic rxode2 model start classic model: Next input estimated parameters: also covariances; , easiest way create named covariance matrix use lotri():","code":"rx1 <- rxode2({   cl <- tcl*(1+crcl.cl*(CLCR-65)) * exp(eta.cl)   v <- tv * WT * exp(eta.v)   ka <- tka * exp(eta.ka)   ipred <- linCmt()   obs <- ipred * (1 + prop.sd) + add.sd  }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ theta <- c(tcl=2.63E+01, tv=1.35E+00, tka=4.20E+00, tlag=2.08E-01,            prop.sd=2.05E-01, add.sd=1.06E-02, crcl.cl=7.17E-03,            ## Note that since we are using the separation strategy the ETA variances are here too            eta.cl=7.30E-02,  eta.v=3.80E-02, eta.ka=1.91E+00) thetaMat <- lotri(     tcl + tv + tka + tlag + prop.sd + add.sd + crcl.cl + eta.cl + eta.v + eta.ka ~         c(7.95E-01,           2.05E-02, 1.92E-03,           7.22E-02, -8.30E-03, 6.55E-01,           -3.45E-03, -6.42E-05, 3.22E-03, 2.47E-04,           8.71E-04, 2.53E-04, -4.71E-03, -5.79E-05, 5.04E-04,           6.30E-04, -3.17E-06, -6.52E-04, -1.53E-05, -3.14E-05, 1.34E-05,           -3.30E-04, 5.46E-06, -3.15E-04, 2.46E-06, 3.15E-06, -1.58E-06, 2.88E-06,           -1.29E-03, -7.97E-05, 1.68E-03, -2.75E-05, -8.26E-05, 1.13E-05, -1.66E-06, 1.58E-04,           -1.23E-03, -1.27E-05, -1.33E-03, -1.47E-05, -1.03E-04, 1.02E-05, 1.67E-06, 6.68E-05, 1.56E-04,           7.69E-02, -7.23E-03, 3.74E-01, 1.79E-03, -2.85E-03, 1.18E-05, -2.54E-04, 1.61E-03, -9.03E-04, 3.12E-01))  evw <- et(amount.units=\"mg\", time.units=\"hours\") %>%     et(amt=100) %>%     ## For this problem we will simulate with sampling windows     et(list(c(0, 0.5),        c(0.5, 1),        c(1, 3),        c(3, 6),        c(6, 12))) %>%     et(id=1:1000)  ## From the run we know that: ##   total number of observations is: 476 ##    Total number of individuals:     74 sim  <- rxSolve(rx1, theta, evw,  nSub=100, nStud=10,                 thetaMat=thetaMat,                 ## Match boundaries of problem                 thetaLower=0,                  sigma=c(\"prop.sd\", \"add.sd\"), ## Sigmas are standard deviations                 sigmaXform=\"identity\", # default sigma xform=\"identity\"                 omega=c(\"eta.cl\", \"eta.v\", \"eta.ka\"), ## etas are variances                 omegaXform=\"variance\", # default omega xform=\"variance\"                 iCov=data.frame(WT=rnorm(1000, 70, 15), CLCR=rnorm(1000, 65, 25)),                 dfSub=74, dfObs=476); #> ℹ thetaMat has too many items, ignored: 'tlag' print(sim) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 10,000 × 9 #>    sim.id id      tcl crcl.cl  eta.cl    tv   eta.v   tka  eta.ka #>     <int> <fct> <dbl>   <dbl>   <dbl> <dbl>   <dbl> <dbl>   <dbl> #>  1      1 1      27.0   0.828 -0.884   1.73  0.531   5.38  0.114  #>  2      1 2      27.0   0.828  1.11    1.73 -0.906   5.38 -0.738  #>  3      1 3      27.0   0.828 -0.690   1.73  0.280   5.38 -0.365  #>  4      1 4      27.0   0.828  0.0751  1.73 -1.17    5.38 -0.575  #>  5      1 5      27.0   0.828 -0.299   1.73 -0.121   5.38 -0.601  #>  6      1 6      27.0   0.828 -0.638   1.73 -0.281   5.38 -0.309  #>  7      1 7      27.0   0.828  1.05    1.73 -0.0495  5.38  0.567  #>  8      1 8      27.0   0.828 -0.371   1.73 -0.372   5.38 -0.903  #>  9      1 9      27.0   0.828 -0.206   1.73  1.97    5.38  0.0203 #> 10      1 10     27.0   0.828  0.754   1.73  0.166   5.38 -1.73   #> # ℹ 9,990 more rows #> ── Initial Conditions ($inits): ── #> named numeric(0) #>  #> Simulation with uncertainty in: #> • parameters ($thetaMat for changes) #> • omega matrix ($omegaList) #> • sigma matrix ($sigmaList) #>  #> ── First part of data (object): ── #> # A tibble: 50,000 × 10 #>   sim.id    id   time     cl     v    ka  ipred    obs    WT  CLCR #>    <int> <int>    [h]  <dbl> <dbl> <dbl>  <dbl>  <dbl> <dbl> <dbl> #> 1      1     1 0.0155   50.4 183.   6.03 0.0487  3.56   62.2  69.3 #> 2      1     1 0.749    50.4 183.   6.03 0.460  -1.59   62.2  69.3 #> 3      1     1 1.02     50.4 183.   6.03 0.431   1.15   62.2  69.3 #> 4      1     1 3.41     50.4 183.   6.03 0.224   1.92   62.2  69.3 #> 5      1     1 7.81     50.4 183.   6.03 0.0666  0.731  62.2  69.3 #> 6      1     2 0.0833 2816.   24.9  2.57 0.0754 -1.20   35.7 105.  #> # ℹ 49,994 more rows ## Notice that the simulation time-points change for the individual  ## If you want the same sampling time-points you can do that as well: evw <- et(amount.units=\"mg\", time.units=\"hours\") %>%     et(amt=100) %>%     et(0, 24, length.out=50) %>%     et(id=1:100)  sim  <- rxSolve(rx1, theta, evw,  nSub=100, nStud=10,                 thetaMat=thetaMat,                 ## Match boundaries of problem                 thetaLower=0,                  sigma=c(\"prop.sd\", \"add.sd\"), ## Sigmas are standard deviations                 sigmaXform=\"identity\", # default sigma xform=\"identity\"                 omega=c(\"eta.cl\", \"eta.v\", \"eta.ka\"), ## etas are variances                 omegaXform=\"variance\", # default omega xform=\"variance\"                 iCov=data.frame(WT=rnorm(100, 70, 15), CLCR=rnorm(100, 65, 25)),                 dfSub=74, dfObs=476,                 resample=TRUE) #> ℹ thetaMat has too many items, ignored: 'tlag' s <-sim %>% confint(c(\"ipred\")) #> summarizing data... #> done plot(s)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulate-without-uncertainty-in-omega-or-sigma-parameters","dir":"Articles","previous_headings":"Simulation of Clinical Trials","what":"Simulate without uncertainty in omega or sigma parameters","title":"rxode2 Simulation","text":"wish sample prior distributions either omega sigma matrices, can turn feature specifying simVariability = FALSE option solving:  Note since realizations omega sigma simulated, $omegaList $sigmaList return NULL.","code":"sim  <- rxSolve(mod2, ev, nSub=100, thetaMat=tMat, nStud=10,                 simVariability=FALSE)  s <-sim %>% confint(c(\"centr\", \"eff\")) #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(s)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-single-subject.html","id":"single-subject-solving","dir":"Articles","previous_headings":"","what":"Single Subject solving","title":"Single Subject ODE solving -- differences from multiple subject","text":"Originally, rxode2 created solve ODEs one individual. single system without changes individual parameters. course still supported, classic examples found rxode2 intro. article discusses differences multiple subject single subject solving. three differences: Single solving solve ID parallel Single solving lacks id column parameters($params) well actual dataset. Single solving allows parameter exploration easier parameter can modified. multiple subject solves, make sure update individual parameter. first obvious difference speed; multiple subjects can run subject ID parallel. information examples speed gains multiple subject solving see Speeding rxode2 vignette. next difference amount information output final data. Taking 2 compartment indirect response model originally tutorial: Now simple solve  better see differences single solve, can solve 2 individuals  observing two solves, can see: multiple subject solve contains id column data frame data frame parameters subject. last feature obvious, modifying individual parameters. single subject data, can modify rxode2 data frame changing initial conditions parameter values part data frame, described rxode2 Data Frames. multiple subject solving, feature still works, requires care supplying individual’s parameter value, otherwise may change solve drop parameter key individuals.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod1 <- function() {   ini({     KA=2.94E-01     CL=1.86E+01     V2=4.02E+01     Q=1.05E+01     V3=2.97E+02     Kin=1     Kout=1     EC50=200   })   model({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) =-KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  =                    Q*C2 - Q*C3     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1   }) }    et <- et(amount.units='mg', time.units='hours') %>%     et(dose=10000, addl=9, ii=12) %>%     et(amt=20000, nbr.doses=5, start.time=120, dosing.interval=24) %>%     et(0:240) # sampling x <- rxSolve(mod1, et) x #> ── Solved rxode2 object ── #> ── Parameters (x$params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions (x$inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows print(x) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions ($inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows plot(x, C2, eff) x2 <- rxSolve(mod1, et %>% et(id=1:2), params=data.frame(CL=c(18.6, 7.6))) print(x2) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 2 × 9 #>   id       KA    CL    V2     Q    V3   Kin  Kout  EC50 #>   <fct> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 1     0.294  18.6  40.2  10.5   297     1     1   200 #> 2 2     0.294   7.6  40.2  10.5   297     1     1   200 #> ── Initial Conditions ($inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 482 × 8 #>      id time    C2    C3  depot centr  peri   eff #>   <int>  [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1     1    0   0   0     10000     0     0   1    #> 2     1    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3     1    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4     1    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5     1    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6     1    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 476 more rows plot(x2, C2, eff)"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"a-note-about-the-speed-of-the-functional-form-for-rxode2","dir":"Articles","previous_headings":"","what":"A note about the speed of the functional form for rxode2","title":"Speeding up rxode2","text":"functional form benefit supported nlmixr2 therefore one interface solving estimating, takes computation time get underlying “classic” simulation code. models form : can also specify end-points simulation/estimation just like nlmixr2: every solve, compile (cached compile) underlying model. wish speed process can use two underlying rxode2 classic models. takes two steps: Parsing/evaluating model Creating simulation model first step can done rxode2(mod1) mod1() (second model ). second step create underlying “classic” rxode2 model, can done two different methods:$simulationModel $simulationIniModel. $simulationModel provide simulation code without initial conditions pre-pended, $simulationIniModel pre-pend values. endpoints specified, simulation code endpoint also output. can see differences : wish speed multiple simualtions rxode2 functions, need pre-calculate care steps : functions can act like normal ui model solved. can convert back UI .rxUi() function .function() needed. increase speed multiple simulations model use lower level simulation model (ie $simulationModel $simulationIniModel depending need)","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod1 <- function() {   ini({     KA   <- 0.3     CL   <- 7     V2   <- 40     Q    <- 10     V3   <- 300     Kin  <- 0.2     Kout <- 0.2     EC50 <- 8   })   model({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     d/dt(eff) = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1   }) } mod2f <- function() {   ini({     TKA   <- 0.3     TCL   <- 7     TV2   <- 40     TQ    <- 10     TV3   <- 300     TKin  <- 0.2     TKout <- 0.2     TEC50 <- 8     eta.cl + eta.v ~ c(0.09,                        0.08, 0.25)     c2.prop.sd <- 0.1     eff.add.sd <- 0.1   })   model({     KA <- TKA     CL <- TCL*exp(eta.cl)     V2  <- TV2*exp(eta.v)     Q   <- TQ     V3  <- TV3     Kin  <- TKin     Kout <- TKout     EC50 <- TEC50     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     d/dt(eff) = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1     C2 ~ prop(c2.prop.sd)     eff ~ add(eff.add.sd)   }) } mod1 <- mod1() mod2f <- rxode2(mod2f) summary(mod1$simulationModel) #> rxode2 2.1.3 model named rx_4e26e9d8f25df5e07a0521b7b177da87 model (✔ ready).  #> DLL: /tmp/RtmpSPRppR/rxode2/rx_4e26e9d8f25df5e07a0521b7b177da87__.rxd/rx_4e26e9d8f25df5e07a0521b7b177da87_.so #> NULL #>  #> Calculated Variables: #> [1] \"C2\" \"C3\" #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(KA, CL, V2, Q, V3, Kin, Kout, EC50) #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #> }) summary(mod1$simulationIniModel) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_5c7a85ad8a3c6825cf7c3e1131132e2a model (✔ ready).  #> DLL: /tmp/RtmpSPRppR/rxode2/rx_5c7a85ad8a3c6825cf7c3e1131132e2a__.rxd/rx_5c7a85ad8a3c6825cf7c3e1131132e2a_.so #> NULL #>  #> Calculated Variables: #> [1] \"C2\" \"C3\" #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(KA, CL, V2, Q, V3, Kin, Kout, EC50) #>     KA = 0.3 #>     CL = 7 #>     V2 = 40 #>     Q = 10 #>     V3 = 300 #>     Kin = 0.2 #>     Kout = 0.2 #>     EC50 = 8 #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #> }) summary(mod2f$simulationModel) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_ea7cfb51903394d4a9db941ee311eca2 model (✔ ready).  #> DLL: /tmp/RtmpSPRppR/rxode2/rx_ea7cfb51903394d4a9db941ee311eca2__.rxd/rx_ea7cfb51903394d4a9db941ee311eca2_.so #> NULL #>  #> Calculated Variables: #>  [1] \"KA\"       \"CL\"       \"V2\"       \"Q\"        \"V3\"       \"Kin\"      #>  [7] \"Kout\"     \"EC50\"     \"C2\"       \"C3\"       \"ipredSim\" \"sim\"      #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(TKA, TCL, TV2, TQ, TV3, TKin, TKout, TEC50, c2.prop.sd,  #>         eff.add.sd, eta.cl, eta.v) #>     KA = TKA #>     CL = TCL * exp(eta.cl) #>     V2 = TV2 * exp(eta.v) #>     Q = TQ #>     V3 = TV3 #>     Kin = TKin #>     Kout = TKout #>     EC50 = TEC50 #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #>     if (CMT == 5) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ C2 #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (rx_pred_f_ * c2.prop.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.C2, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     if (CMT == 4) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ eff #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (eff.add.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.eff, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     cmt(C2) #>     dvid(5, 4) #> }) summary(mod2f$simulationIniModel) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_d7291e04a9631f78bdda73dc6745fe46 model (✔ ready).  #> DLL: /tmp/RtmpSPRppR/rxode2/rx_d7291e04a9631f78bdda73dc6745fe46__.rxd/rx_d7291e04a9631f78bdda73dc6745fe46_.so #> NULL #>  #> Calculated Variables: #>  [1] \"KA\"       \"CL\"       \"V2\"       \"Q\"        \"V3\"       \"Kin\"      #>  [7] \"Kout\"     \"EC50\"     \"C2\"       \"C3\"       \"ipredSim\" \"sim\"      #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(TKA, TCL, TV2, TQ, TV3, TKin, TKout, TEC50, c2.prop.sd,  #>         eff.add.sd, eta.cl, eta.v) #>     rxerr.C2 = 1 #>     rxerr.eff = 1 #>     TKA = 0.3 #>     TCL = 7 #>     TV2 = 40 #>     TQ = 10 #>     TV3 = 300 #>     TKin = 0.2 #>     TKout = 0.2 #>     TEC50 = 8 #>     c2.prop.sd = 0.1 #>     eff.add.sd = 0.1 #>     eta.cl = 0 #>     eta.v = 0 #>     KA = TKA #>     CL = TCL * exp(eta.cl) #>     V2 = TV2 * exp(eta.v) #>     Q = TQ #>     V3 = TV3 #>     Kin = TKin #>     Kout = TKout #>     EC50 = TEC50 #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #>     if (CMT == 5) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ C2 #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (rx_pred_f_ * c2.prop.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.C2, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     if (CMT == 4) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ eff #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (eff.add.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.eff, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     cmt(C2) #>     dvid(5, 4) #> }) mod1 <- mod1$simulationModel  mod2 <- mod2f$simulationModel"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"increasing-rxode2-speed-by-multi-subject-parallel-solving","dir":"Articles","previous_headings":"","what":"Increasing rxode2 speed by multi-subject parallel solving","title":"Speeding up rxode2","text":"Using classic rxode2 model specification (can convert functional/ui model style) continue discussion rxode2 speed enhancements. rxode2 originally developed ODE solver allowed ODE solve single subject. flexibility still supported. original code rxode2 tutorial :","code":"library(rxode2)  library(microbenchmark) library(ggplot2)  mod1 <- rxode2({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     d/dt(eff) = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1 }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## Create an event table  ev <- et() %>%     et(amt=10000, addl=9,ii=12) %>%     et(time=120, amt=20000, addl=4, ii=24) %>%     et(0:240) ## Add Sampling  nsub <- 100 # 100 sub-problems sigma <- matrix(c(0.09,0.08,0.08,0.25),2,2) # IIV covariance matrix mv <- rxRmvn(n=nsub, rep(0,2), sigma) # Sample from covariance matrix CL <- 7*exp(mv[,1]) V2 <- 40*exp(mv[,2]) params.all <- cbind(KA=0.3, CL=CL, V2=V2, Q=10, V3=300,                     Kin=0.2, Kout=0.2, EC50=8)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"for-loop","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"For Loop","title":"Speeding up rxode2","text":"slowest way code use loop. example enclose function compare timing.","code":"runFor <- function(){     res <- NULL     for (i in 1:nsub) {         params <- params.all[i,]         x <- mod1$solve(params, ev)         ##Store results for effect compartment         res <- cbind(res, x[, \"eff\"])     }     return(res) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"running-with-apply","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Running with apply","title":"Speeding up rxode2","text":"general R, apply types functions perform better loop, tutorial also suggests speed enhancement","code":"runSapply <- function(){     res <- apply(params.all, 1, function(theta)         mod1$run(theta, ev)[, \"eff\"]) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"run-using-a-single-threaded-solve","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Run using a single-threaded solve","title":"Speeding up rxode2","text":"can also rxode2 solve subject simultaneously without collecting results R, using single threaded solve. data output slightly different , still gives information:","code":"runSingleThread <- function(){   solve(mod1, params.all, ev, cores=1)[,c(\"sim.id\", \"time\", \"eff\")] }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"run-a-2-threaded-solve","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Run a 2 threaded solve","title":"Speeding up rxode2","text":"rxode2 supports multi-threaded solves, another option 2 threads (called cores solve options, can see options rxControl() rxSolve()).","code":"run2Thread <- function(){   solve(mod1, params.all, ev, cores=2)[,c(\"sim.id\", \"time\", \"eff\")] }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"compare-the-times-between-all-the-methods","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Compare the times between all the methods","title":"Speeding up rxode2","text":"Now moment truth, timings:  clear largest jump performance using solve method providing parameters rxode2 solve without looping subject either sapply. number cores/threads applied solve also plays role solving. can explore number threads following code:  can suite spot speed vs number cores. system type (mac, linux, windows /processor), complexity ODE solving number subjects may affect arbitrary number threads. 4 threads good number use without prior knowledge systems days least 4 threads (2 processors 4 threads).","code":"bench <- microbenchmark(runFor(), runSapply(), runSingleThread(),run2Thread()) print(bench) #> Unit: milliseconds #>               expr       min        lq      mean    median        uq       max #>           runFor() 265.60642 270.91812 280.10508 273.27375 276.22522 390.10634 #>        runSapply() 264.82305 271.37265 280.13567 273.25948 277.09826 440.15521 #>  runSingleThread()  27.71541  27.86995  28.73338  27.96453  28.08604  48.11461 #>       run2Thread()  16.31607  16.43347  17.25248  16.52140  16.67760  27.95446 #>  neval #>    100 #>    100 #>    100 #>    100 autoplot(bench) runThread <- function(n){     solve(mod1, params.all, ev, cores=n)[,c(\"sim.id\", \"time\", \"eff\")] }  bench <- eval(parse(text=sprintf(\"microbenchmark(%s)\",                                      paste(paste0(\"runThread(\", seq(1, 2 * rxCores()),\")\"),                                            collapse=\",\")))) print(bench) #> Unit: milliseconds #>          expr      min       lq     mean   median       uq      max neval #>  runThread(1) 27.73454 27.90294 29.28469 28.52594 29.34674 48.13157   100 #>  runThread(2) 16.29712 16.53740 18.44910 17.22503 17.44130 27.90322   100 #>  runThread(3) 17.40113 19.47790 19.87893 19.73595 20.33017 25.56011   100 #>  runThread(4) 15.93134 16.07433 18.29293 16.73419 16.86182 38.19777   100 autoplot(bench)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"increasing-speed-with-compiler-options","dir":"Articles","previous_headings":"","what":"Increasing speed with compiler options","title":"Speeding up rxode2","text":"One way allows faster ODE solving make approximations make math operators like exp() faster technically accurate enough follow IEEE standard math functions values (implications cover ). optimizations opt-Julia since compile everything session, CRAN conservative approach since individuals compile R function running . Still, rxode2 models can compiled option without disturbing CRAN policies. key set option. example:  Note compiler settings can tricky setup system wide Makevars may interact setting. example use ccache compile may produced options since cached options. example, github runner (generates page), advantage \"fast\" compile. However, development laptop minimal speed increase. probably check using . disabled default since minimum increase speed.","code":"# Using the first example subset to PK mod2f <- function() {   ini({     TKA   <- 0.3     TCL   <- 7     TV2   <- 40     TQ    <- 10     TV3   <- 300     TKin  <- 0.2     TKout <- 0.2     TEC50 <- 8     eta.cl + eta.v ~ c(0.09,                        0.08, 0.25)     c2.prop.sd <- 0.1   })   model({     KA <- TKA     CL <- TCL*exp(eta.cl)     V2  <- TV2*exp(eta.v)     Q   <- TQ     V3  <- TV3     Kin  <- TKin     Kout <- TKout     EC50 <- TEC50     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     C2 ~ prop(c2.prop.sd)   }) }  mod2f <- mod2f()  mod2s <- mod2f$simulationIniModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ev  <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(0:240) # Add sampling  bench1 <- microbenchmark(standardCompile=rxSolve(mod2s, ev, nSub=1000))  # Now clear the cache of models so we can change the compile options for the same model rxClean()  # Use withr to preserve the options withr::with_options(list(rxode2.compile.O=\"fast\"), {   mod2s <- mod2f$simulationIniModel }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ bench2 <- microbenchmark(fastCompile=rxSolve(mod2s, ev, nSub=1000))  bench <- rbind(bench1, bench2)  print(bench) #> Unit: milliseconds #>             expr      min       lq     mean   median       uq      max neval #>  standardCompile 176.6202 183.5591 219.5272 191.7218 238.6262 350.7461   100 #>      fastCompile 176.9396 183.0521 219.0194 187.1348 235.6830 351.4314   100 autoplot(bench)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"a-real-life-example","dir":"Articles","previous_headings":"","what":"A real life example","title":"Speeding up rxode2","text":"cBefore parallel solving implemented, fastest way run rxode2 lapply. Rik Schoemaker created data-set nlmixr comparisons, reduced run faster automatic building pkgdown website. applying new parallel solving concepts can simply run simulation less code faster: can see striking time difference two methods; things keep mind: rxode2 use thread-safe sitmo threefry routines simulation eta values. Therefore results expected different (also random samples taken different order different) prior simulation run R 3.5, different random number generator results simulation different actual nlmixr comparison using slower simulation. speed comparison used data.table. rxode2 uses data.table internally (available) try speed sorting, different installations data.table installed. can force rxode2 use order() sorting using forderForceBase(TRUE). case little difference two, though examples data.table’s presence leads speed increase (less likely lead slowdown).","code":"library(rxode2) library(data.table) #Define the rxode2 model   ode1 <- \"   d/dt(abs)    = -KA*abs;   d/dt(centr)  =  KA*abs-(CL/V)*centr;   C2=centr/V;   \"  #Create the rxode2 simulation object mod1 <- rxode2(model = ode1) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #Population parameter values on log-scale   paramsl <- c(CL = log(4),                V = log(70),                KA = log(1)) #make 10,000 subjects to sample from:   nsubg <- 300 # subjects per dose   doses <- c(10, 30, 60, 120)   nsub <- nsubg * length(doses) #IIV of 30% for each parameter   omega <- diag(c(0.09, 0.09, 0.09))# IIV covariance matrix   sigma <- 0.2 #Sample from the multivariate normal   set.seed(98176247)   rxSetSeed(98176247)   library(MASS)   mv <-     mvrnorm(nsub, rep(0, dim(omega)[1]), omega) # Sample from covariance matrix #Combine population parameters with IIV   params.all <-     data.table(       \"ID\" = seq(1:nsub),       \"CL\" = exp(paramsl['CL'] + mv[, 1]),       \"V\" = exp(paramsl['V'] + mv[, 2]),       \"KA\" = exp(paramsl['KA'] + mv[, 3])     ) #set the doses (looping through the 4 doses) params.all[, AMT := rep(100 * doses,nsubg)]  Startlapply <- Sys.time()  #Run the simulations using lapply for speed   s = lapply(1:nsub, function(i) { #selects the parameters associated with the subject to be simulated     params <- params.all[i] #creates an eventTable with 7 doses every 24 hours     ev <- eventTable()     ev$add.dosing(       dose = params$AMT,       nbr.doses = 1,       dosing.to = 1,       rate = NULL,       start.time = 0     ) #generates 4 random samples in a 24 hour period     ev$add.sampling(c(0, sort(round(sample(runif(600, 0, 1440), 4) / 60, 2)))) #runs the rxode2 simulation     x <- as.data.table(mod1$run(params, ev)) #merges the parameters and ID number to the simulation output     x[, names(params) := params]   })  #runs the entire sequence of 100 subjects and binds the results to the object res   res = as.data.table(do.call(\"rbind\", s))  Stoplapply <- Sys.time()  print(Stoplapply - Startlapply) #> Time difference of 11.31994 secs rx <- rxode2({     CL =  log(4)     V = log(70)     KA = log(1)     CL = exp(CL + eta.CL)     V = exp(V + eta.V)     KA = exp(KA + eta.KA)     d/dt(abs)    = -KA*abs;     d/dt(centr)  =  KA*abs-(CL/V)*centr;     C2=centr/V; }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ omega <- lotri(eta.CL ~ 0.09,                eta.V ~ 0.09,                eta.KA ~ 0.09)  doses <- c(10, 30, 60, 120)   startParallel <- Sys.time() ev <- do.call(\"rbind\",         lapply(seq_along(doses), function(i){             et() %>%                 et(amt=doses[i]) %>% # Add single dose                 et(0) %>% # Add 0 observation                 ## Generate 4 samples in 24 hour period                 et(lapply(1:4, function(...){c(0, 24)})) %>%                 et(id=seq(1, nsubg) + (i - 1) * nsubg) %>%                 ## Convert to data frame to skip sorting the data                 ## When binding the data together                 as.data.frame         })) ## To better compare, use the same output, that is data.table res <- rxSolve(rx, ev, omega=omega, returnType=\"data.table\") endParallel <- Sys.time() print(endParallel - startParallel) #> Time difference of 0.1070809 secs"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"want-more-ways-to-run-multi-subject-simulations","dir":"Articles","previous_headings":"A real life example","what":"Want more ways to run multi-subject simulations","title":"Speeding up rxode2","text":"version since tutorial even ways run multi-subject simulations, including adding variability sampling dosing times et() (see rxode2 events information), ability supply omega sigma matrix well adding thetaMat R simulate uncertainty omega, sigma theta matrices; see rxode2 simulation vignette.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"session-information","dir":"Articles","previous_headings":"A real life example","what":"Session Information","title":"Speeding up rxode2","text":"session information:","code":"sessionInfo() #> R version 4.4.0 (2024-04-24) #> Platform: x86_64-pc-linux-gnu #> Running under: Ubuntu 22.04.4 LTS #>  #> Matrix products: default #> BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3  #> LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.20.so;  LAPACK version 3.10.0 #>  #> locale: #>  [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8        #>  [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8    #>  [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C           #> [10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C    #>  #> time zone: UTC #> tzcode source: system (glibc) #>  #> attached base packages: #> [1] stats     graphics  grDevices utils     datasets  methods   base      #>  #> other attached packages: #> [1] MASS_7.3-60.2         data.table_1.15.4     ggplot2_3.5.1         #> [4] microbenchmark_1.4.10 rxode2_2.1.3          #>  #> loaded via a namespace (and not attached): #>  [1] gtable_0.3.5         xfun_0.44            bslib_0.7.0          #>  [4] htmlwidgets_1.6.4    RApiSerialize_0.1.3  lattice_0.22-6       #>  [7] vctrs_0.6.5          tools_4.4.0          generics_0.1.3       #> [10] tibble_3.2.1         symengine_0.2.6      fansi_1.0.6          #> [13] highr_0.11           pkgconfig_2.0.3      checkmate_2.3.1      #> [16] desc_1.4.3           RcppParallel_5.1.7   lifecycle_1.0.4      #> [19] compiler_4.4.0       farver_2.1.2         textshaping_0.4.0    #> [22] munsell_0.5.1        qs_0.26.3            htmltools_0.5.8.1    #> [25] sys_3.4.2            sass_0.4.9           yaml_2.3.8           #> [28] lazyeval_0.2.2       pillar_1.9.0         pkgdown_2.0.9        #> [31] crayon_1.5.2         jquerylib_0.1.4      cachem_1.1.0         #> [34] nlme_3.1-164         tidyselect_1.2.1     digest_0.6.35        #> [37] lotri_0.4.4          dplyr_1.1.4          purrr_1.0.2          #> [40] labeling_0.4.3       rxode2ll_2.0.11.9000 fastmap_1.2.0        #> [43] grid_4.4.0           colorspace_2.1-0     rxode2parse_2.0.19   #> [46] cli_3.6.2            dparser_1.3.1-11     magrittr_2.0.3       #> [49] utf8_1.2.4           withr_3.0.0          scales_1.3.0         #> [52] backports_1.5.0      rmarkdown_2.27       ragg_1.3.2           #> [55] stringfish_0.16.0    memoise_2.0.1        evaluate_0.23        #> [58] knitr_1.46           rex_1.2.1            rxode2et_2.0.13      #> [61] rxode2random_2.1.1   PreciseSums_0.6      rlang_1.1.3          #> [64] Rcpp_1.0.12          glue_1.7.0           jsonlite_1.8.8       #> [67] R6_2.5.1             systemfonts_1.1.0    fs_1.6.4             #> [70] units_0.8-5"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-stiff.html","id":"stiff-odes-with-jacobian-specification","dir":"Articles","previous_headings":"","what":"Stiff ODEs with Jacobian Specification","title":"rxode2 Jacobian specification and Stiff Systems","text":"Occasionally, may come across stiff differential equation, differential equation numerically unstable small variations parameters cause different solutions ODEs. One way tackle choose stiff-solver, hybrid stiff solver (like default LSODA). Typically enough. However exact Jacobian solutions may increase stability ODE. (Note Jacobian derivative ODE specification respect variable). rxode2 can specify Jacobian df(state)/dy(variable)= statement. classic ODE stiff properties various conditions Van der Pol differential equations. rxode2 can specified following: stiff mu=1, mu=1000 stiff system easy enough , bit tedious. rxode2 setup appropriately, can use computer algebra system sympy calculate Jacobian automatically. done rxode2 option calcJac option:","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` Vtpol2 <- function() {   ini({     mu <- 1 ## nonstiff; 10 moderately stiff; 1000 stiff   })   model({     d/dt(y)       <- dy     d/dt(dy)      <- mu*(1-y^2)*dy - y     ## Jacobian     df(y)/dy(dy)  <- 1     df(dy)/dy(y)  <- -2*dy*mu*y - 1     df(dy)/dy(dy) <- mu*(1-y^2)     ## Initial conditions     y(0)          <- 2     dy(0)         <- 0   }) }  et <- et(0, 10, length.out=200) %>%   et(amt=0)  s1 <- Vtpol2 %>%  solve(et, method=\"lsoda\") ## i parameter labels from comments will be replaced by 'label()' ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(s1) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ## mu  ##  1  ## -- Initial Conditions ($inits): -- ##  y dy  ##  2  0  ## -- First part of data (object): -- ## # A tibble: 200 x 3 ##     time     y      dy ##    <dbl> <dbl>   <dbl> ## 1 0       2     0      ## 2 0.0503  2.00 -0.0933 ## 3 0.101   1.99 -0.173  ## 4 0.151   1.98 -0.242  ## 5 0.201   1.97 -0.302  ## 6 0.251   1.95 -0.353  ## # i 194 more rows s2 <- Vtpol2 %>%  solve(c(mu=1000), et) ## i parameter labels from comments will be replaced by 'label()' ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(s2) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##   mu  ## 1000  ## -- Initial Conditions ($inits): -- ##  y dy  ##  2  0  ## -- First part of data (object): -- ## # A tibble: 200 x 3 ##     time     y        dy ##    <dbl> <dbl>     <dbl> ## 1 0       2     0        ## 2 0.0503  2.00 -0.000667 ## 3 0.101   2.00 -0.000667 ## 4 0.151   2.00 -0.000667 ## 5 0.201   2.00 -0.000667 ## 6 0.251   2.00 -0.000667 ## # i 194 more rows Vtpol <- function() {   ini({     mu <- 1 ## nonstiff; 10 moderately stiff; 1000 stiff   })   model({     d/dt(y)       <- dy     d/dt(dy)      <- mu*(1-y^2)*dy - y     y(0)          <- 2     dy(0)         <- 0   }) }  Vtpol <- Vtpol()  # you can also use $symengineModelPrune if there is if/else blocks # that need to be converted: Vtpol <- rxode2(Vtpol$symengineModelNoPrune,  calcJac=TRUE) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## > pruning branches (`if`/`else`)... ## v done ## > loading into symengine environment... ## v done ## > calculate jacobian ## [====|====|====|====|====|====|====|====|====|====] 0:00:00 ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ summary(Vtpol) ## rxode2 2.1.3 model named rx_cc493e863ffdb1841288028f0470c2e6 model (ready).  ## DLL: /tmp/RtmpSPRppR/rxode2/rx_cc493e863ffdb1841288028f0470c2e6__.rxd/rx_cc493e863ffdb1841288028f0470c2e6_.so ## NULL ## -- rxode2 Model Syntax -- ## rxode2({ ##     cmt(y) ##     cmt(dy) ##     d/dt(y) = dy ##     d/dt(dy) = -y + mu * dy * (1 - Rx_pow_di(y, 2)) ##     y(0) = 2 ##     dy(0) = 0 ##     df(y)/dy(y) = 0 ##     df(dy)/dy(y) = -1 - 2 * y * mu * dy ##     df(y)/dy(dy) = 1 ##     df(dy)/dy(dy) = mu * (1 - Rx_pow_di(y, 2)) ##     df(y)/dy(mu) = 0 ##     df(dy)/dy(mu) = dy * (1 - Rx_pow_di(y, 2)) ## })"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"introduction","dir":"Articles","previous_headings":"","what":"Introduction","title":"rxode2 ODE solving syntax","text":"briefly describes syntax used define models rxode2 translate R-callable compiled code. also describes communication variables R rxode2 modeling specification.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"creating-rxode2-models","dir":"Articles","previous_headings":"","what":"Creating rxode2 models","title":"rxode2 ODE solving syntax","text":"ODE-based model specification may coded inside four places: Inside rxode2({}) block statements: Inside rxode2(\"\") string statement: file name loaded rxode2: model function can parsed rxode2: block statement, character string text file internal rxode2 compilation manager translates ODE system C, compiles loads R session. call rxode2 produces object class rxode2 consists list-like structure (environment) various member functions. last type model (model function), call rxode2 creates parsed rxode2 ui can translated rxode2 compilation model. type function required nlmixr2 estimation can extended modified model piping. reason focused documentation.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod <- rxode2({   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ mod <- rxode2(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \") #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ writeLines(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \", \"modelFile.rxode2\") mod <- rxode2(filename='modelFile.rxode2') unlink(\"modelFile.rxode2\") mod <- function() {   model({     # simple assignment     C2 <- centr/V2      # time-derivative assignment     d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;   }) }  mod <- rxode2(mod) # or simply mod() if the model is at the end of the function  # These model functions often have residual components and initial # (`ini({})`) conditions attached as well. For example the # theophylline model can be written as:  one.compartment <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  # after parsing the model mod <- one.compartment() mod$simulationModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_b794dde852cc7c8fe9aebc3eea58fddc model (✔ ready).  #> x$state: depot, center #> x$stateExtra: cp #> x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp #> x$lhs: ka, cl, v, cp, ipredSim, sim # or  mod$simulationIniModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_a9629bdbda112e498cbaf24cafcccfe9 model (✔ ready).  #> x$state: depot, center #> x$stateExtra: cp #> x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp #> x$lhs: ka, cl, v, cp, ipredSim, sim"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"syntax","dir":"Articles","previous_headings":"","what":"Syntax","title":"rxode2 ODE solving syntax","text":"basic model specification consists one statements optionally terminated semi-colons ; optional comments (comments delimited # end--line). block statements set statements delimited curly braces, { ... }. Statements can either assignments, conditional /else /else, loops (can exited break), special statements, printing statements (debugging/testing). Assignment statements can : simple assignments, left hand identifier (.e., variable) special time-derivative assignments, left hand specifies change amount corresponding state variable (compartment) respect time e.g., d/dt(depot): special initial-condition assignments left hand specifies compartment initial condition specified, e.g. depot(0) = 0 special model event changes including bioavailability (f(depot)=1), lag time (alag(depot)=0), modeled rate (rate(depot)=2) modeled duration (dur(depot)=2). example model features event specification modeled infusions rxode2 data specification found rxode2 events vignette. special change point syntax, model times. model times specified mtime(var)=time special Jacobian-derivative assignments, left hand specifies change compartment ode respect variable. example, d/dt(y) = dy, Jacobian compartment can specified df(y)/dy(dy) = 1. may advantage obtaining solution specifying Jacobian stiff ODE systems. However, stiff systems tried LSODA, actually slightly slowed solving. Note assignment can done =, <- ~. assigning ~ operator, simple assignments time-derivative assignments output. Note rxode2 model functions assignment ~ can also overloaded residual distribution specification. Special statements can : Compartment declaration statements, can change default dosing compartment assumed compartment number(s) well add extra compartment names end (useful multiple-endpoint nlmixr models); specified cmt(compartmentName) Parameter declaration statements, can make sure input parameters certain order instead ordering parameters order parsed. useful keeping parameter order using 2 different ODE models. specified param(par1, par2,...) example model shown : Expressions assignment statements can numeric logical. Numeric expressions can include following numeric operators +, -, *, /, ^ mathematical functions defined C R math libraries (e.g., fabs, exp, log, sin, abs). may also access R’s functions R math libraries, like lgammafn log gamma function. rxode2 syntax case-sensitive, .e., ABC different abc, Abc, ABc, etc.","code":"# simple assignment    C2 <- centr/V2     # time-derivative assignment    d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"identifiers","dir":"Articles","previous_headings":"Syntax","what":"Identifiers","title":"rxode2 ODE solving syntax","text":"Like R, Identifiers (variable names) may consist one alphanumeric, underscore _ period . characters, first character digit underscore _. Identifiers model specification can refer : State variables dynamic system (e.g., compartments pharmacokinetics model). Implied input variable, t (time), tlast (last time point), podo (oral dose, undocumented case absorption transit models). Special constants like pi R’s predefined constants. Model parameters (e.g., ka rate absorption, CL clearance, etc.) Others, created assignments part model specification; referred LHS (left-hand side) variable. Currently, rxode2 modeling language recognizes system state variables “parameters”, thus, values need passed R ODE model (e.g., age) either passed params argument integrator function rxSolve() supplied event data-set. certain variable names rxode2 event tables. avoid confusion, following event table-related items assigned, used state can accessed rxode2 code: cmt dvid addl ss rate id However following variables used model specification: evid ii Sometimes rxode2 generates variables fed back rxode2. Similarly, nlmixr2 generates variables used nlmixr estimation simulation. variables start either rx nlmixr prefixes. avoid problems, suggested use variables starting either rx nlmixr prefixes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"logical-operators","dir":"Articles","previous_headings":"","what":"Logical Operators","title":"rxode2 ODE solving syntax","text":"Logical operators support standard R operators ==, != >= <= > <. Like R can () () statements, ifelse() expressions. Additionally can standard assignment. instance, following valid: Notice can also use character expressions comparisons. convenience comes cost since character comparisons slower numeric expressions. Unlike R, .numeric .integer logical statements needed, cause syntax error try use function.","code":"cov1 = covm*(sexf == \"female\") + covm*(sexf != \"female\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"supported-functions","dir":"Articles","previous_headings":"","what":"Supported functions","title":"rxode2 ODE solving syntax","text":"supported functions rxode2 can seen rxSupportedFuns(). brief description built-functions following table: Note lag(cmt) = equivalent alag(cmt) = = lag(wt)","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"reserved-keywords","dir":"Articles","previous_headings":"","what":"Reserved keywords","title":"rxode2 ODE solving syntax","text":"reserved keywords rxode2 model. following table: Note rxFlag always output 11 calc_lhs since final variables calculated, though can tweak test certain parts rxode2 using flag.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"residual-functions-when-using-rxode2-functions","dir":"Articles","previous_headings":"","what":"Residual functions when using rxode2 functions","title":"rxode2 ODE solving syntax","text":"addition ~ hiding output certain types output, also used specify residual output endpoint input rxode2 model function (includes residual model({}) block). specifications form: Indicating variable var variable represents individual central tendencies model also represents compartment specification data-set. can also change compartment name using | syntax, : case var represents central tendency cmt represents compartment dvid specification.","code":"var ~ add(add.sd) var ~ add(add.sd) | cmt"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"transformations","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Transformations","title":"rxode2 ODE solving syntax","text":"normal related distributions, can apply transformation sides using keywords/functions apply transformations. default likelihood transformations calculated untransformed scale. bounded variables like logit-normal probit-normal low high values defaulted 0 1 missing. models wish proportional model one transformation can replace standard deviation NA allow transformations, lnorm(), probitNorm() logitNorm() can combined variance stabilizing yeoJohnson() transformation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"normal-and-t-related-distributions","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Normal and t-related distributions","title":"rxode2 ODE solving syntax","text":"normal t-related distributions, wanted keep ability use skewed distributions additive proportional t/cauchy-space, distributions specified differently comparison supported distributions within nlmixr2: Note normal t-related distributions nlmixr2 calculate cwres npde normal assumption help assess goodness fit model. Also note +dnorm() mostly testing purposes slow estimation procedure nlmixr2. suggest adding (except explicit testing). multiple endpoint models mix non-normal normal distributions, whole problem shifted log-likelihood method estimation nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"notes-on-additive-proportional-models","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Notes on additive + proportional models","title":"rxode2 ODE solving syntax","text":"two different ways specify additive proportional models, call combined1 combined2, way Monolix calls two distributions (avoid software differences naming). first, combined1, assumes additive proportional differences standard deviation scale, : y=f+(+b* f^c)*err second, combined2, assumes additive proportional differences combined variance scale: y=f+\\[sqrt(^2+b^2 *f^(2c))\\]*err default nlmixr2/rxode2 otherwise specified combined2 since mirrors adding 2 normal distributions statistics add variances (standard deviations). However, combined1 can describe data possibly even better combined2 possible options rxode2/nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"distributions-of-known-likelihoods","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Distributions of known likelihoods","title":"rxode2 ODE solving syntax","text":"residuals related normal, t-distribution cauchy, often residual specification form: compartment specification left handed side specification. generalized likelihood can specify:","code":"cmt ~ dbeta(alpha, beta) ll(cmt) ~ llik specification"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"ordinal-likelihoods","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Ordinal likelihoods","title":"rxode2 ODE solving syntax","text":"Finally, ordinal likelihoods/simulations can specified 2 ways. first : err represents compartment p0 probability specific category: model ensure sum p values less 1. Additionally can write arbitrary number categories ordinal model described . seems little p0 probability category 1 sometimes scores non-whole numbers. can modeled follows: numeric categories specified explicitly, probabilities remain :","code":"err ~ c(p0, p1, p2) err ~ c(p0=0, p1=1, p2=2, 3)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"general-table-of-supported-residual-distributions","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"General table of supported residual distributions","title":"rxode2 ODE solving syntax","text":"general supported following table (available rxode2::rxResidualError)","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"bugs-andor-deficiencies","dir":"Articles","previous_headings":"","what":"Bugs and/or deficiencies","title":"rxode2 ODE solving syntax","text":"modulo operator %% currently unsupported.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"note","dir":"Articles","previous_headings":"","what":"Note","title":"rxode2 ODE solving syntax","text":"ODE specification mini-language parsed help open source tool , Plevyak (2015).","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"example-model","dir":"Articles","previous_headings":"","what":"Example model","title":"rxode2 ODE solving syntax","text":"commented example quickly show capabilities rxode2 syntax.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"example","dir":"Articles","previous_headings":"","what":"Example","title":"rxode2 ODE solving syntax","text":"","code":"f <- function() {   ini({        })   model({     # An rxode2 model specification (this line is a comment).      if(comed==0) {  # concomitant medication (con-med)?       F <- 1.0     # full bioavailability w.o. con-med     } else {       F <- 0.80    # 20% reduced bioavailability     }      C2 <- centr/V2  # concentration in the central compartment     C3 <- peri/V3   # concentration in the peripheral compartment      # ODE describing the PK and PD      d/dt(depot) <- -KA*depot     d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                      Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0)      <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"interface-and-data-handling-between-r-and-the-generated-c-code","dir":"Articles","previous_headings":"","what":"Interface and data handling between R and the generated C code","title":"rxode2 ODE solving syntax","text":"Users specify variables dynamic system’s state variables via d/dt(identifier) operator part model specification, model parameters via params= argument rxode2 solve() method: values variables pre-specified time points saved model fitting/integration returned part fitted values (see function et(), define set time points capture values variables) returned part modeling output. ODE specification mini-language parsed help open source tool DParser, Plevyak (2015).","code":"m1 <- rxode2(model = ode, modName = \"m1\")  # model parameters -- a named vector is required theta <-     c(KA=0.29, CL=18.6, V2=40.2, Q=10.5, V3=297, Kin=1, Kout=1, EC50=200)  # state variables and their amounts at time 0 (the use of names is # encouraged, but not required) inits <- c(depot=0, centr=0, peri=0, eff=1)  # qd1 is an eventTable specification with a set of dosing and sampling  # records (code not shown here)  solve(theta, event = qd1, inits = inits)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-tutorials.html","id":"interactive-tutorials","dir":"Articles","previous_headings":"","what":"Interactive tutorials","title":"Interactive Tutorials","text":"rxode2 comes interactive tutorials can run R session; built Rstudio 1.3 can run R session :","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` ## rxode2 Syntax learnr::run_tutorial(\"rxode2-00-syntax\", \"rxode2\")  ## rxode2 event tables   learnr::run_tutorial(\"rxode2-00-events\", \"rxode2\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"creating-a-2-compartment-model-in-rxode2","dir":"Articles","previous_headings":"","what":"Creating a 2-compartment model in rxode2","title":"Weight Based Dosing in rxode2","text":"","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` ## Note the time covariate is not included in the simulation m1 <- function() {   model({     CL ~ (1-0.2*SEX)*(0.807+0.00514*(CRCL-91.2))*exp(eta.cl)     V1 ~ 4.8*exp(eta.v1)     Q ~ (3.46+0.0593*(WT-75.1))*exp(eta.q);     V2 ~ 1.93*(3.13+0.0458*(WT-75.1))*exp(eta.v2)     A1 ~ centr;     A2 ~ peri;     d/dt(centr) <- - A1*(CL/V1 + Q/V1) + A2*Q/V2;     d/dt(peri) <- A1*Q/V1 - A2*Q/V2;     DV = centr / V1 * (1 + prop.err)   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"simulating-covariates","dir":"Articles","previous_headings":"","what":"Simulating Covariates","title":"Weight Based Dosing in rxode2","text":"simulation correlates age, sex, weight. Since using weight based dosing, needs simulated first","code":"set.seed(42) rxSetSeed(42) library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union nsub=30 # Simulate Weight based on age and gender AGE<-round(runif(nsub,min=18,max=70)) SEX<-round(runif(nsub,min=0,max=1)) HTm<-round(rnorm(nsub,176.3,0.17*sqrt(4482)),digits=1) HTf<-round(rnorm(nsub,162.2,0.16*sqrt(4857)),digits=1) WTm<-round(exp(3.28+1.92*log(HTm/100))*exp(rnorm(nsub,0,0.14)),digits=1) WTf<-round(exp(3.49+1.45*log(HTf/100))*exp(rnorm(nsub,0,0.17)),digits=1) WT<-ifelse(SEX==1,WTf,WTm) CRCL<-round(runif(nsub,30,140)) ## id is in lower case to match the event table cov.df <- tibble(id=seq_along(AGE), AGE=AGE, SEX=SEX, WT=WT, CRCL=CRCL) print(cov.df) #> # A tibble: 30 x 5 #>       id   AGE   SEX    WT  CRCL #>    <int> <dbl> <dbl> <dbl> <dbl> #>  1     1    66     1  49.4    83 #>  2     2    67     1  52.5    79 #>  3     3    33     0  97.9    37 #>  4     4    61     1  63.8    66 #>  5     5    51     0  71.8   127 #>  6     6    45     1  69.6   132 #>  7     7    56     0  61      73 #>  8     8    25     0  57.7    47 #>  9     9    52     1  58.7    65 #> 10    10    55     1  73.1    64 #> # i 20 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"creating-weight-based-event-table","dir":"Articles","previous_headings":"","what":"Creating weight based event table","title":"Weight Based Dosing in rxode2","text":"","code":"s<-c(0,0.25,0.5,0.75,1,1.5,seq(2,24,by=1)) s <- lapply(s, function(x){.x <- 0.1 * x; c(x - .x, x + .x)})  e <- et() %>%     ## Specify the id and weight based dosing from covariate data.frame     ## This requires rxode2 XXX      et(id=cov.df$id, amt=6*cov.df$WT, rate=6 * cov.df$WT) %>%     ## Sampling is added for each ID     et(s) %>%     as.data.frame %>%     ## Merge the event table with the covarite information     merge(cov.df, by=\"id\") %>%     as_tibble   e #> # A tibble: 900 x 12 #>       id    low  time   high cmt         amt  rate  evid   AGE   SEX    WT  CRCL #>    <int>  <dbl> <dbl>  <dbl> <chr>     <dbl> <dbl> <int> <dbl> <dbl> <dbl> <dbl> #>  1     1  0     0      0     (obs)       NA    NA      0    66     1  49.4    83 #>  2     1 NA     0     NA     (default)  296.  296.     1    66     1  49.4    83 #>  3     1  0.225 0.246  0.275 (obs)       NA    NA      0    66     1  49.4    83 #>  4     1  0.45  0.516  0.55  (obs)       NA    NA      0    66     1  49.4    83 #>  5     1  0.675 0.729  0.825 (obs)       NA    NA      0    66     1  49.4    83 #>  6     1  0.9   0.921  1.1   (obs)       NA    NA      0    66     1  49.4    83 #>  7     1  1.35  1.42   1.65  (obs)       NA    NA      0    66     1  49.4    83 #>  8     1  1.8   1.82   2.2   (obs)       NA    NA      0    66     1  49.4    83 #>  9     1  2.7   2.97   3.3   (obs)       NA    NA      0    66     1  49.4    83 #> 10     1  3.6   3.87   4.4   (obs)       NA    NA      0    66     1  49.4    83 #> # i 890 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"solving-daptomycin-simulation","dir":"Articles","previous_headings":"","what":"Solving Daptomycin simulation","title":"Weight Based Dosing in rxode2","text":"","code":"data <- rxSolve(m1, e,              ## Lotri uses lower-triangular matrix rep. for named matrix              omega=lotri(eta.cl ~ .306,                           eta.q ~0.0652,                          eta.v1 ~.567,                          eta.v2 ~ .191),              sigma=lotri(prop.err ~ 0.15),              addDosing = TRUE, addCov = TRUE) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(data) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #> # A tibble: 30 x 5 #>    id    eta.cl  eta.v1      eta.q  eta.v2 #>    <fct>  <dbl>   <dbl>      <dbl>   <dbl> #>  1 1      0.343 -1.25   -0.0936    -0.253  #>  2 2      0.444 -0.0581  0.187      0.233  #>  3 3      0.142  0.203   0.0617     1.03   #>  4 4      0.274 -0.733  -0.0000116 -0.321  #>  5 5      0.505 -0.104  -0.406     -0.518  #>  6 6      0.472  1.48   -0.537      0.210  #>  7 7      0.156 -0.0295  0.112     -0.226  #>  8 8     -0.747  0.519  -0.347      0.476  #>  9 9      0.326 -0.373   0.340      0.470  #> 10 10     0.428 -1.28    0.131     -0.0240 #> # i 20 more rows #> -- Initial Conditions ($inits): -- #> centr  peri  #>     0     0  #> -- First part of data (object): -- #> # A tibble: 900 x 12 #>      id  evid   cmt   amt  rate  time    DV centr  peri   SEX    WT  CRCL #>   <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1     1     1     1  296.  296. 0       0     0    0        1  49.4    83 #> 2     1     0    NA   NA    NA  0       0     0    0        1  49.4    83 #> 3     1     0    NA   NA    NA  0.246  41.3  58.7  9.45     1  49.4    83 #> 4     1     0    NA   NA    NA  0.516  31.0 100.  34.0      1  49.4    83 #> 5     1     0    NA   NA    NA  0.729  72.2 124.  58.9      1  49.4    83 #> 6     1     0    NA   NA    NA  0.921  82.0 140.  83.1      1  49.4    83 #> # i 894 more rows plot(data, log=\"y\") #> Warning in transformation$transform(x): NaNs produced #> Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = #> minor_breaks, : log-10 transformation introduced infinite #> values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"daptomycin-reference","dir":"Articles","previous_headings":"","what":"Daptomycin Reference","title":"Weight Based Dosing in rxode2","text":"weight-based simulation adapted Daptomycin article : Dvorchik B, Arbeit RD, Chung J, Liu S, Knebel W, Kastrissios H. Population pharmacokinetics daptomycin. Antimicrob Agents Che mother 2004; 48: 2799-2807. doi:(10.1128/AAC.48.8.2799-2807.2004)[https://dx.doi.org/10.1128%2FAAC.48.8.2799-2807.2004] simulation example made available work Sherwin Sy modifications Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors","text":"Matthew L. Fidler (core team/developer/manual) Melissa Hallow (tutorial writer) Wenping Wang (core team/developer)","code":""},{"path":"https://nlmixr2.github.io/rxode2/authors.html","id":"contributors","dir":"","previous_headings":"","what":"Contributors","title":"Authors","text":"Zufar Mulyukov – Wrote initial version rxShiny() modifications Matthew Fidler Alan Hindmarsh – Lsoda author Awad H. Al-Mohy – Al-Mohy matrix exponential author Ernst Hairer – dop853 author Gerhard Wanner – dop853 author Goro Fuji – Timsort author Hadley Wickham – Author original findLhs RxODE, also original author .s3register (used permission anyone, changed Matthew Fidler) Jack Dongarra – LApack author Linda Petzold – LSODA Martin Maechler – expm author, used routines inductive linearization Morwenn – Timsort author Nicholas J. Higham – Author Al-mohy matrix exponential Roger B. Sidje – expokit matrix exponential author Simon Frost – thread safe C implementation liblsoda Kevin Ushey – Original author fast factor, modified Matthew Filder Yu Feng – thread safe liblsoda Cleve Moler – LApack author David Cooley – Author fast_factor modified now used RxODE quickly create factors IDs without sorting like R Drew Schmidt – Drew Schmidt author edits exponential matrix utility taken R package expm Matt Dowle & Arun Srinivasan – data.table authors openMp framework based upon","code":""},{"path":"https://nlmixr2.github.io/rxode2/authors.html","id":"rxode-acknowledgments","dir":"","previous_headings":"","what":"RxODE acknowledgments:","title":"Authors","text":"Sherwin Sy – Weight based dosing example Justin Wilkins – Documentation updates, logo testing Emma Schwager – R IJK distribution author J Coligne – dop853 fortran author Bill Denney – Documentation updates, manual minor bug fixes Tim Waterhouse – Fixed one bug mac working directories Richard Upton – Helped solving ADVAN linCmt() solutions Dirk Eddelbuettel – Made fixes Rcpp changes require R strict headers Ross Ihaka – R author Robert Gentleman – R author R core team – R authors","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"overview","dir":"","previous_headings":"","what":"Overview","title":"Facilities for Simulating from ODE-Based Models","text":"rxode2 R package solving simulating ode-based models. models convert rxode2 mini-language C create compiled dll fast solving. ODE solving using rxode2 key parts: rxode2() creates C code fast ODE solving based simple syntax related Leibnitz notation. NONMEM deSolve compatible data frame, created et() eventTable() easy simulation events data frame can augmented adding time-varying adding individual covariates (iCov= needed) multiple subject data, may parallelized. single subject output data frame adaptive Covariances metrics uncertanty can used simulate solving","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"installation","dir":"","previous_headings":"","what":"Installation","title":"Facilities for Simulating from ODE-Based Models","text":"can install released version rxode2 CRAN : fastest way install development version rxode2 use r-universe service. service compiles binaries development version MacOS Windows don’t wait package compilation: doesn’t work install development version rxode2 build models rxode2, need working c compiler. use parallel threaded solving rxode2, c compiler needs support open-mp. can check see R working c compiler can check : toolchain, can set described platform information :","code":"install.packages(\"rxode2\") install.packages(c(\"dparser\", \"rxode2ll\", \"rxode2parse\",                    \"rxode2random\", \"rxode2et\", \"rxode2\"),                  repos=c(nlmixr2=\"https://nlmixr2.r-universe.dev\",                          CRAN=\"https://cloud.r-project.org\")) devtools::install_github(\"nlmixr2/rxode2parse\") devtools::install_github(\"nlmixr2/rxode2random\") devtools::install_github(\"nlmixr2/rxode2et\") devtools::install_github(\"nlmixr2/rxode2ll\") devtools::install_github(\"nlmixr2/rxode2\") ## install.packages(\"pkgbuild\") pkgbuild::has_build_tools(debug = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"windows","dir":"","previous_headings":"Installation","what":"Windows","title":"Facilities for Simulating from ODE-Based Models","text":"windows may simply use installr install rtools: Alternatively can download install rtools directly.","code":"install.packages(\"installr\") library(installr) install.rtools()"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"mac-osx","dir":"","previous_headings":"Installation","what":"Mac OSX","title":"Facilities for Simulating from ODE-Based Models","text":"get speed need OpenMP enabled compile rxode2 compiler. various options date discussion likely data.table installation FAQ MacOS. last thing keep mind rxode2 uses code similar original lsoda requires gfortran compiler setup well OpenMP compilers. going using rxode2 nlmixr together older mac computer, suggest trying following: crashes R session binary work Mac machine. able run nlmixr, need compile package manually. proceed assuming homebrew installed system. system terminal need install dependencies compile symengine: installing dependencies, need re-install symengine:","code":"library(symengine) brew install cmake gmp mpfr libmpc install.packages(\"symengine\", type=\"source\") library(symengine)"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"linux","dir":"","previous_headings":"Installation","what":"Linux","title":"Facilities for Simulating from ODE-Based Models","text":"install linux make sure install gcc (openmp support) gfortran using distribution’s package manager. also install system dependencies like udunits symengine dependencies complete installation work linux. also done system packages package manager add appropriate repositories. point r2u project.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"r-versions-40-and-41","dir":"","previous_headings":"Installation","what":"R versions 4.0 and 4.1","title":"Facilities for Simulating from ODE-Based Models","text":"installation R versions 4.0.x 4.1.x, please see instructions install symengine nlmixr2 installation instructions: https://github.com/nlmixr2/nlmixr2#r-package-installation","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"development-version","dir":"","previous_headings":"","what":"Development version","title":"Facilities for Simulating from ODE-Based Models","text":"Since development version rxode2 uses StanHeaders, need make sure compiler setup support C++14, described rstan setup page. R 4.0, believe requires modifying windows toolchain longer (much easier setup). C++ toolchain setup appropriately, can install development version GitHub :","code":"# install.packages(\"devtools\") devtools::install_github(\"nlmixr2/rxode2parse\") devtools::install_github(\"nlmixr2/rxode2random\") devtools::install_github(\"nlmixr2/rxode2et\") devtools::install_github(\"nlmixr2/rxode2ll\") devtools::install_github(\"nlmixr2/rxode2\")"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"illustrated-example","dir":"","previous_headings":"","what":"Illustrated Example","title":"Facilities for Simulating from ODE-Based Models","text":"model equations can specified text string, model file R expression. differential algebraic equations permitted. Differential equations specified d/dt(var_name) =. equation can separated semicolon. load rxode2 package compile model: Model parameters may specified ini({}) model block, initial conditions can specified within model cmt(0)= X, like model eff(0) <- 1. may also specify subject variability initial conditions residual error components just like nlmixr2. allows single interface nlmixr2/rxode2 models. Also note, classic rxode2 interface still works just like past (don’t worry breaking code time). fact, can get classic rxode2 model $simulationModel ui object:","code":"library(rxode2) #> rxode2 2.1.2.9000 using 8 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()`  mod1 <- function() {   ini({     # central      KA=2.94E-01     CL=1.86E+01     V2=4.02E+01     # peripheral     Q=1.05E+01     V3=2.97E+02     # effects     Kin=1     Kout=1     EC50=200    })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <- Q*C2 - Q*C3     eff(0) <- 1     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff   }) } mod1 <- mod1() # create the ui object (can also use `rxode2(mod1)`) mod1  summary(mod1$simulationModel)"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"specify-dosing-and-sampling-in-rxode2","dir":"","previous_headings":"","what":"Specify Dosing and sampling in rxode2","title":"Facilities for Simulating from ODE-Based Models","text":"rxode2 provides simple flexible way specify dosing sampling functions generate event table. First, empty event table generated “et()” function. interface similar NONMEM event tables: can see code, can dose compartment named rxode2 model. slight deviation NONMEM can reduce need compartment renumbering. events can also combined expanded (multi-subject events complex regimens) rbind, c, seq, rep. information creating complex dosing regimens using rxode2 see rxode2 events vignette.","code":"ev  <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(0:240) # Add sampling"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"solving-odes","dir":"","previous_headings":"","what":"Solving ODEs","title":"Facilities for Simulating from ODE-Based Models","text":"ODE can now solved using rxSolve: returns modified data frame. can see compartment values plot :  ,  Note labels automatically labeled units initial event table. rxode2 extracts units label plot (present).","code":"x <- mod1 %>% rxSolve(ev) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> In file included from /usr/share/R/include/R.h:71, #>                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, #>                  from /tmp/RtmpMKFh3q/temp_libpath2b2b3c8645df/rxode2/include/rxode2.h:9, #>                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, #>                  from rx_f5e1c8d22ddcc3e6890ad688a5f37e59_.c:117: #> /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] #>    80 |     }; #>       |      ^ x #> ── Solved rxode2 object ── #> ── Parameters (x$params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions (x$inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows library(ggplot2) plot(x,C2) + ylab(\"Central Concentration\") plot(x,eff)  + ylab(\"Effect\")"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"ode-solving","dir":"","previous_headings":"","what":"ODE solving","title":"Facilities for Simulating from ODE-Based Models","text":"brief comparison pharmacometric ODE solving R packages rxode2. several R packages differential equations. popular deSolve. However pharmacometrics-specific ODE solving, 2 packages rxode2 released CRAN. uses compiled code faster ODE solving. mrgsolve, uses C++ lsoda solver solve ODE systems. user required write hybrid R/C++ code create mrgsolve model translated C++ solving. contrast, rxode2 R-like mini-language parsed C code solves ODE system. Unlike rxode2, mrgsolve currently support symbolic manipulation ODE systems, like automatic Jacobian calculation forward sensitivity calculation (rxode2 currently supports basis nlmixr2’s FOCEi algorithm) dMod, uses unique syntax create “reactions”. reactions create underlying ODEs created c code compiled deSolve model. contrast rxode2 defines ODE systems lower level. rxode2’s parsing mini-language comes C, whereas dMod’s parsing comes R. Like rxode2, dMod supports symbolic manipulation ODE systems calculates forward sensitivities adjoint sensitivities systems. Unlike rxode2, dMod thread-safe since deSolve yet thread-safe. PKPDsim defines models R-like syntax converts system compiled code. Like mrgsolve, PKPDsim currently support symbolic manipulation ODE systems. PKPDsim thread-safe. open pharmacometrics open source community fairly friendly, rxode2 maintainers positive interactions ODE-solving pharmacometric projects listed.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"pk-solved-systems","dir":"","previous_headings":"","what":"PK Solved systems","title":"Facilities for Simulating from ODE-Based Models","text":"rxode2 supports 1-3 compartment models gradients (using stan math’s auto-differentiation). currently uses equations PKADVAN allow time-varying covariates. rxode2 can mix ODEs solved systems.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"the-following-packages-for-solved-pk-systems-are-on-cran","dir":"","previous_headings":"PK Solved systems","what":"The following packages for solved PK systems are on CRAN","title":"Facilities for Simulating from ODE-Based Models","text":"mrgsolve currently 1-2 compartment (poly-exponential models) models built-. solved systems ODEs currently mixed. pmxTools currently 1-3 compartment (super-positioning) models built-. R-implementation. PKPDsim uses 1-3 “ADVAN” solutions using non-superpositioning. PKPDmodels one-compartment model gradients.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"non-cran-libraries","dir":"","previous_headings":"PK Solved systems","what":"Non-CRAN libraries:","title":"Facilities for Simulating from ODE-Based Models","text":"PKADVAN Provides 1-3 compartment models using non-superpositioning. allows time-varying covariates.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":null,"dir":"Reference","previous_headings":"","what":"Turn into an ini block for initialization — as.ini","title":"Turn into an ini block for initialization — as.ini","text":"Turn ini block initialization","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Turn into an ini block for initialization — as.ini","text":"","code":"as.ini(x)  # S3 method for character as.ini(x)  # S3 method for data.frame as.ini(x)  # S3 method for call as.ini(x)  # S3 method for lotriFix as.ini(x)  # S3 method for matrix as.ini(x)  # S3 method for default as.ini(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Turn into an ini block for initialization — as.ini","text":"x Item convert rxode2/nlmixr2 ui ini expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Turn into an ini block for initialization — as.ini","text":"rxode2 ini expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Turn into an ini block for initialization — as.ini","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Turn into an ini block for initialization — as.ini","text":"","code":"ini <- quote(ini({   tka <- log(1.57)   tcl <- log(2.72)   tv <- log(31.5)   eta.ka ~ 0.6   eta.cl ~ 0.3   eta.v ~ 0.1   add.sd <- 0.7 }))  as.ini(ini) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  l <- quote(lotri({   tka <- log(1.57)   tcl <- log(2.72)   tv <- log(31.5)   eta.ka ~ 0.6   eta.cl ~ 0.3   eta.v ~ 0.1   add.sd <- 0.7  }))  as.ini(l) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  m <- lotri({    eta.ka ~ 0.6    eta.cl ~ 0.3    eta.v ~ 0.1 })  as.ini(m) #> ini({ #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  one.compartment <- function() {   ini({     tka <- log(1.57)     tcl <- log(2.72)     tv <- log(31.5)     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  as.ini(one.compartment) #>   #>   #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- c(0, 0.7) #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ui <- one.compartment()  as.ini(ui) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- c(0, 0.7) #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ui$iniDf #>   ntheta neta1 neta2   name lower       est upper   fix label backTransform #> 1      1    NA    NA    tka  -Inf 0.4510756   Inf FALSE  <NA>          <NA> #> 2      2    NA    NA    tcl  -Inf 1.0006319   Inf FALSE  <NA>          <NA> #> 3      3    NA    NA     tv  -Inf 3.4499875   Inf FALSE  <NA>          <NA> #> 4      4    NA    NA add.sd     0 0.7000000   Inf FALSE  <NA>          <NA> #> 5     NA     1     1 eta.ka  -Inf 0.6000000   Inf FALSE  <NA>          <NA> #> 6     NA     2     2 eta.cl  -Inf 0.3000000   Inf FALSE  <NA>          <NA> #> 7     NA     3     3  eta.v  -Inf 0.1000000   Inf FALSE  <NA>          <NA> #>   condition  err #> 1      <NA> <NA> #> 2      <NA> <NA> #> 3      <NA> <NA> #> 4        cp  add #> 5        id <NA> #> 6        id <NA> #> 7        id <NA>  as.ini(ui$iniDf) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- c(0, 0.7) #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ini <- c(\"ini({\",           \"tka <- log(1.57)\",           \"tcl <- log(2.72)\",           \"tv <- log(31.5)\",           \"eta.ka ~ 0.6\",           \"eta.cl ~ 0.3\",           \"eta.v ~ 0.1\",           \"add.sd <- 0.7\",           \"})\")  as.ini(ini) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ini <- paste(ini, collapse=\"\\n\")  as.ini(ini) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":null,"dir":"Reference","previous_headings":"","what":"Turn into a model expression — as.model","title":"Turn into a model expression — as.model","text":"Turn model expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Turn into a model expression — as.model","text":"","code":"as.model(x)  # S3 method for character as.model(x)  # S3 method for call as.model(x)  # S3 method for list as.model(x)  # S3 method for default as.model(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Turn into a model expression — as.model","text":"x item convert model({}) expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Turn into a model expression — as.model","text":"model expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Turn into a model expression — as.model","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Turn into a model expression — as.model","text":"","code":"model <- quote(model({   ka <- exp(tka + eta.ka)   cl <- exp(tcl + eta.cl)   v <- exp(tv + eta.v)   d/dt(depot) = -ka * depot   d/dt(center) = ka * depot - cl / v * center   cp = center / v   cp ~ add(add.sd) }))   as.model(model) #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })  one.compartment <- function() {   ini({     tka <- log(1.57)     tcl <- log(2.72)     tv <- log(31.5)     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  as.model(one.compartment) #>   #>   #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })   ui <- one.compartment()  as.model(ui) #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })  model <- c(\"model({\",            \"ka <- exp(tka + eta.ka)\",            \"cl <- exp(tcl + eta.cl)\",            \"v <- exp(tv + eta.v)\",            \"d/dt(depot) = -ka * depot\",            \"d/dt(center) = ka * depot - cl / v * center\",            \"cp = center / v\",            \"cp ~ add(add.sd)\",            \"})\")  as.model(model) #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })  model <- paste(model, collapse=\"\\n\")  as.model(model) #> model({ #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> })"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":null,"dir":"Reference","previous_headings":"","what":"As rxode2 ui — as.rxUi","title":"As rxode2 ui — as.rxUi","text":"rxode2 ui","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"As rxode2 ui — as.rxUi","text":"","code":"as.rxUi(x)  # S3 method for rxode2 as.rxUi(x)  # S3 method for rxode2tos as.rxUi(x)  # S3 method for rxModelVars as.rxUi(x)  # S3 method for `function` as.rxUi(x)  # S3 method for rxUi as.rxUi(x)  # S3 method for default as.rxUi(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"As rxode2 ui — as.rxUi","text":"x Object convert rxUi object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"As rxode2 ui — as.rxUi","text":"rxUi object (error converted)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"As rxode2 ui — as.rxUi","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"As rxode2 ui — as.rxUi","text":"","code":"mod1 <- function() {  ini({    # central     KA=2.94E-01    CL=1.86E+01    V2=4.02E+01    # peripheral    Q=1.05E+01    V3=2.97E+02    # effects    Kin=1    Kout=1    EC50=200   })  model({    C2 <- centr/V2    C3 <- peri/V3    d/dt(depot) <- -KA*depot    d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3    d/dt(peri)  <- Q*C2 - Q*C3    eff(0) <- 1    d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff  }) }  as.rxUi(mod1) #>   #>   #>  ── rxode2-based free-form 4-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2            centr #> 3                  3             peri #> 4                  4              eff #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         KA <- 0.294 #>         CL <- 18.6 #>         V2 <- 40.2 #>         Q <- 10.5 #>         V3 <- 297 #>         Kin <- 1 #>         Kout <- 1 #>         EC50 <- 200 #>     }) #>     model({ #>         C2 <- centr/V2 #>         C3 <- peri/V3 #>         d/dt(depot) <- -KA * depot #>         d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3 #>         d/dt(peri) <- Q * C2 - Q * C3 #>         eff(0) <- 1 #>         d/dt(eff) <- Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":null,"dir":"Reference","previous_headings":"","what":"Assert properties of the rxUi models — assertRxUi","title":"Assert properties of the rxUi models — assertRxUi","text":"Assert properties rxUi models","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assert properties of the rxUi models — assertRxUi","text":"","code":"assertRxUi(model, extra = \"\", .var.name = .vname(model))  assertRxUiPrediction(model, extra = \"\", .var.name = .vname(model))  assertRxUiSingleEndpoint(model, extra = \"\", .var.name = .vname(model))  assertRxUiTransformNormal(model, extra = \"\", .var.name = .vname(model))  assertRxUiNormal(model, extra = \"\", .var.name = .vname(model))  assertRxUiMuRefOnly(model, extra = \"\", .var.name = .vname(model))  assertRxUiEstimatedResiduals(model, extra = \"\", .var.name = .vname(model))  assertRxUiPopulationOnly(model, extra = \"\", .var.name = .vname(model))  assertRxUiMixedOnly(model, extra = \"\", .var.name = .vname(model))  assertRxUiRandomOnIdOnly(model, extra = \"\", .var.name = .vname(model))"},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assert properties of the rxUi models — assertRxUi","text":"model Model check extra Extra text append error message (like \"focei\") .var.name [character(1)] Name checked object print assertions. Defaults heuristic implemented vname.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assert properties of the rxUi models — assertRxUi","text":"rxUi model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Assert properties of the rxUi models — assertRxUi","text":"functions different types assertions assertRxUi -- Make sure proper rxode2 model (throw error) assertRxUiSingleEndpoint -- Make sure rxode2 model single endpoint model (throw error) assertRxUiTransformNormal -- Make sure model residual distribution normal transformably normal assertRxUiNormal -- Make sure model residual distribution normal assertRxUiEstimatedResiduals -- Make sure residual error parameters estimated (modeled). assertRxUiPopulationOnly -- Make sure model population model (mixed effects) assertRxUiMixedOnly -- Make sure model mixed effect model (population effect, ) assertRxUiPrediction -- Make sure model predictions assertRxUiMuRefOnly -- Make sure parameters mu-referenced assertRxUiRandomOnIdOnly -- Make sure random effects ID level","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assert properties of the rxUi models — assertRxUi","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Assert properties of the rxUi models — assertRxUi","text":"","code":"# \\donttest{ one.cmt <- function() {  ini({    tka <- 0.45; label(\"Ka\")    tcl <- log(c(0, 2.7, 100)); label(\"Cl\")    tv <- 3.45; label(\"V\")    eta.ka ~ 0.6    eta.cl ~ 0.3    eta.v ~ 0.1    add.sd <- 0.7  })  model({    ka <- exp(tka + eta.ka)    cl <- exp(tcl + eta.cl)    v <- exp(tv + eta.v)    linCmt() ~ add(add.sd)  }) }  assertRxUi(one.cmt) #>   #>   # assertRxUi(rnorm) # will fail  assertRxUiSingleEndpoint(one.cmt) #>   #>   # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"meant perform way quantile() can drop replacement code using quantile() using distributional assumptions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"","code":"binomProbs(x, ...)  # S3 method for default binomProbs(   x,   probs = c(0.025, 0.05, 0.5, 0.95, 0.975),   na.rm = FALSE,   names = TRUE,   onlyProbs = TRUE,   n = 0L,   m = 0L,   pred = FALSE,   piMethod = c(\"lim\"),   M = 5e+05,   tol = .Machine$double.eps^0.25,   ciMethod = c(\"wilson\", \"wilsonCorrect\", \"agrestiCoull\", \"wald\", \"wc\", \"ac\"),   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"x numeric vector whose mean probability based confidence values wanted, NA NaN values allowed numeric vectors unless na.rm TRUE. ... Arguments passed default method, allows many different methods applied. probs numeric vector probabilities values interval 0 1, inclusive. 0, represents maximum observed, 1, represents maximum observed. 0.5 represents expected probability (mean). na.rm logical; true, NA NaN's removed x quantiles computed. names logical; true, result names attribute. onlyProbs logical; true, return probability based confidence interval/prediction interval estimates, otherwise return extra statistics. n integer/integerish; n used calculate prediction confidence interval.  n=0 (default) use number non-NA observations.  calculating prediction interval, represents number observations used input (\"true\") distribution. m integer.  using prediction interval represents number samples observed future prediction interval. pred Use prediction interval instead confidence interval.  default FALSE. piMethod gives prediction interval method (currently lim) Lu 2020 M number simulations run LIM PI. tol tolerance root finding LIM prediction interval ciMethod gives method calculating confidence interval. Can : \"argestiCoull\" \"ac\" -- Agresti-Coull method. 95\\ interval, method use concept   \"adding 2 successes 2 failures,\" rather uses formulas explicitly described following link: https://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval#Agresti-Coull_Interval. \"wilson\" -- Wilson Method \"wilsonCorrect\" \"wc\" -- Wilson method continuity correction \"wald\" -- Wald confidence interval standard z approximation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"default return probabilities names (named) points expected distribution located given sampling mean standard deviation. onlyProbs=FALSE prepend mean, variance, standard deviation, minimum, maximum number non-NA observations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"used confidence intervals rxode2 solved objects using confint(mean=\"binom\")","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"Newcombe, R. G. (1998). \"Two-sided confidence intervals single proportion: comparison seven methods\". Statistics Medicine. 17 (8): 857–872. doi:10.1002/(SICI)1097-0258(19980430)17:8<857::AID-SIM777>3.0.CO;2-E. PMID 9595616. Hezhi Lu, Hua Jin, new prediction interval binomial random variable based inferential models, Journal Statistical Planning Inference, Volume 205, 2020, Pages 156-174, ISSN 0378-3758, https://doi.org/10.1016/j.jspi.2019.07.001.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"","code":"x<- rbinom(7001, p=0.375, size=1) binomProbs(x) #>      2.5%        5%       50%       95%     97.5%  #> 0.3716257 0.3734362 0.3829453 0.3925448 0.3943933   # you can also use the prediction interval # \\donttest{ binomProbs(x, pred=TRUE) #>      2.5%        5%       50%       95%     97.5%  #> 0.3668047 0.3695186 0.3829453 0.3965148 0.3990858  # }  # Can get some extra statistics if you request onlyProbs=FALSE binomProbs(x, onlyProbs=FALSE) #>         mean          var           sd            n         2.5%           5%  #>    0.3829453    0.2362982    0.4861051 7001.0000000    0.3716257    0.3734362  #>          50%          95%        97.5%  #>    0.3829453    0.3925448    0.3943933   x[2] <- NA_real_  binomProbs(x, onlyProbs=FALSE) #>  mean   var    sd     n  2.5%    5%   50%   95% 97.5%  #>    NA    NA    NA    NA    NA    NA    NA    NA    NA   binomProbs(x, na.rm=TRUE) #>      2.5%        5%       50%       95%     97.5%  #> 0.3715373 0.3733478 0.3828571 0.3924570 0.3943055"},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the rxode2 coefficients — coef.rxode2","title":"Return the rxode2 coefficients — coef.rxode2","text":"returns parameters , state variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the rxode2 coefficients — coef.rxode2","text":"","code":"# S3 method for rxode2 coef(object, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the rxode2 coefficients — coef.rxode2","text":"object rxode2 object ... ignored arguments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the rxode2 coefficients — coef.rxode2","text":"rxCoef object following params  list strings parameters rxode2 object state list strings names state rxode2 object. ini model specified default values parameters. rxode2 referring rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the rxode2 coefficients — coef.rxode2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":null,"dir":"Reference","previous_headings":"","what":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"returns new expression, old expression, new variable old variable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"","code":".assertRenameErrorModelLine(line, vars)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"line quoted call information line vars Variables contained within mdel","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"list(new, old, newChar, oldChar)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":null,"dir":"Reference","previous_headings":"","what":"This copies the rxode2 UI object so it can be modified — .copyUi","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"copies rxode2 UI object can modified","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"","code":".copyUi(ui)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"ui Original UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"Copied UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":null,"dir":"Reference","previous_headings":"","what":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"Handle single error normal t distributions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"","code":".handleSingleErrTypeNormOrTFoceiBase(   env,   pred1,   errNum = 1L,   rxPredLlik = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"env Environment parsed model pred1 data.frame current error errNum number error specification nlmixr2 model rxPredLlik boolean indicating log likelihood calculated non-normal distributions.  default TRUE.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"list lines added.  lines contain rx_yj_ integer corresponds transformation type. rx_lambda_ transformation lambda rx_low_ lower boundary transformation rx_hi_ upper boundary transformation rx_pred_f_ prediction function rx_pred_ transformed prediction function rx_r_ transformed variance","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":null,"dir":"Reference","previous_headings":"","what":"This gets the append arg for the ini() piping — .iniGetAppendArg","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"gets append arg ini() piping","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"","code":".iniGetAppendArg(f, s)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"f try(force(append)) argument, s .character(substitute(append)) argument","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"corrected ini piping argument exported creating new ini methods requirements piping","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":null,"dir":"Reference","previous_headings":"","what":"Reorder rows in iniDf — .iniHandleAppend","title":"Reorder rows in iniDf — .iniHandleAppend","text":"Reorder rows iniDf","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reorder rows in iniDf — .iniHandleAppend","text":"","code":".iniHandleAppend(expr, rxui, envir, append)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reorder rows in iniDf — .iniHandleAppend","text":"expr Expression parsing rxui User interface function envir Environment parsing append Reorder theta parameters.  NULL means change parameter order.  parameter name (character string) means put new parameter named parameter.  number less equal zero means put parameter beginning list.  number greater last parameter number means put parameter end list.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Reorder rows in iniDf — .iniHandleAppend","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":null,"dir":"Reference","previous_headings":"","what":"Update the iniDf of a model — .iniHandleLine","title":"Update the iniDf of a model — .iniHandleLine","text":"Update iniDf model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update the iniDf of a model — .iniHandleLine","text":"","code":".iniHandleLine(expr, rxui, envir = parent.frame(), append = NULL)  .iniHandleFixOrUnfix(expr, rxui, envir = parent.frame(), append = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update the iniDf of a model — .iniHandleLine","text":"expr Expression parsing rxui User interface function envir Environment parsing append Reorder theta parameters.  NULL means change parameter order.  parameter name (character string) means put new parameter named parameter.  number less equal zero means put parameter beginning list.  number greater last parameter number means put parameter end list.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update the iniDf of a model — .iniHandleLine","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Update the iniDf of a model — .iniHandleLine","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":null,"dir":"Reference","previous_headings":"","what":"Check if a language object matches a template language object — .matchesLangTemplate","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"template == str2lang(\".\"), match anything. template == str2lang(\".name\"), match name. template == str2lang(\".call()\"), match call.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"","code":".matchesLangTemplate(x, template)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"x object check template template object match","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"TRUE matches, FALSE, otherwise","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"","code":".matchesLangTemplate(str2lang(\"d/dt(foo)\"), str2lang(\"d/dt(.name)\")) #> [1] TRUE .matchesLangTemplate(str2lang(\"d/dt(foo)\"), str2lang(\"d/foo(.name)\")) #> [1] FALSE .matchesLangTemplate(str2lang(\"d/dt(foo)\"), str2lang(\"d/.\")) #> [1] TRUE"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal messaging statements — .minfo","title":"Internal messaging statements — .minfo","text":"Internal messaging statements","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal messaging statements — .minfo","text":"","code":".minfo(text, ..., .envir = parent.frame())  .malert(text, ..., .envir = parent.frame())  .mwarn(text, ..., .envir = parent.frame())  .msuccess(text, ..., .envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal messaging statements — .minfo","text":"text Text ... arguments .envir Environment evaluate ","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal messaging statements — .minfo","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal messaging statements — .minfo","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":null,"dir":"Reference","previous_headings":"","what":"Handle model lines — .modelHandleModelLines","title":"Handle model lines — .modelHandleModelLines","text":"Handle model lines","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Handle model lines — .modelHandleModelLines","text":"","code":".modelHandleModelLines(   modelLines,   rxui,   modifyIni = FALSE,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir )"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Handle model lines — .modelHandleModelLines","text":"modelLines model lines considered rxui rxode2 UI object modifyIni ini() considered append boolean determine lines appended piping.  possible values : TRUE lines appended model instead replaced FALSE lines replaced model (default) NA lines pre-pended model instead replaced lhs expression, append lines last observed line expression lhs auto boolean tells piping automatically selects parameters characterized population parameter, subject variability, covariate.  TRUE automatic selection occurs.  FALSE automatic selection turned everything added covariate (can promoted parameter ini statement). default TRUE, can changed options(rxode2.autoVarPiping=FALSE). cov character vector variables assumed covariates.  override automatic promotion population parameter estimate (eta) envir Environment evaluation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Handle model lines — .modelHandleModelLines","text":"New UI","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Handle model lines — .modelHandleModelLines","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":null,"dir":"Reference","previous_headings":"","what":"Returns quoted call information — .quoteCallInfoLines","title":"Returns quoted call information — .quoteCallInfoLines","text":"Returns quoted call information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Returns quoted call information — .quoteCallInfoLines","text":"","code":".quoteCallInfoLines(callInfo, envir = parent.frame(), iniDf = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Returns quoted call information — .quoteCallInfoLines","text":"callInfo Call information envir Environment evaluation (needed) iniDf parent model iniDf piping ini block (NULL otherwise)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Returns quoted call information — .quoteCallInfoLines","text":"Quote call information.  name=expression, change name<-expression quoted call list. expressions within brackets ie {}, unlist brackets called one single sequence.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Returns quoted call information — .quoteCallInfoLines","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":null,"dir":"Reference","previous_headings":"","what":"Print out a table in the documentation — .rxDocTable","title":"Print out a table in the documentation — .rxDocTable","text":"Print table documentation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print out a table in the documentation — .rxDocTable","text":"","code":".rxDocTable(table, caption = \"none\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print out a table in the documentation — .rxDocTable","text":"table data frame caption character vector representing caption latex table","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print out a table in the documentation — .rxDocTable","text":"based knitr context: output kableExtra::kbl latex output output DT::datatable html output otherwise output knitr::kable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print out a table in the documentation — .rxDocTable","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Print out a table in the documentation — .rxDocTable","text":"","code":".rxDocTable(rxReservedKeywords) #>  #>  #> |Reserved Name  |Meaning                                                                                                                                                                           |Alias  | #> |:--------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------| #> |time           |solver time                                                                                                                                                                       |t      | #> |podo           |In Transit compartment models, last dose amount                                                                                                                                   |       | #> |tlast          |Time of Last dose                                                                                                                                                                 |       | #> |M_E            |Exp(1)                                                                                                                                                                            |       | #> |M_LOG2E        |log2(e)                                                                                                                                                                           |       | #> |M_LOG10E       |log10(e)                                                                                                                                                                          |       | #> |M_LN2          |log(2)                                                                                                                                                                            |       | #> |M_LN10         |log(10)                                                                                                                                                                           |       | #> |M_PI           |pi                                                                                                                                                                                |       | #> |M_PI_2         |pi/2                                                                                                                                                                              |       | #> |M_PI_4         |pi/4                                                                                                                                                                              |       | #> |M_1_PI         |1/pi                                                                                                                                                                              |       | #> |M_2_PI         |2/pi                                                                                                                                                                              |       | #> |M_2_SQRTPI     |2/sqrt(pi)                                                                                                                                                                        |       | #> |M_SQRT2        |sqrt(2)                                                                                                                                                                           |       | #> |M_SQRT1_2      |1/sqrt(2)                                                                                                                                                                         |       | #> |M_SQRT_3       |sqrt(3)                                                                                                                                                                           |       | #> |M_SQRT_32      |sqrt(32)                                                                                                                                                                          |       | #> |M_LOG10_2      |Log10(2)                                                                                                                                                                          |       | #> |M_2PI          |2*pi                                                                                                                                                                              |       | #> |M_SQRT_PI      |sqrt(pi)                                                                                                                                                                          |       | #> |M_1_SQRT_2PI   |1/(sqrt(2*pi))                                                                                                                                                                    |       | #> |M_LN_SQRT_PI   |log(sqrt(pi))                                                                                                                                                                     |       | #> |M_LN_SQRT_2PI  |log(sqrt(2*pi))                                                                                                                                                                   |       | #> |M_LN_SQRT_PId2 |log(sqrt(pi/2))                                                                                                                                                                   |       | #> |pi             |pi                                                                                                                                                                                |       | #> |NA             |R's NA value                                                                                                                                                                      |       | #> |NaN            |Not a Number Value                                                                                                                                                                |       | #> |Inf            |Infinite Value                                                                                                                                                                    |       | #> |newind         |1: First record of individual; 2: Subsequent record of individual                                                                                                                 |NEWIND | #> |rxFlag         |Flag for what part of the rxode2 model is being run; 1: ddt; 2: jac; 3: ini; 4: F; 5: lag; 6: rate; 7: dur; 8: mtime; 9: matrix exponential; 10: inductive linearization; 11: lhs |       |"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"Get upper boundary condition transformation ","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"","code":".rxGetHiBoundaryPred1AndIni(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"env Environment parsed model pred1 data.frame current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"Upper Boundary","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"Get lambda value based pred information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"","code":".rxGetLambdaFromPred1AndIni(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"env Environment environment pred1 Single error data frame","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"Lambda expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"Get lower boundary condition transformation requires ","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"","code":".rxGetLowBoundaryPred1AndIni(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"env Environment parsed model pred1 data.frame current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"Lower Boundary","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the Observation transformation — .rxGetPredictionDVTransform","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"Get Observation transformation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"","code":".rxGetPredictionDVTransform(env, pred1, yj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"env Environment parsed model pred1 data.frame current error yj transformation number current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"transformation expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the prediction name — .rxGetPredictionF","title":"Get the prediction name — .rxGetPredictionF","text":"Get prediction name","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the prediction name — .rxGetPredictionF","text":"","code":".rxGetPredictionF(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the prediction name — .rxGetPredictionF","text":"env Environment parsed model pred1 data.frame current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the prediction name — .rxGetPredictionF","text":"prediction symbol","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the prediction name — .rxGetPredictionF","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the prediction transformation — .rxGetPredictionFTransform","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"Get prediction transformation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"","code":".rxGetPredictionFTransform(env, pred1, yj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"env Environment parsed model pred1 data.frame current error yj transformation number current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"transformation expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":null,"dir":"Reference","previous_headings":"","what":"Get Variance for error type — .rxGetVarianceForErrorType","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"Get Variance error type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"","code":".rxGetVarianceForErrorType(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"env Environment pred1 Pred one end-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"Variance error type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"Internal function tell linCmt() model variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"","code":".rxIsLinCmt()"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"0 1","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function for calculating the Jacobian — .rxJacobian","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"Internal function calculating Jacobian","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"","code":".rxJacobian(model, vars = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"model symengine environment vars Variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"Jacobian information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"Internal function generate model variables linCmt() model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"","code":".rxLinCmtGen(lenState, vars)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"lenState Length state vars Variables model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"Model variables expanded linCmt model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal Pruning function — .rxPrune","title":"Internal Pruning function — .rxPrune","text":"Internal Pruning function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal Pruning function — .rxPrune","text":"","code":".rxPrune(x, envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal Pruning function — .rxPrune","text":"x List quoted lines envir Environment information stored","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal Pruning function — .rxPrune","text":"Pruned model code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal Pruning function — .rxPrune","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":null,"dir":"Reference","previous_headings":"","what":"Sensitivity for model — .rxSens","title":"Sensitivity for model — .rxSens","text":"Sensitivity model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Sensitivity for model — .rxSens","text":"","code":".rxSens(model, vars, vars2)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Sensitivity for model — .rxSens","text":"model symengine model environment vars Variables single sensitivity vars2 present, 2 parameter sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Sensitivity for model — .rxSens","text":"Sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Sensitivity for model — .rxSens","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":null,"dir":"Reference","previous_headings":"","what":"Temporarily set options then restore them while running code — .rxWithOptions","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"Temporarily set options restore running code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"","code":".rxWithOptions(ops, code)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"ops list options temporarily set code code code run sink","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"value code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"","code":".rxWithOptions(list(digits = 21), {   print(pi) }) #> [1] 3.141592653589793116  print(pi) #> [1] 3.141593"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":null,"dir":"Reference","previous_headings":"","what":"With one sink, then release — .rxWithSink","title":"With one sink, then release — .rxWithSink","text":"one sink, release","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"With one sink, then release — .rxWithSink","text":"","code":".rxWithSink(file, code)  .rxWithSinkBoth(file, code)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"With one sink, then release — .rxWithSink","text":"file path file sink running code code code run sink","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"With one sink, then release — .rxWithSink","text":"return results code section","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"With one sink, then release — .rxWithSink","text":".rxWithSink captures output cat .rxWithSinkBoth captures output cat message","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"With one sink, then release — .rxWithSink","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"With one sink, then release — .rxWithSink","text":"","code":"t <- tempfile() .rxWithSink(t, cat(\"message\\n\")) cat(\"cat2\\n\") # now you can see the cat2 #> cat2 lines <- readLines(t) unlink(t)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":null,"dir":"Reference","previous_headings":"","what":"Temporarily set options then restore them while running code — .rxWithWd","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"Temporarily set options restore running code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"","code":".rxWithWd(wd, code)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"wd working directory temporarily set system evaluating code code code run sink","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"value code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"","code":".rxWithWd(tempdir(), {   getwd() }) #> [1] \"/tmp/RtmpSPRppR\"  getwd() #> [1] \"/home/runner/work/rxode2/rxode2/docs/reference\""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-useUtf.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to figure out if this session supports Unicode — .useUtf","title":"Internal function to figure out if this session supports Unicode — .useUtf","text":"Internal function figure session supports Unicode","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-useUtf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to figure out if this session supports Unicode — .useUtf","text":"","code":".useUtf()"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-useUtf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function to figure out if this session supports Unicode — .useUtf","text":"boolean indicating session supports Unicode","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":null,"dir":"Reference","previous_headings":"","what":"Error function — erf","title":"Error function — erf","text":"Error function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Error function — erf","text":"","code":"erf(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Error function — erf","text":"x vector real values","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Error function — erf","text":"erf x","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Error function — erf","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Error function — erf","text":"","code":"erf(1.0) #> [1] 0.8427008"},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":null,"dir":"Reference","previous_headings":"","what":"Event translation for rxode2 — etTrans","title":"Event translation for rxode2 — etTrans","text":"Event translation rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Event translation for rxode2 — etTrans","text":"","code":"etTrans(   inData,   obj,   addCmt = FALSE,   dropUnits = FALSE,   allTimeVar = FALSE,   keepDosingOnly = FALSE,   combineDvid = NULL,   keep = character(0),   addlKeepsCov = FALSE,   addlDropSs = TRUE,   ssAtDoseTime = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Event translation for rxode2 — etTrans","text":"inData Data frame translate obj Model translate data addCmt Add compartment data frame (default FALSE). dropUnits Boolean drop units (default FALSE). allTimeVar Treat covariates time-varying keepDosingOnly keep individuals dosing records trailing dosing records last observation. combineDvid boolean indicating rxode2 use DVID observation records change cmt value; Useful multiple-endpoint nlmixr models.  default determined option(\"rxode2.combine.dvid\") option set, TRUE. typically affect rxode2 simulations. keep named vector items want keep final rxode2 dataset. added rxode2 event records (seen), last observation carried forward used. addlKeepsCov determines additional dosing items repeats dose (FALSE) keeps covariates record dose (TRUE) addlDropSs steady state doses addl specification steady state flag dropped repeated doses (TRUE) retained (FALSE) ssAtDoseTime Boolean TRUE back calculates steady concentration actual time dose, otherwise FALSE doses shifted","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Event translation for rxode2 — etTrans","text":"Object solving rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":null,"dir":"Reference","previous_headings":"","what":"Find the assignments in R expression — findLhs","title":"Find the assignments in R expression — findLhs","text":"Find assignments R expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Find the assignments in R expression — findLhs","text":"","code":"findLhs(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Find the assignments in R expression — findLhs","text":"x R expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Find the assignments in R expression — findLhs","text":"list assigned parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Find the assignments in R expression — findLhs","text":"Hadley Wickham Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":null,"dir":"Reference","previous_headings":"","what":"Gammap: normalized lower incomplete gamma function — gammap","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"gamma_p boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"","code":"gammap(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"numeric '' parameter normalized lower incomplete gamma z numeric 'z' parameter normalized lower incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"gammap results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"gamma p function given : gammap = lowergamma(, z)/gamma()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"","code":"gammap(1, 3) #> [1] 0.9502129 gammap(1:3, 3) #> [1] 0.9502129 0.8008517 0.5768099 gammap(1, 1:3) #> [1] 0.6321206 0.8646647 0.9502129"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":null,"dir":"Reference","previous_headings":"","what":"gammapDer: derivative of gammap — gammapDer","title":"gammapDer: derivative of gammap — gammapDer","text":"gamma_p_derivative boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"gammapDer: derivative of gammap — gammapDer","text":"","code":"gammapDer(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"gammapDer: derivative of gammap — gammapDer","text":"numeric '' parameter upper incomplete gamma z numeric 'z' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"gammapDer: derivative of gammap — gammapDer","text":"lowergamma results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gammapDer: derivative of gammap — gammapDer","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"gammapDer: derivative of gammap — gammapDer","text":"","code":"gammapDer(1:3, 3) #> [1] 0.04978707 0.14936121 0.22404181  gammapDer(1, 1:3) #> [1] 0.36787944 0.13533528 0.04978707"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":null,"dir":"Reference","previous_headings":"","what":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"gammapInv gammapInva:  Inverses normalized gammap function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"","code":"gammapInv(a, p)  gammapInva(x, p)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"numeric '' parameter upper incomplete gamma p numeric 'p' parameter upper incomplete gamma x numeric 'x' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"inverse gammap results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"equation: p = gammap(, x) 'gammapInv' function returns value 'x' satisfies equation 'gammapInva' function returns value 'q' satisfies equation NOTE: gammapInva slow","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"","code":"gammapInv(1:3, 0.5) #> [1] 0.6931472 1.6783470 2.6740603  gammapInv(1, 1:3 / 3.1) #> [1] 0.3894648 1.0360919 3.4339872  gammapInv(1:3, 1:3 / 3.1) #> [1] 0.3894648 2.1988984 6.8872929  gammapInva(1:3, 1:3 / 3.1) #> [1] 0.3894648 2.1988984 6.8872929"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":null,"dir":"Reference","previous_headings":"","what":"Gammaq: normalized upper incomplete gamma function — gammaq","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"gamma_q boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"","code":"gammaq(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"numeric '' parameter normalized upper incomplete gamma z numeric 'z' parameter normalized upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"gammaq results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"gamma q function given : gammaq = uppergamma(, z)/gamma()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"","code":"gammaq(1, 3) #> [1] 0.04978707 gammaq(1:3, 3) #> [1] 0.04978707 0.19914827 0.42319008 gammaq(1, 1:3) #> [1] 0.36787944 0.13533528 0.04978707"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":null,"dir":"Reference","previous_headings":"","what":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"gammaqInv gammaqInva:  Inverses normalized gammaq function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"","code":"gammaqInv(a, q)  gammaqInva(x, q)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"numeric '' parameter upper incomplete gamma q numeric 'q' parameter upper incomplete gamma x numeric 'x' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"inverse gammaq results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"equation: q = gammaq(, x) 'gammaqInv' function returns value 'x' satisfies equation 'gammaqInva' function returns value '' satisfies equation NOTE: gammaqInva slow","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"","code":"gammaqInv(1:3, 0.5) #> [1] 0.6931472 1.6783470 2.6740603  gammaqInv(1, 1:3 / 3) #> [1] 1.0986123 0.4054651 0.0000000  gammaqInv(1:3, 1:3 / 3.1) #> [1] 1.131402 1.248533 0.684289  gammaqInva(1:3, 1:3 / 3.1) #> [1] 1.131402 1.248533 0.684289"},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"Create complete shiny application exploring dosing regimens given (hardcoded) PK/PD model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"","code":"genShinyApp.template(   appDir = \"shinyExample\",   verbose = TRUE,   ODE.config = list(ode = \"model\", params = c(KA = 0.294), inits = c(eff = 1), method =     \"lsoda\", atol = 1e-08, rtol = 1e-06) )  write.template.server(appDir)  write.template.ui(appDir, statevars)"},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"appDir string directory store shiny app, default \"shinyExample\". directory appDir created exist. verbose logical specifying whether write messages shiny app generated. Defaults TRUE. ODE.config model name compiled list parameters sent rxSolve(). statevars List statevars passed write.template.ui() function.  usually called directly. PK/PD model defined using rxode2(), set parameters initial values defined.  appropriate R scripts shiny's user interface ui.R server logic server.R created directory appDir. function evaluates following PK/PD model default: can changed ODE.config parameter. launch shiny app, simply issue runApp(appDir) R command.","code":"C2 = centr/V2;     C3 = peri/V3;     d/dt(depot) =-KA*depot;     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  =                    Q*C2 - Q*C3;     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff;"},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"None, functions used side effects.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"functions create simple, working example dosing regimen simulation web application. Users may want modify code experiment creating shiny applications specific rxode2 models.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"","code":"# \\donttest{ # remove myapp when the example is complete on.exit(unlink(\"myapp\", recursive = TRUE, force = TRUE)) # create the shiny app example (template) genShinyApp.template(appDir = \"myapp\") #>  #> Generating an example (template) for a dosing regimen shiny app #>  #> Using the following PK/PD model: #>    C2 = centr/V2; #>    C3 = peri/V3; #>    d/dt(depot) =-KA*depot; #>    d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3; #>    d/dt(peri)  =                    Q*C2 - Q*C3; #>    d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff; #>      #> Translating the PK/PD ODE model into C, compiling, etc. #>  #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #>  #> Parameters and their values: #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #>  #> Initial values in each compartment: #> depot centr  pari   eff  #>     0     0     0     1  #>  #> Warning: 2 arguments not used by format ' #>       # #>       # Dosing regimen template generated by rxode2::genShinyApp.template() #>       # #>  #>       debug = TRUE #>       #wd = sprintf(\"%%s/../\", getwd()) #>       #setwd(wd) #>  #>       # Server inputs: Dose, dosing regimen, dosing frequency, #>       # dosing cycle definition, number of dosing cycles #>  #>       library(shiny) #>       library(rxode2) #>  #>       # read objects from \"rx_shiny_data.rda\" in the  AppDir folder, #>       # objects include, mod1, params, inits, method, atol, rtol.] #>  #>       load(\"./rx_shiny_data.rda\") #>       if (!rxDynLoad(mod1)) mod1 <- rxode2(mod1, modName=\"mod1\") #>       # Define server logic #>       shinyServer(function(input, output) { #>  #>         get.cp <- reactive({ #>           ds <- input$Dose #>           reg <- switch(input$regimen, \"QD\"=1, \"BID\"=2) #>           cyc <- switch(input$cycle, #>               \"continous\"=c(7,0), #>               \"1wkon 1wkoff\"=c(7,7), #>               \"2wkon 1wkoff\"=c(14,7), #>               \"3wkon 1wkoff\"=c(21,7) #>           ) #>           cyc <- [... truncated] #> Shiny files (ui.R, server.R) plus R data saved. #>  #> To launch the Shiny app, type the following two R commands: #>  #> \tlibrary(shiny) #> \trunApp(\"myapp\") #>  # run the shiny app if (requireNamespace(\"shiny\", quietly=TRUE)) {   library(shiny)   # runApp(\"myapp\") # Won't launch in environments without browsers } # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"Get base simulation model simulation inis underlying rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"","code":"getBaseIniSimModel(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"obj Fit Object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"Simulation object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the base simulation model for simulation — getBaseSimModel","title":"Get the base simulation model for simulation — getBaseSimModel","text":"Get base simulation model simulation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the base simulation model for simulation — getBaseSimModel","text":"","code":"getBaseSimModel(obj)  # S3 method for default getBaseSimModel(obj)  # S3 method for default getBaseIniSimModel(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the base simulation model for simulation — getBaseSimModel","text":"obj Fit Object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the base simulation model for simulation — getBaseSimModel","text":"Simulation object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the base simulation model for simulation — getBaseSimModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"Get symengine loading symengine rxS()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"","code":"getBaseSymengineModel(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"obj Object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"Simulation model ready load symeng","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":null,"dir":"Reference","previous_headings":"","what":"Get/Set the number of threads that rxode2 uses — getRxThreads","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"Get/Set number threads rxode2 uses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"","code":"getRxThreads(verbose = FALSE)  setRxThreads(threads = NULL, percent = NULL, throttle = NULL)  rxCores(verbose = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"verbose Display value relevant OpenMP settings threads NULL (default) rereads environment variables. 0 means use logical CPUs available. Otherwise number >= 1 percent provided number 2 100; percentage logical CPUs use. default startup, 50 percent. throttle 2 (default) means , roughly speaking, single thread used number subjects solved <=2, 2 threads number points <=4, etc. throttle speed small data tasks (especially repeated many times) incurring overhead managing multiple threads. throttle also suppress sorting ID solved first (nsubject solved)*throttle <= nthreads.  rxode2 sorting occurs minimize time waiting another thread finish. last item solved long solving time, solving wait last costly solving occur. items likely take time solved first, wait less likely impact overall solving time. rxode2 IDs sorted individual number solving points (largest first). also C interface allows IDs resorted total time spent solving equation.  allows packages like nlmixr sort solving time needed. Overall number threads throttled (restricted) small tasks sorting IDs suppressed.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"number threads rxode2 uses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":null,"dir":"Reference","previous_headings":"","what":"Empty Guide — guide_none","title":"Empty Guide — guide_none","text":"empty guide draws nothing; included rxode2 compatibility ggplot 3.2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Empty Guide — guide_none","text":"","code":"guide_none(title = waiver(), position = waiver())"},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Empty Guide — guide_none","text":"title character string expression indicating title guide. NULL, title shown. default (waiver()), name scale object name specified labs() used title. position guide drawn: one top, bottom, left, right.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Empty Guide — guide_none","text":"nothing, simply included compatible ggplot 3.2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign the ini block in the rxode2 related object — ini<-","title":"Assign the ini block in the rxode2 related object — ini<-","text":"Assign ini block rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign the ini block in the rxode2 related object — ini<-","text":"","code":"ini(x, envir = environment(x)) <- value"},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign the ini block in the rxode2 related object — ini<-","text":"x rxode2 related object envir Environment assignment occurs value Value object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign the ini block in the rxode2 related object — ini<-","text":"rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign the ini block in the rxode2 related object — ini<-","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":null,"dir":"Reference","previous_headings":"","what":"Ini block for rxode2/nlmixr models — ini.rxUi","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"ini block controls initial conditions 'theta' (fixed effects), 'omega' (random effects), 'sigma' (residual error) elements model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"","code":"# S3 method for rxUi ini(x, ..., envir = parent.frame(), append = NULL)  # S3 method for default ini(x, ..., envir = parent.frame(), append = NULL)  ini(x, ..., envir = parent.frame(), append = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"x expression ... expressions ini() function envir environment unevaluated model expressions evaluated.  May also NULL, list, data frame, pairlist integer specified sys.call. append Reorder theta parameters.  NULL means change parameter order.  parameter name (character string) means put new parameter named parameter.  number less equal zero means put parameter beginning list.  number greater last parameter number means put parameter end list.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"ini block","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"ini() function used two different ways.  main way used set initial conditions associated attributes (described ) model.  way used updating initial conditions model, often using pipe operator. 'theta' 'sigma' can set using either <- = tvCL <- 1 equivalently tvCL = 1.  'omega' can set ~ etaCL ~ 0.1. Parameters can named unnamed (though named parameters preferred). named parameter set using name left assignment unnamed parameters set without assignment operator.  tvCL <- 1 set named parameter tvCL 1.  Unnamed parameters set using just value, 1. estimation methods, lower upper bounds can set 'theta' 'sigma' values.  set lower /upper bound, use vector values.  vector c(lower, estimate, upper).  vector may given just estimate (estimate), lower bound estimate (c(lower, estimate)), three (c(lower, estimate, upper)).  set estimate upper bound without lower bound, set lower bound -Inf, c(-Inf, estimate, upper).  estimation method support bounds, bounds ignored warning. 'omega' values can set single value values lower-triangular matrix.  values may set either variance-covariance matrix (default) correlation matrix -diagonals standard deviations diagonals.  Names may set left side ~.  set variance-covariance matrix variance values 2 3 covariance -2.5 use ~c(2, 2.5, 3).  set matrix names iivKa iivCL, use iivKa + iivCL~c(2, 2.5, 3).  set correlation matrix standard deviations diagonal, use cor() like iivKa + iivCL~cor(2, -0.5, 3). Values may fixed (therefore estimated) using either name fixed end assignment calling fixed() function value fix.  'theta' 'sigma', either estimate full definition (including lower upper bounds) may included fixed setting.  example, following effectively equivalent set 'theta' 'sigma' fixed value (lower upper bounds ignored fixed value): tvCL <- fixed(1), tvCL <- fixed(0, 1), tvCL <- fixed(0, 1, 2), tvCL <- c(0, fixed(1), 2), tvCL <- c(0, 1, fixed).  'omega' assignment, full block none block must set fixed.  Examples setting 'omega' value fixed : iivKa~fixed(1), iivKa + iivCL~fixed(1, 2, 3), iivKa + iivCL~c(1, 2, 3, fixed).  Anywhere fixed used, FIX, FIXED, fix may used equivalently. value, standard mathematical operators functions may used define value.  example, log(2) 24*30 may used define value anywhere number can used (e.g. lower bound, estimate, upper bound, variance, etc.). Values may labeled using label() function assignment. Labels used make reporting easier giving human-readable description parameter, labels effect estimation.  typical way set label parameter tvCL label \"Typical Value Clearance (L/hr)\" tvCL <- 1; label(\"Typical Value Clearance (L/hr)\"). rxode2/nlmixr2 attempt determine back-transformations user.  example, CL <- exp(tvCL) detect tvCL must back-transformed exp() easier interpretation.  want control back-transformation, can specify back-transformation using backTransform() assignment.  example, set back-transformation exp(), can use tvCL <- 1; backTransform(exp()).","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"","code":"# Set the ini() block in a model one.compartment <- function() {   ini({     tka <- log(1.57); label(\"Ka\")     tcl <- log(2.72); label(\"Cl\")     tv <- log(31.5); label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  # Use piping to update initial conditions one.compartment %>% ini(tka <- log(2)) #>   #>   #> ℹ change initial estimate of `tka` to `0.693147180559945` #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.6931472 1.0006319 3.4499875 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.693147180559945 #>         label(\"Ka\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } one.compartment %>% ini(tka <- label(\"Absorption rate, Ka (1/hr)\")) #>   #>   #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.4510756 1.0006319 3.4499875 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.451075619360217 #>         label(\"Absorption rate, Ka (1/hr)\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } # Move the tka parameter to be just below the tv parameter (affects parameter # summary table, only) one.compartment %>% ini(tka <- label(\"Absorption rate, Ka (1/hr)\"), append = \"tv\") #>   #>   #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tcl        tv       tka    add.sd  #> 1.0006319 3.4499875 0.4510756 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         tka <- 0.451075619360217 #>         label(\"Absorption rate, Ka (1/hr)\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } # When programming with rxode2/nlmixr2, it may be easier to pass strings in # to modify the ini one.compartment %>% ini(\"tka <- log(2)\") #>   #>   #> ℹ change initial estimate of `tka` to `0.693147180559945` #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.6931472 1.0006319 3.4499875 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.693147180559945 #>         label(\"Ka\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Check to see if this is an rxSolve object. — is.rxSolve","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"Check see rxSolve object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"","code":"is.rxSolve(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"x object check see rxSolve rxSolve object expired strip rxSolve information.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"boolean indicating rxSolve object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"Calculate log likelihood binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"","code":"llikBeta(x, shape1, shape2, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"x Observation shape1, shape2 non-negative parameters Beta distribution. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"data frame fx log pdf value dShape1 dShape2 derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"rxode2() model, can use llikBeta() use arguments.  can also get derivative shape1 shape2 llikBetaDshape1() llikBetaDshape2().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"","code":"# \\donttest{  x <- seq(1e-4, 1 - 1e-4, length.out = 21)  llikBeta(x, 0.5, 0.5) #>             fx     dShape1     dShape2 #> 1   3.46049030 -7.82404601  1.38619436 #> 2   0.37793108 -1.60763953  1.33490633 #> 3   0.05888752 -0.91549105  1.28084495 #> 4  -0.11510253 -0.51035907  1.22369308 #> 5  -0.22855163 -0.22284360  1.16307581 #> 6  -0.30780832  0.00019998  1.09854562 #> 7  -0.36444410  0.18245488  1.02956227 #> 8  -0.40444714  0.33655795  0.95546529 #> 9  -0.43118004  0.47005363  0.87543540 #> 10 -0.44655956  0.58780889  0.78843918 #> 11 -0.45158271  0.69314718  0.69314718 #> 12 -0.44655956  0.78843918  0.58780889 #> 13 -0.43118004  0.87543540  0.47005363 #> 14 -0.40444714  0.95546529  0.33655795 #> 15 -0.36444410  1.02956227  0.18245488 #> 16 -0.30780832  1.09854562  0.00019998 #> 17 -0.22855163  1.16307581 -0.22284360 #> 18 -0.11510253  1.22369308 -0.51035907 #> 19  0.05888752  1.28084495 -0.91549105 #> 20  0.37793108  1.33490633 -1.60763953 #> 21  3.46049030  1.38619436 -7.82404601  llikBeta(x, 1, 3, TRUE) #>          x shape1 shape2           fx     dShape1     dShape2 #> 1  0.00010      1      3   1.09841228 -7.37700704  0.33323333 #> 2  0.05009      1      3   0.99583622 -1.16060056  0.28194530 #> 3  0.10008      1      3   0.88771347 -0.46845208  0.22788392 #> 4  0.15007      1      3   0.77340972 -0.06332009  0.17073205 #> 5  0.20006      1      3   0.65217518  0.22419538  0.11011478 #> 6  0.25005      1      3   0.52311481  0.44723895  0.04558459 #> 7  0.30004      1      3   0.38514811  0.62949385 -0.02339876 #> 8  0.35003      1      3   0.23695415  0.78359692 -0.09749574 #> 9  0.40002      1      3   0.07689437  0.91709260 -0.17752562 #> 10 0.45001      1      3  -0.09709808  1.03484786 -0.26452185 #> 11 0.50000      1      3  -0.28768207  1.14018615 -0.35981385 #> 12 0.54999      1      3  -0.49835866  1.23547815 -0.46515214 #> 13 0.59998      1      3  -0.73386918  1.32247438 -0.58290740 #> 14 0.64997      1      3  -1.00086054  1.40250426 -0.71640308 #> 15 0.69996      1      3  -1.30906667  1.47660124 -0.87050615 #> 16 0.74995      1      3  -1.67357647  1.54558459 -1.05276105 #> 17 0.79994      1      3  -2.11966363  1.61011478 -1.27580462 #> 18 0.84993      1      3  -2.69469457  1.67073205 -1.56332009 #> 19 0.89992      1      3  -3.50495854  1.72788392 -1.96845208 #> 20 0.94991      1      3  -4.88925549  1.78194530 -2.66060056 #> 21 0.99990      1      3 -17.32206846  1.83323333 -8.87700704  et <- et(seq(1e-4, 1-1e-4, length.out=21)) et$shape1 <- 0.5 et$shape2 <- 1.5  model <- function() {   model({     fx <- llikBeta(time, shape1, shape2)     dShape1 <- llikBetaDshape1(time, shape1, shape2)     dShape2 <- llikBetaDshape2(time, shape1, shape2)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 21 × 6 #>     time     fx dShape1 dShape2 shape1 shape2 #>    <dbl>  <dbl>   <dbl>   <dbl>  <dbl>  <dbl> #> 1 0.0001 4.15   -6.82    0.386     0.5    1.5 #> 2 0.0501 1.02   -0.608   0.335     0.5    1.5 #> 3 0.100  0.647   0.0845  0.281     0.5    1.5 #> 4 0.150  0.415   0.490   0.224     0.5    1.5 #> 5 0.200  0.241   0.777   0.163     0.5    1.5 #> 6 0.250  0.0976  1.00    0.0985    0.5    1.5 #> # ℹ 15 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"Calculate log likelihood binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"","code":"llikBinom(x, size, prob, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"x Number successes size Size trial prob probability success full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"data frame fx pdf value dProb derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"rxode2() model, can use llikBinom() use arguments.  can also get derivative prob llikBinomDprob()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"","code":"# \\donttest{ llikBinom(46:54, 100, 0.5) #>          fx dProb #> 1 -2.848030   -16 #> 2 -2.709193   -12 #> 3 -2.610102    -8 #> 4 -2.550679    -4 #> 5 -2.530876     0 #> 6 -2.550679     4 #> 7 -2.610102     8 #> 8 -2.709193    12 #> 9 -2.848030    16  llikBinom(46:54, 100, 0.5, TRUE) #>    x size prob        fx dProb #> 1 46  100  0.5 -2.848030   -16 #> 2 47  100  0.5 -2.709193   -12 #> 3 48  100  0.5 -2.610102    -8 #> 4 49  100  0.5 -2.550679    -4 #> 5 50  100  0.5 -2.530876     0 #> 6 51  100  0.5 -2.550679     4 #> 7 52  100  0.5 -2.610102     8 #> 8 53  100  0.5 -2.709193    12 #> 9 54  100  0.5 -2.848030    16  # In rxode2 you can use:  et <- et(46:54) et$size <- 100 et$prob <-0.5  model <- function() {   model({     fx <- llikBinom(time, size, prob)     dProb <- llikBinomDprob(time, size, prob)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 9 × 5 #>    time    fx dProb  size  prob #>   <dbl> <dbl> <dbl> <dbl> <dbl> #> 1    46 -2.85   -16   100   0.5 #> 2    47 -2.71   -12   100   0.5 #> 3    48 -2.61    -8   100   0.5 #> 4    49 -2.55    -4   100   0.5 #> 5    50 -2.53     0   100   0.5 #> 6    51 -2.55     4   100   0.5 #> # ℹ 3 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"log likelihood Cauchy distribution derivatives (stan)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"","code":"llikCauchy(x, location = 0, scale = 1, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"x Observation location, scale location scale parameters. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"data frame fx log pdf value dLocation dScale derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"rxode2() model, can use llikCauchy() use arguments.  can also get derivative location scale llikCauchyDlocation() llikCauchyDscale().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"","code":"# \\donttest{ x <- seq(-3, 3, length.out = 21)  llikCauchy(x, 0, 1) #>           fx  dLocation     dScale #> 1  -3.447315 -0.6000000  0.8000000 #> 2  -3.259780 -0.6513872  0.7587455 #> 3  -3.055753 -0.7100592  0.7041420 #> 4  -2.832979 -0.7763401  0.6303142 #> 5  -2.589293 -0.8490566  0.5283019 #> 6  -2.323385 -0.9230769  0.3846154 #> 7  -2.036728 -0.9836066  0.1803279 #> 8  -1.738057 -0.9944751 -0.1049724 #> 9  -1.452215 -0.8823529 -0.4705882 #> 10 -1.230908 -0.5504587 -0.8348624 #> 11 -1.144730  0.0000000 -1.0000000 #> 12 -1.230908  0.5504587 -0.8348624 #> 13 -1.452215  0.8823529 -0.4705882 #> 14 -1.738057  0.9944751 -0.1049724 #> 15 -2.036728  0.9836066  0.1803279 #> 16 -2.323385  0.9230769  0.3846154 #> 17 -2.589293  0.8490566  0.5283019 #> 18 -2.832979  0.7763401  0.6303142 #> 19 -3.055753  0.7100592  0.7041420 #> 20 -3.259780  0.6513872  0.7587455 #> 21 -3.447315  0.6000000  0.8000000  llikCauchy(x, 3, 1, full=TRUE) #>       x location scale        fx  dLocation     dScale #> 1  -3.0        3     1 -4.755648 -0.3243243  0.9459459 #> 2  -2.7        3     1 -4.655977 -0.3404001  0.9402807 #> 3  -2.4        3     1 -4.551246 -0.3580902  0.9336870 #> 4  -2.1        3     1 -4.440937 -0.3776379  0.9259534 #> 5  -1.8        3     1 -4.324449 -0.3993344  0.9168053 #> 6  -1.5        3     1 -4.201087 -0.4235294  0.9058824 #> 7  -1.2        3     1 -4.070040 -0.4506438  0.8927039 #> 8  -0.9        3     1 -3.930358 -0.4811845  0.8766194 #> 9  -0.6        3     1 -3.780926 -0.5157593  0.8567335 #> 10 -0.3        3     1 -3.620428 -0.5550883  0.8317914 #> 11  0.0        3     1 -3.447315 -0.6000000  0.8000000 #> 12  0.3        3     1 -3.259780 -0.6513872  0.7587455 #> 13  0.6        3     1 -3.055753 -0.7100592  0.7041420 #> 14  0.9        3     1 -2.832979 -0.7763401  0.6303142 #> 15  1.2        3     1 -2.589293 -0.8490566  0.5283019 #> 16  1.5        3     1 -2.323385 -0.9230769  0.3846154 #> 17  1.8        3     1 -2.036728 -0.9836066  0.1803279 #> 18  2.1        3     1 -1.738057 -0.9944751 -0.1049724 #> 19  2.4        3     1 -1.452215 -0.8823529 -0.4705882 #> 20  2.7        3     1 -1.230908 -0.5504587 -0.8348624 #> 21  3.0        3     1 -1.144730  0.0000000 -1.0000000  et <- et(-3, 3, length.out=10) et$location <- 0 et$scale <- 1  model <- function() {   model({     fx <- llikCauchy(time, location, scale)     dLocation <- llikCauchyDlocation(time, location, scale)     dScale <- llikCauchyDscale(time, location, scale)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> Warning:  #> with negative times, compartments initialize at first negative observed time #> with positive times, compartments initialize at time zero #> use 'rxSetIni0(FALSE)' to initialize at first observed time #> this warning is displayed once per session #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>     time    fx dLocation dScale location scale #>    <dbl> <dbl>     <dbl>  <dbl>    <dbl> <dbl> #> 1 -3     -3.45    -0.6    0.8          0     1 #> 2 -2.33  -3.01    -0.724  0.690        0     1 #> 3 -1.67  -2.47    -0.882  0.471        0     1 #> 4 -1     -1.84    -1      0            0     1 #> 5 -0.333 -1.25    -0.6   -0.8          0     1 #> 6  0.333 -1.25     0.600 -0.8          0     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for chi-squared distribution — llikChisq","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"log likelihood derivatives chi-squared distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"","code":"llikChisq(x, df, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"x variable distributed chi-squared distribution df degrees freedom (non-negative, can non-integer). full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"data frame fx log pdf value dDf derivatives respect df parameter observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"rxode2() model, can use llikChisq() use x df arguments.  can also get derivative df llikChisqDdf().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"","code":"# \\donttest{ llikChisq(1, df = 1:3, full=TRUE) #>   x df        fx         dDf #> 1 1  1 -1.418939  0.63518142 #> 2 1  2 -1.193147 -0.05796576 #> 3 1  3 -1.418939 -0.36481858  llikChisq(1, df = 6:9) #>          fx        dDf #> 1 -3.272589 -0.8079658 #> 2 -4.126989 -0.8981519 #> 3 -5.064348 -0.9746324 #> 4 -6.072899 -1.0410091  et <- et(1:3) et$x <- 1  model <- function() {   model({    fx <- llikChisq(x, time)    dDf <- llikChisqDdf(x, time)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 3 × 4 #>    time    fx     dDf     x #>   <dbl> <dbl>   <dbl> <dbl> #> 1     1 -1.42  0.635      1 #> 2     2 -1.19 -0.0580     1 #> 3     3 -1.42 -0.365      1 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for exponential distribution — llikExp","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"log likelihood derivatives exponential distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"","code":"llikExp(x, rate, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"x variable distributed exponential distribution rate vector rates. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"data frame fx log pdf value dRate derivatives respect rate parameter observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"rxode2() model, can use llikExp() use x rate arguments.  can also get derivative rate llikExpDrate().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"","code":"# \\donttest{ llikExp(1, 1:3) #>          fx      dRate #> 1 -1.000000  0.0000000 #> 2 -1.306853 -0.5000000 #> 3 -1.901388 -0.6666667  llikExp(1, 1:3, full=TRUE) #>   x rate        fx      dRate #> 1 1    1 -1.000000  0.0000000 #> 2 1    2 -1.306853 -0.5000000 #> 3 1    3 -1.901388 -0.6666667  # You can use rxode2 for these too:  et <- et(1:3) et$x <- 1  model <- function() {   model({     fx <- llikExp(x, time)     dRate <- llikExpDrate(x, time)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 3 × 4 #>    time    fx  dRate     x #>   <dbl> <dbl>  <dbl> <dbl> #> 1     1 -1     0         1 #> 2     2 -1.31 -0.5       1 #> 3     3 -1.90 -0.667     1 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for F distribution — llikF","title":"log likelihood and derivatives for F distribution — llikF","text":"log likelihood derivatives F distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for F distribution — llikF","text":"","code":"llikF(x, df1, df2, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for F distribution — llikF","text":"x variable distributed f distribution df1, df2 degrees freedom.  Inf allowed. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for F distribution — llikF","text":"data frame fx log pdf value dDf1 dDf2 derivatives respect df1/df2 parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for F distribution — llikF","text":"rxode2() model, can use llikF() use x rate arguments.  can also get derivative df1 df2 llikFDdf1() llikFDdf2().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for F distribution — llikF","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for F distribution — llikF","text":"","code":"# \\donttest{ x <- seq(0.001, 5, length.out = 100)  llikF(x^2, 1, 5) #>               fx         dDf1          dDf2 #> 1    5.939135090 -5.769327755  0.0098138672 #> 2    1.996061339 -1.829698617  0.0098667834 #> 3    1.308027230 -1.151720596  0.0100203751 #> 4    0.898151814 -0.758467992  0.0102713542 #> 5    0.601410048 -0.484842238  0.0106143384 #> 6    0.365567397 -0.278507461  0.0110419510 #> 7    0.167275143 -0.115987118  0.0115449567 #> 8   -0.005928407  0.015333168  0.0121124287 #> 9   -0.161467803  0.123053577  0.0127319420 #> 10  -0.304110944  0.212138450  0.0133897879 #> 11  -0.437089645  0.286034532  0.0140712037 #> 12  -0.562684604  0.347253504  0.0147606108 #> 13  -0.682555520  0.397699568  0.0154418570 #> 14  -0.797939098  0.438864917  0.0160984557 #> 15  -0.909773848  0.471951996  0.0167138175 #> 16  -1.018782007  0.497952966  0.0172714700 #> 17  -1.125525217  0.517703019  0.0177552614 #> 18  -1.230443500  0.531917117  0.0181495454 #> 19  -1.333883243  0.541215913  0.0184393449 #> 20  -1.436117763  0.546144408  0.0186104926 #> 21  -1.537362702  0.547185650  0.0186497493 #> 22  -1.637787764  0.544770962  0.0185448973 #> 23  -1.737525805  0.539287723  0.0182848126 #> 24  -1.836679954  0.531085385  0.0178595139 #> 25  -1.935329266  0.520480216  0.0172601925 #> 26  -2.033533263  0.507759100  0.0164792220 #> 27  -2.131335592  0.493182655  0.0155101530 #> 28  -2.228767013  0.476987816  0.0143476919 #> 29  -2.325847849  0.459390036  0.0129876676 #> 30  -2.422589998  0.440585182  0.0114269877 #> 31  -2.518998603  0.420751193  0.0096635860 #> 32  -2.615073425  0.400049565  0.0076963639 #> 33  -2.710810000  0.378626673  0.0055251263 #> 34  -2.806200581  0.356614982  0.0031505147 #> 35  -2.901234940  0.334134152  0.0005739382 #> 36  -2.995901017  0.311292055 -0.0022024960 #> 37  -3.090185467  0.288185720 -0.0051760516 #> 38  -3.184074111  0.264902202 -0.0083434311 #> 39  -3.277552304  0.241519400 -0.0117008417 #> 40  -3.370605236  0.218106810 -0.0152440579 #> 41  -3.463218177  0.194726229 -0.0189684823 #> 42  -3.555376677  0.171432411 -0.0228692014 #> 43  -3.647066718  0.148273673 -0.0269410395 #> 44  -3.738274839  0.125292465 -0.0311786073 #> 45  -3.828988227  0.102525891 -0.0355763478 #> 46  -3.919194789  0.080006197 -0.0401285781 #> 47  -4.008883199  0.057761220 -0.0448295271 #> 48  -4.098042931  0.035814802 -0.0496733702 #> 49  -4.186664276  0.014187176 -0.0546542601 #> 50  -4.274738350 -0.007104684 -0.0597663547 #> 51  -4.362257091 -0.028046739 -0.0650038410 #> 52  -4.449213248 -0.048627585 -0.0703609577 #> 53  -4.535600366 -0.068838184 -0.0758320130 #> 54  -4.621412763 -0.088671614 -0.0814114015 #> 55  -4.706645505 -0.108122840 -0.0870936184 #> 56  -4.791294380 -0.127188508 -0.0928732707 #> 57  -4.875355868 -0.145866755 -0.0987450876 #> 58  -4.958827105 -0.164157039 -0.1047039284 #> 59  -5.041705859 -0.182059980 -0.1107447891 #> 60  -5.123990491 -0.199577223 -0.1168628074 #> 61  -5.205679923 -0.216711307 -0.1230532668 #> 62  -5.286773611 -0.233465550 -0.1293115988 #> 63  -5.367271506 -0.249843943 -0.1356333846 #> 64  -5.447174028 -0.265851061 -0.1420143563 #> 65  -5.526482033 -0.281491973 -0.1484503963 #> 66  -5.605196787 -0.296772169 -0.1549375368 #> 67  -5.683319932 -0.311697491 -0.1614719584 #> 68  -5.760853464 -0.326274073 -0.1680499885 #> 69  -5.837799706 -0.340508283 -0.1746680988 #> 70  -5.914161282 -0.354406680 -0.1813229026 #> 71  -5.989941093 -0.367975967 -0.1880111521 #> 72  -6.065142296 -0.381222952 -0.1947297352 #> 73  -6.139768282 -0.394154517 -0.2014756717 #> 74  -6.213822655 -0.406777586 -0.2082461105 #> 75  -6.287309215 -0.419099097 -0.2150383254 #> 76  -6.360231938 -0.431125983 -0.2218497115 #> 77  -6.432594959 -0.442865149 -0.2286777813 #> 78  -6.504402560 -0.454323453 -0.2355201612 #> 79  -6.575659147 -0.465507693 -0.2423745875 #> 80  -6.646369247 -0.476424593 -0.2492389023 #> 81  -6.716537483 -0.487080791 -0.2561110501 #> 82  -6.786168572 -0.497482830 -0.2629890742 #> 83  -6.855267306 -0.507637151 -0.2698711122 #> 84  -6.923838547 -0.517550083 -0.2767553934 #> 85  -6.991887212 -0.527227840 -0.2836402345 #> 86  -7.059418267 -0.536676516 -0.2905240366 #> 87  -7.126436719 -0.545902081 -0.2974052813 #> 88  -7.192947604 -0.554910380 -0.3042825282 #> 89  -7.258955983 -0.563707126 -0.3111544109 #> 90  -7.324466935 -0.572297907 -0.3180196342 #> 91  -7.389485548 -0.580688177 -0.3248769714 #> 92  -7.454016916 -0.588883261 -0.3317252608 #> 93  -7.518066132 -0.596888352 -0.3385634034 #> 94  -7.581638284 -0.604708517 -0.3453903600 #> 95  -7.644738447 -0.612348689 -0.3522051484 #> 96  -7.707371685 -0.619813679 -0.3590068413 #> 97  -7.769543041 -0.627108167 -0.3657945633 #> 98  -7.831257537 -0.634236713 -0.3725674892 #> 99  -7.892520171 -0.641203751 -0.3793248413 #> 100 -7.953335909 -0.648013598 -0.3860658874  model <- function(){   model({     fx <- llikF(time, df1, df2)     dMean <- llikFDdf1(time, df1, df2)     dSd <- llikFDdf2(time, df1, df2)   }) }  et <- et(x) et$df1 <- 1 et$df2 <- 5  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 100 × 6 #>     time     fx   dMean     dSd   df1   df2 #>    <dbl>  <dbl>   <dbl>   <dbl> <dbl> <dbl> #> 1 0.001   2.48  -2.32   0.00983     1     5 #> 2 0.0515  0.484 -0.380  0.0108      1     5 #> 3 0.102   0.112 -0.0731 0.0117      1     5 #> 4 0.152  -0.118  0.0943 0.0125      1     5 #> 5 0.203  -0.291  0.204  0.0133      1     5 #> 6 0.253  -0.431  0.283  0.0140      1     5 #> # ℹ 94 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Gamma distribution — llikGamma","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"log likelihood derivatives Gamma distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"","code":"llikGamma(x, shape, rate, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"x variable distributed gamma distribution shape distribution's shape parameter. Must positive. rate distribution's rate parameters.  Must positive. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"rxode2() model, can use llikGamma() use x rate arguments.  can also get derivative shape rate llikGammaDshape() llikGammaDrate().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"","code":"# \\donttest{  llikGamma(1, 1, 10) #>          fx   dShape dRate #> 1 -7.697415 2.879801  -0.9  # You can use this in `rxode2` too:  et  <- et(seq(0.001, 1, length.out=10)) et$shape <- 1 et$rate <- 10   model <- function() {   model({     fx <- llikGamma(time, shape, rate)     dShape<- llikGammaDshape(time, shape, rate)     dRate <- llikGammaDrate(time, shape, rate)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>    time      fx dShape  dRate  rate shape #>   <dbl>   <dbl>  <dbl>  <dbl> <dbl> <dbl> #> 1 0.001  2.29   -4.03   0.099    10     1 #> 2 0.112  1.18    0.691 -0.012    10     1 #> 3 0.223  0.0726  1.38  -0.123    10     1 #> 4 0.334 -1.04    1.78  -0.234    10     1 #> 5 0.445 -2.15    2.07  -0.345    10     1 #> 6 0.556 -3.26    2.29  -0.456    10     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Geom distribution — llikGeom","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"log likelihood derivatives Geom distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"","code":"llikGeom(x, prob, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"x variable distributed geom distribution prob probability success trial. 0 < prob <= 1. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"rxode2() model, can use llikGeom() use x rate arguments.  can also get derivative prob llikGeomDprob().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"","code":"# \\donttest{  llikGeom(1:10, 0.2) #>           fx dProb #> 1  -1.832581  3.75 #> 2  -2.055725  2.50 #> 3  -2.278869  1.25 #> 4  -2.502012  0.00 #> 5  -2.725156 -1.25 #> 6  -2.948299 -2.50 #> 7  -3.171443 -3.75 #> 8  -3.394586 -5.00 #> 9  -3.617730 -6.25 #> 10 -3.840873 -7.50  et  <- et(1:10) et$prob <- 0.2   model <- function() {   model({     fx <- llikGeom(time, prob)     dProb <- llikGeomDprob(time, prob)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time    fx dProb  prob #>   <dbl> <dbl> <dbl> <dbl> #> 1     1 -1.83  3.75   0.2 #> 2     2 -2.06  2.5    0.2 #> 3     3 -2.28  1.25   0.2 #> 4     4 -2.50  0      0.2 #> 5     5 -2.73 -1.25   0.2 #> 6     6 -2.95 -2.5    0.2 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"Calculate log likelihood negative binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"","code":"llikNbinom(x, size, prob, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"x Number successes size Size trial prob probability success full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"data frame fx pdf value dProb derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"rxode2() model, can use llikNbinom() use arguments.  can also get derivative prob llikNbinomDprob()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"","code":"# \\donttest{ llikNbinom(46:54, 100, 0.5) #>          fx dProb #> 1 -13.25200   108 #> 2 -12.81168   106 #> 3 -12.38560   104 #> 4 -11.97335   102 #> 5 -11.57458   100 #> 6 -11.18892    98 #> 7 -10.81603    96 #> 8 -10.45559    94 #> 9 -10.10728    92  llikNbinom(46:54, 100, 0.5, TRUE) #>    x size prob        fx dProb #> 1 46  100  0.5 -13.25200   108 #> 2 47  100  0.5 -12.81168   106 #> 3 48  100  0.5 -12.38560   104 #> 4 49  100  0.5 -11.97335   102 #> 5 50  100  0.5 -11.57458   100 #> 6 51  100  0.5 -11.18892    98 #> 7 52  100  0.5 -10.81603    96 #> 8 53  100  0.5 -10.45559    94 #> 9 54  100  0.5 -10.10728    92  # In rxode2 you can use:  et <- et(46:54) et$size <- 100 et$prob <-0.5  model <- function() {   model({     fx <- llikNbinom(time, size, prob)     dProb <- llikNbinomDprob(time, size, prob)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 9 × 5 #>    time    fx dProb  size  prob #>   <dbl> <dbl> <dbl> <dbl> <dbl> #> 1    46 -13.3   108   100   0.5 #> 2    47 -12.8   106   100   0.5 #> 3    48 -12.4   104   100   0.5 #> 4    49 -12.0   102   100   0.5 #> 5    50 -11.6   100   100   0.5 #> 6    51 -11.2    98   100   0.5 #> # ℹ 3 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"Calculate log likelihood negative binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"","code":"llikNbinomMu(x, size, mu, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"x Number successes size Size trial mu mu parameter negative binomial full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"data frame fx pdf value dProb derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"rxode2() model, can use llikNbinomMu() use arguments.  can also get derivative mu llikNbinomMuDmu()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"","code":"# \\donttest{ llikNbinomMu(46:54, 100, 40) #>          fx       dMu #> 1 -3.326828 0.1071429 #> 2 -3.446132 0.1250000 #> 3 -3.579663 0.1428571 #> 4 -3.727034 0.1607143 #> 5 -3.887874 0.1785714 #> 6 -4.061827 0.1964286 #> 7 -4.248554 0.2142857 #> 8 -4.447728 0.2321429 #> 9 -4.659037 0.2500000  llikNbinomMu(46:54, 100, 40, TRUE) #>    x size mu        fx       dMu #> 1 46  100 40 -3.326828 0.1071429 #> 2 47  100 40 -3.446132 0.1250000 #> 3 48  100 40 -3.579663 0.1428571 #> 4 49  100 40 -3.727034 0.1607143 #> 5 50  100 40 -3.887874 0.1785714 #> 6 51  100 40 -4.061827 0.1964286 #> 7 52  100 40 -4.248554 0.2142857 #> 8 53  100 40 -4.447728 0.2321429 #> 9 54  100 40 -4.659037 0.2500000  et <- et(46:54) et$size <- 100 et$mu <- 40  model <- function() {   model({     fx <- llikNbinomMu(time, size, mu)     dProb <- llikNbinomMuDmu(time, size, mu)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 9 × 5 #>    time    fx dProb  size    mu #>   <dbl> <dbl> <dbl> <dbl> <dbl> #> 1    46 -3.33 0.107   100    40 #> 2    47 -3.45 0.125   100    40 #> 3    48 -3.58 0.143   100    40 #> 4    49 -3.73 0.161   100    40 #> 5    50 -3.89 0.179   100    40 #> 6    51 -4.06 0.196   100    40 #> # ℹ 3 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":null,"dir":"Reference","previous_headings":"","what":"Log likelihood for normal distribution — llikNorm","title":"Log likelihood for normal distribution — llikNorm","text":"Log likelihood normal distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Log likelihood for normal distribution — llikNorm","text":"","code":"llikNorm(x, mean = 0, sd = 1, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Log likelihood for normal distribution — llikNorm","text":"x Observation mean Mean likelihood sd Standard deviation likelihood full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Log likelihood for normal distribution — llikNorm","text":"data frame fx pdf value dMean dSd derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Log likelihood for normal distribution — llikNorm","text":"rxode2() model, can use llikNorm() use arguments.  can also get derivatives llikNormDmean() llikNormDsd()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Log likelihood for normal distribution — llikNorm","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Log likelihood for normal distribution — llikNorm","text":"","code":"# \\donttest{  llikNorm(0) #>           fx dMean dSd #> 1 -0.9189385     0  -1  llikNorm(seq(-2,2,length.out=10), full=TRUE) #>             x mean sd         fx      dMean        dSd #> 1  -2.0000000    0  1 -2.9189385 -2.0000000  3.0000000 #> 2  -1.5555556    0  1 -2.1288151 -1.5555556  1.4197531 #> 3  -1.1111111    0  1 -1.5362225 -1.1111111  0.2345679 #> 4  -0.6666667    0  1 -1.1411608 -0.6666667 -0.5555556 #> 5  -0.2222222    0  1 -0.9436299 -0.2222222 -0.9506173 #> 6   0.2222222    0  1 -0.9436299  0.2222222 -0.9506173 #> 7   0.6666667    0  1 -1.1411608  0.6666667 -0.5555556 #> 8   1.1111111    0  1 -1.5362225  1.1111111  0.2345679 #> 9   1.5555556    0  1 -2.1288151  1.5555556  1.4197531 #> 10  2.0000000    0  1 -2.9189385  2.0000000  3.0000000  # With rxode2 you can use:   et <- et(-3, 3, length.out=10) et$mu <- 0 et$sigma <- 1  model <- function(){   model({     fx <- llikNorm(time, mu, sigma)     dMean <- llikNormDmean(time, mu, sigma)     dSd <- llikNormDsd(time, mu, sigma)   })  }  ret <- rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ret #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>     time     fx  dMean    dSd    mu sigma #>    <dbl>  <dbl>  <dbl>  <dbl> <dbl> <dbl> #> 1 -3     -5.42  -3      8         0     1 #> 2 -2.33  -3.64  -2.33   4.44      0     1 #> 3 -1.67  -2.31  -1.67   1.78      0     1 #> 4 -1     -1.42  -1      0         0     1 #> 5 -0.333 -0.974 -0.333 -0.889     0     1 #> 6  0.333 -0.974  0.333 -0.889     0     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":null,"dir":"Reference","previous_headings":"","what":"log-likelihood for the Poisson distribution — llikPois","title":"log-likelihood for the Poisson distribution — llikPois","text":"log-likelihood Poisson distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log-likelihood for the Poisson distribution — llikPois","text":"","code":"llikPois(x, lambda, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log-likelihood for the Poisson distribution — llikPois","text":"x non negative integers lambda non-negative means full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log-likelihood for the Poisson distribution — llikPois","text":"data frame fx pdf value dLambda derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log-likelihood for the Poisson distribution — llikPois","text":"rxode2() model, can use llikPois() use arguments.  can also get derivatives llikPoisDlambda()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log-likelihood for the Poisson distribution — llikPois","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log-likelihood for the Poisson distribution — llikPois","text":"","code":"# \\donttest{ llikPois(0:7, lambda = 1) #>          fx dLambda #> 1 -1.000000      -1 #> 2 -1.000000       0 #> 3 -1.693147       1 #> 4 -2.791759       2 #> 5 -4.178054       3 #> 6 -5.787492       4 #> 7 -7.579251       5 #> 8 -9.525161       6  llikPois(0:7, lambda = 4, full=TRUE) #>   x lambda        fx dLambda #> 1 0      4 -4.000000   -1.00 #> 2 1      4 -2.613706   -0.75 #> 3 2      4 -1.920558   -0.50 #> 4 3      4 -1.632876   -0.25 #> 5 4      4 -1.632876    0.00 #> 6 5      4 -1.856020    0.25 #> 7 6      4 -2.261485    0.50 #> 8 7      4 -2.821101    0.75  # In rxode2 you can use:  et <- et(0:10) et$lambda <- 0.5  model <- function() {   model({     fx <- llikPois(time, lambda)     dLambda <- llikPoisDlambda(time, lambda)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 11 × 4 #>    time    fx dLambda lambda #>   <dbl> <dbl>   <dbl>  <dbl> #> 1     0 -0.5       -1    0.5 #> 2     1 -1.19       1    0.5 #> 3     2 -2.58       3    0.5 #> 4     3 -4.37       5    0.5 #> 5     4 -6.45       7    0.5 #> 6     5 -8.75       9    0.5 #> # ℹ 5 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":null,"dir":"Reference","previous_headings":"","what":"Log likelihood of T and it's derivatives (from stan) — llikT","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"Log likelihood T derivatives (stan)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"","code":"llikT(x, df, mean = 0, sd = 1, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"x Observation df degrees freedom (\\(> 0\\), maybe non-integer).  df       = Inf allowed. mean Mean likelihood sd Standard deviation likelihood full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"data frame fx log pdf value dDf dMean dSd derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"rxode2() model, can use llikT() use arguments.  can also get derivative df, mean sd llikTDdf(), llikTDmean() llikTDsd().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"","code":"# \\donttest{ x <- seq(-3, 3, length.out = 21)  llikT(x, 7, 0, 1) #>            fx          dDf      dMean        dSd #> 1  -4.2612484 -0.086858773 -1.5000000  3.5000000 #> 2  -3.8091335 -0.060260433 -1.5115465  3.0811756 #> 3  -3.3561547 -0.037201674 -1.5047022  2.6112853 #> 4  -2.9088542 -0.018379169 -1.4723926  2.0920245 #> 5  -2.4761000 -0.004340221 -1.4062500  1.5312500 #> 6  -2.0693878  0.004691380 -1.2972973  0.9459459 #> 7  -1.7028228  0.009010785 -1.1374408  0.3649289 #> 8  -1.3925134  0.009569214 -0.9218950 -0.1702945 #> 9  -1.1551333  0.007927361 -0.6521739 -0.6086957 #> 10 -1.0056349  0.005918024 -0.3385049 -0.8984485 #> 11 -0.9545342  0.005051942  0.0000000 -1.0000000 #> 12 -1.0056349  0.005918024  0.3385049 -0.8984485 #> 13 -1.1551333  0.007927361  0.6521739 -0.6086957 #> 14 -1.3925134  0.009569214  0.9218950 -0.1702945 #> 15 -1.7028228  0.009010785  1.1374408  0.3649289 #> 16 -2.0693878  0.004691380  1.2972973  0.9459459 #> 17 -2.4761000 -0.004340221  1.4062500  1.5312500 #> 18 -2.9088542 -0.018379169  1.4723926  2.0920245 #> 19 -3.3561547 -0.037201674  1.5047022  2.6112853 #> 20 -3.8091335 -0.060260433  1.5115465  3.0811756 #> 21 -4.2612484 -0.086858773  1.5000000  3.5000000  llikT(x, 15, 0, 1, full=TRUE) #>            fx           dDf      dMean        dSd #> 1  -4.6956220 -0.0338931511 -2.0000000  5.0000000 #> 2  -4.1042965 -0.0225073150 -1.9380888  4.2328398 #> 3  -3.5354158 -0.0134033865 -1.8497110  3.4393064 #> 4  -2.9974985 -0.0065857823 -1.7310665  2.6352396 #> 5  -2.5001272 -0.0019378861 -1.5789474  1.8421053 #> 6  -2.0536885  0.0007929097 -1.3913043  1.0869565 #> 7  -1.6689304  0.0019903977 -1.1678832  0.4014599 #> 8  -1.3563325  0.0021369166 -0.9108159 -0.1802657 #> 9  -1.1253251  0.0017504002 -0.6250000 -0.6250000 #> 10 -0.9834495  0.0012985422 -0.3180915 -0.9045726 #> 11 -0.9355929  0.0011086635  0.0000000 -1.0000000 #> 12 -0.9834495  0.0012985422  0.3180915 -0.9045726 #> 13 -1.1253251  0.0017504002  0.6250000 -0.6250000 #> 14 -1.3563325  0.0021369166  0.9108159 -0.1802657 #> 15 -1.6689304  0.0019903977  1.1678832  0.4014599 #> 16 -2.0536885  0.0007929097  1.3913043  1.0869565 #> 17 -2.5001272 -0.0019378861  1.5789474  1.8421053 #> 18 -2.9974985 -0.0065857823  1.7310665  2.6352396 #> 19 -3.5354158 -0.0134033865  1.8497110  3.4393064 #> 20 -4.1042965 -0.0225073150  1.9380888  4.2328398 #> 21 -4.6956220 -0.0338931511  2.0000000  5.0000000  et <- et(-3, 3, length.out=10) et$nu <- 7 et$mean <- 0 et$sd <- 1  model <- function() {   model({     fx <- llikT(time, nu, mean, sd)     dDf <- llikTDdf(time, nu, mean, sd)     dMean <- llikTDmean(time, nu, mean, sd)     dSd   <- llikTDsd(time, nu, mean, sd)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 8 #>     time    fx       dDf  dMean    dSd    nu  mean    sd #>    <dbl> <dbl>     <dbl>  <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1 -3     -4.26 -0.0869   -1.5    3.5       7     0     1 #> 2 -2.33  -3.26 -0.0326   -1.5    2.5       7     0     1 #> 3 -1.67  -2.29  0.000289 -1.36   1.27      7     0     1 #> 4 -1     -1.49  0.00971  -1      0         7     0     1 #> 5 -0.333 -1.02  0.00611  -0.375 -0.875     7     0     1 #> 6  0.333 -1.02  0.00611   0.375 -0.875     7     0     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Unif distribution — llikUnif","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"log likelihood derivatives Unif distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"","code":"llikUnif(x, alpha, beta, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"x variable distributed uniform distribution alpha lower limit uniform distribution beta upper limit distribution full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"rxode2() model, can use llikUnif() use x rate arguments.  can also get derivative alpha beta llikUnifDalpha() llikUnifDbeta().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"","code":"# \\donttest{  llikUnif(1, -2, 2) #>          fx dAlpha dBeta #> 1 -1.386294   0.25 -0.25  et  <- et(seq(1,1, length.out=4)) et$alpha <- -2 et$beta <- 2   model <- function() {   model({     fx <- llikUnif(time, alpha, beta)     dAlpha<- llikUnifDalpha(time, alpha, beta)     dBeta <- llikUnifDbeta(time, alpha, beta)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 4 × 6 #>    time    fx dAlpha dBeta alpha  beta #>   <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1     1 -1.39   0.25 -0.25    -2     2 #> 2     1 -1.39   0.25 -0.25    -2     2 #> 3     1 -1.39   0.25 -0.25    -2     2 #> 4     1 -1.39   0.25 -0.25    -2     2 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Weibull distribution — llikWeibull","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"log likelihood derivatives Weibull distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"","code":"llikWeibull(x, shape, scale, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"x variable distributed Weibull distribution shape, scale shape scale parameters, latter defaulting 1. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"rxode2() model, can use llikWeibull() use x rate arguments.  can also get derivative shape scale llikWeibullDshape() llikWeibullDscale().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"","code":"# \\donttest{ llikWeibull(1, 1, 10) #>          fx    dShape dScale #> 1 -2.402585 -1.072327  -0.09  # rxode2 can use this too:  et  <- et(seq(0.001, 1, length.out=10)) et$shape <- 1 et$scale <- 10   model <- function() {   model({     fx <- llikWeibull(time, shape, scale)     dShape<- llikWeibullDshape(time, shape, scale)     dScale <- llikWeibullDscale(time, shape, scale)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>    time    fx dShape  dScale shape scale #>   <dbl> <dbl>  <dbl>   <dbl> <dbl> <dbl> #> 1 0.001 -2.30  -8.21 -0.100      1    10 #> 2 0.112 -2.31  -3.44 -0.0989     1    10 #> 3 0.223 -2.32  -2.72 -0.0978     1    10 #> 4 0.334 -2.34  -2.29 -0.0967     1    10 #> 5 0.445 -2.35  -1.97 -0.0956     1    10 #> 6 0.556 -2.36  -1.73 -0.0944     1    10 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":null,"dir":"Reference","previous_headings":"","what":"logit and inverse logit (expit) functions — logit","title":"logit and inverse logit (expit) functions — logit","text":"logit inverse logit (expit) functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"logit and inverse logit (expit) functions — logit","text":"","code":"logit(x, low = 0, high = 1)  expit(alpha, low = 0, high = 1)  logitNormInfo(mean = 0, sd = 1, low = 0, high = 1, abs.tol = 1e-06, ...)  probitNormInfo(mean = 0, sd = 1, low = 0, high = 1, abs.tol = 1e-06, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"logit and inverse logit (expit) functions — logit","text":"x Input value(s) range [low,high] translate -Inf Inf low Lowest value range high Highest value range alpha Infinite value(s) translate range [low, high] mean logit-scale mean sd logit-scale standard deviation abs.tol absolute accuracy requested. ... parameters passed integrate()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"logit and inverse logit (expit) functions — logit","text":"values logit expit","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"logit and inverse logit (expit) functions — logit","text":"logit given : logit(p) = -log(1/p-1) : p = x-low/high-low expit given : expit(p, low, high) = (high-low)/(1+exp(-alpha)) + low logitNormInfo() gives mean, variance coefficient variability untransformed scale.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"logit and inverse logit (expit) functions — logit","text":"","code":"logit(0.25) #> [1] -1.098612  expit(-1.09) #> [1] 0.2516183  logitNormInfo(logit(0.25), sd = 0.1) #>         mean          var           cv  #> 0.2504672899 0.0003515538 0.0748591440   logitNormInfo(logit(1, 0, 10), sd = 1, low = 0, high = 10) #>      mean       var        cv  #> 1.3386966 1.2662276 0.8405697"},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":null,"dir":"Reference","previous_headings":"","what":"lowergamma: upper incomplete gamma function — lowergamma","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"tgamma_lower boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"","code":"lowergamma(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"numeric '' parameter upper incomplete gamma z numeric 'z' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"lowergamma results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"lowergamma function given : $$lowergamma(, z) = \\int_{0}^{z}t^{-1}\\cdot e^{-t} dt$$","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"","code":"lowergamma(1, 3) #> [1] 0.9502129  lowergamma(1:3, 3) #> [1] 0.9502129 0.8008517 1.1536198  lowergamma(1, 1:3) #> [1] 0.6321206 0.8646647 0.9502129"},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"generic function meanProbs produces expected confidence bands either t distribution normal sampling distribution. uses qnorm() qt() mean standard deviation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"","code":"meanProbs(x, ...)  # S3 method for default meanProbs(   x,   probs = seq(0, 1, 0.25),   na.rm = FALSE,   names = TRUE,   useT = TRUE,   onlyProbs = TRUE,   pred = FALSE,   n = 0L,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"x numeric vector whose mean probability based confidence values wanted, NA NaN values allowed numeric vectors unless ‘na.rm’ ‘TRUE’. ... Arguments passed default method, allows many different methods applied. probs numeric vector probabilities values interval 0 1 . na.rm logical; true, NA NaN's removed x quantiles computed. names logical; true, result names attribute. useT logical; true, use t-distribution calculate confidence-based estimates. false use normal distribution calculate confidence based estimates. onlyProbs logical; true, return probability based confidence interval estimates, otherwise return pred logical; true use prediction interval instead confidence interval n integer/integerish; n used calculate prediction confidence interval.  n=0 (default) use number non-NA observations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"default return probabilities names (named) points expected distribution located given sampling mean standard deviation. onlyProbs=FALSE prepend mean, variance, standard deviation, minimum, maximum number non-NA observations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"single probability, p, uses either: mean + qt(p, df=n)*sd/sqrt(n) mean + qnorm(p)*sd/sqrt(n) smallest observation corresponds probability 0 largest probability 1 mean corresponds 0.5. mean standard deviation sample calculated based Welford's method single pass. meant perform way quantile() can drop replacement code using quantile() using distributional assumptions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"","code":"quantile(x<- rnorm(1001)) #>           0%          25%          50%          75%         100%  #> -3.403225585 -0.642325341 -0.001837139  0.656154718  2.780013343  meanProbs(x) #>           0%          25%          50%          75%         100%  #> -3.403225585 -0.023025596 -0.002431956  0.018161685  2.780013343   # Can get some extra statistics if you request onlyProbs=FALSE meanProbs(x, onlyProbs=FALSE) #>          mean           var            sd           min           max  #> -2.431956e-03  9.324674e-01  9.656435e-01 -3.403226e+00  2.780013e+00  #>             n            0%           25%           50%           75%  #>  1.001000e+03 -3.403226e+00 -2.302560e-02 -2.431956e-03  1.816168e-02  #>          100%  #>  2.780013e+00   x[2] <- NA_real_  meanProbs(x, onlyProbs=FALSE) #> mean  var   sd  min  max    n   0%  25%  50%  75% 100%  #>   NA   NA   NA   NA   NA   NA   NA   NA   NA   NA   NA   quantile(x<- rnorm(42)) #>         0%        25%        50%        75%       100%  #> -2.1838542 -0.8532155  0.1476959  0.8088171  1.5847511   meanProbs(x) #>          0%         25%         50%         75%        100%  #> -2.18385418 -0.18647291 -0.07532481  0.03582330  1.58475107   meanProbs(x, useT=FALSE) #>          0%         25%         50%         75%        100%  #> -2.18385418 -0.18548788 -0.07532481  0.03483827  1.58475107"},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign the model block in the rxode2 related object — model<-","title":"Assign the model block in the rxode2 related object — model<-","text":"Assign model block rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign the model block in the rxode2 related object — model<-","text":"","code":"model(x, envir = environment(x)) <- value"},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign the model block in the rxode2 related object — model<-","text":"x rxode2 related object envir Environment assignment occurs value Value object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign the model block in the rxode2 related object — model<-","text":"rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign the model block in the rxode2 related object — model<-","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":null,"dir":"Reference","previous_headings":"","what":"Model block for rxode2/nlmixr models — model.function","title":"Model block for rxode2/nlmixr models — model.function","text":"Model block rxode2/nlmixr models","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Model block for rxode2/nlmixr models — model.function","text":"","code":"# S3 method for `function` model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for rxUi model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for rxode2 model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for rxModelVars model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  model(   x,   ...,   append = FALSE,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for default model(x, ..., append = FALSE, cov = NULL, envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Model block for rxode2/nlmixr models — model.function","text":"x model expression ... arguments append boolean determine lines appended piping.  possible values : TRUE lines appended model instead replaced FALSE lines replaced model (default) NA lines pre-pended model instead replaced lhs expression, append lines last observed line expression lhs auto boolean tells piping automatically selects parameters characterized population parameter, subject variability, covariate.  TRUE automatic selection occurs.  FALSE automatic selection turned everything added covariate (can promoted parameter ini statement). default TRUE, can changed options(rxode2.autoVarPiping=FALSE). cov character vector variables assumed covariates.  override automatic promotion population parameter estimate (eta) envir environment unevaluated model expressions evaluated.  May also NULL, list, data frame, pairlist integer specified sys.call.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Model block for rxode2/nlmixr models — model.function","text":"Model block ini information included.  ini must called model block","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Model block for rxode2/nlmixr models — model.function","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract model lines from a rxui model — modelExtract","title":"Extract model lines from a rxui model — modelExtract","text":"Extract model lines rxui model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract model lines from a rxui model — modelExtract","text":"","code":"modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for `function` modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for rxUi modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for rxode2 modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for rxModelVars modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for default modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract model lines from a rxui model — modelExtract","text":"x model extract lines ... variables extract. missing, extract entire model (conditioned endpoint option ) expression return expressions (TRUE) strings (FALSE) endpoint include endpoint.  can : NA    -- Missing means include endpoint non-endpoint lines TRUE  -- include endpoint lines FALSE -- include non-endpoint lines lines boolean.  TRUE add lines attribute output value ie attr(, \"lines\") envir Environment evaluating variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract model lines from a rxui model — modelExtract","text":"expressions strings extracted lines. Note duplicated lhs expression line, return lines","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Extract model lines from a rxui model — modelExtract","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract model lines from a rxui model — modelExtract","text":"","code":"one.compartment <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v  <- exp(tv + eta.v)     d/dt(depot)  <- -ka * depot     d/dt(center) <-  ka * depot - cl / v * center     cp <- center / v     cp ~ add(add.sd)   })  }   f <- one.compartment()   modelExtract(f, cp) #> [1] \"cp <- center/v\"   modelExtract(one.compartment, d/dt(depot)) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #> [1] \"d/dt(depot) <- -ka * depot\"   # from variable  var <- \"d/dt(depot)\"   modelExtract(one.compartment, var) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #> [1] \"d/dt(depot) <- -ka * depot\"   modelExtract(f, endpoint=NA, lines=TRUE, expression=TRUE) #> [[1]] #> ka <- exp(tka + eta.ka) #>  #> [[2]] #> cl <- exp(tcl + eta.cl) #>  #> [[3]] #> v <- exp(tv + eta.v) #>  #> [[4]] #> d/dt(depot) <- -ka * depot #>  #> [[5]] #> d/dt(center) <- ka * depot - cl/v * center #>  #> [[6]] #> cp <- center/v #>  #> [[7]] #> cp ~ add(add.sd) #>  #> attr(,\"lines\") #> [1] 1 2 3 4 5 6 7"},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":null,"dir":"Reference","previous_headings":"","what":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"NULL given method, choices returned named vector.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"","code":"odeMethodToInt(method = c(\"liblsoda\", \"lsoda\", \"dop853\", \"indLin\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"method method solving ODEs.  Currently supports: \"liblsoda\" thread safe lsoda.  supports parallel thread-based solving, ignores user Jacobian specification. \"lsoda\" -- LSODA solver.  support parallel thread-based solving, allows user Jacobian specification. \"dop853\" -- DOP853 solver.  support parallel thread-based solving user Jacobian specification \"indLin\" -- Solving inductive linearization.  rxode2 dll must setup specially use solving routine.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"integer method (unless input NULL, case, see details)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot rxode2 objects — plot.rxSolve","title":"Plot rxode2 objects — plot.rxSolve","text":"Plot rxode2 objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot rxode2 objects — plot.rxSolve","text":"","code":"# S3 method for rxSolve plot(x, y, ..., log = \"\", xlab = \"Time\", ylab = \"\")  # S3 method for rxSolveConfint1 plot(x, y, ..., xlab = \"Time\", ylab = \"\", log = \"\")  # S3 method for rxSolveConfint2 plot(x, y, ..., xlab = \"Time\", ylab = \"\", log = \"\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot rxode2 objects — plot.rxSolve","text":"x rxode2 object plot y Compartments left-hand-side values plot either bare name character vector ... Ignored log \"\" (neither x y), \"x\", \"y\", \"xy\" (\"yx\") log-scale? xlab, ylab x y axis labels","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot rxode2 objects — plot.rxSolve","text":"ggplot2 object","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Update Solved object with '+' — +.rxSolve","title":"Update Solved object with '+' — +.rxSolve","text":"Update Solved object '+'","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update Solved object with '+' — +.rxSolve","text":"","code":"# S3 method for rxSolve +(solved, new)"},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update Solved object with '+' — +.rxSolve","text":"solved Solved object new New information added table.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update Solved object with '+' — +.rxSolve","text":"new solved object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Update Solved object with '+' — +.rxSolve","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":null,"dir":"Reference","previous_headings":"","what":"Print the rxCoef object — print.rxCoef","title":"Print the rxCoef object — print.rxCoef","text":"prints user supplied arguments rxCoef object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print the rxCoef object — print.rxCoef","text":"","code":"# S3 method for rxCoef print(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print the rxCoef object — print.rxCoef","text":"x rxCoef object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print the rxCoef object — print.rxCoef","text":"original object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print the rxCoef object — print.rxCoef","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":null,"dir":"Reference","previous_headings":"","what":"Print rxDll object — print.rxDll","title":"Print rxDll object — print.rxDll","text":"tells rxDll loaded, ready /deleted.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print rxDll object — print.rxDll","text":"","code":"# S3 method for rxDll print(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print rxDll object — print.rxDll","text":"original object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print rxDll object — print.rxDll","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Print information about the rxode2 object. — print.rxode2","title":"Print information about the rxode2 object. — print.rxode2","text":"prints model name status able solved","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print information about the rxode2 object. — print.rxode2","text":"","code":"# S3 method for rxode2 print(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print information about the rxode2 object. — print.rxode2","text":"x rxode2 object ... Ignored parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print information about the rxode2 object. — print.rxode2","text":"original object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print information about the rxode2 object. — print.rxode2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":null,"dir":"Reference","previous_headings":"","what":"probit and inverse probit functions — probit","title":"probit and inverse probit functions — probit","text":"probit inverse probit functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"probit and inverse probit functions — probit","text":"","code":"probit(x, low = 0, high = 1)  probitInv(x, low = 0, high = 1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"probit and inverse probit functions — probit","text":"x Input value(s) range [low,high] translate -Inf Inf low Lowest value range high Highest value range","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"probit and inverse probit functions — probit","text":"values probit, probitInv probitNormInfo","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"probit and inverse probit functions — probit","text":"","code":"probit(0.25) #> [1] -0.6744898  probitInv(-0.674) #> [1] 0.2501557  probitNormInfo(probit(0.25), sd = 0.1) #>       mean        var         cv  #> 0.25106491 0.00100658 0.12636827   probitNormInfo(probit(1, 0, 10), sd = 1, low = 0, high = 10) #>     mean      var       cv  #> 1.824166 4.307824 1.137796"},{"path":"https://nlmixr2.github.io/rxode2/reference/reexports.html","id":null,"dir":"Reference","previous_headings":"","what":"Objects exported from other packages — reexports","title":"Objects exported from other packages — reexports","text":"objects imported packages. Follow links see documentation. ggplot2 aes, expand_limits, facet_wrap, geom_line, ggplot, label_both, label_context, label_context, label_value, label_wrap_gen, scale_type, scale_x_continuous, scale_x_date, scale_x_discrete, scale_y_continuous, scale_y_date, scale_y_discrete, waiver, xlab, ylab lotri lotri magrittr %>% rxode2et .clearPipe, .collectWarnings, .s3register, add.dosing, add.sampling, .et, .rxEvid, .rxRateDur, et, etExpand, etRbind, etRep, etSeq, eventTable, rxCbindStudyIndividual, rxEtDispatchSolve, rxEvid, rxRateDur, rxReq, rxStack, toTrialDuration rxode2parse .getLastIdLvl, forderForceBase, rxDerived, rxSetIni0 rxode2random .cbindOme, .expandPars, .vecDf, cvPost, invWR1d, phi, rinvchisq, rLKJ1, rxGetSeed, rxGetSeed, rxRmvn, rxSeedEng, rxSetSeed, rxSetSeed, rxSetSeed, rxWithPreserveSeed, rxWithSeed, rxWithSeed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/reexports.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Objects exported from other packages — reexports","text":"Inherited parent routine","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":null,"dir":"Reference","previous_headings":"","what":"Allow unloading of dlls — rxAllowUnload","title":"Allow unloading of dlls — rxAllowUnload","text":"Allow unloading dlls","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Allow unloading of dlls — rxAllowUnload","text":"","code":"rxAllowUnload(allow)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Allow unloading of dlls — rxAllowUnload","text":"allow boolean indicating garbage collection unload rxode2 dlls.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Allow unloading of dlls — rxAllowUnload","text":"Boolean allow; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Allow unloading of dlls — rxAllowUnload","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Allow unloading of dlls — rxAllowUnload","text":"","code":"# Garbage collection will not unload un-used rxode2 dlls rxAllowUnload(FALSE); #> [1] FALSE  # Garbage collection will unload unused rxode2 dlls rxAllowUnload(TRUE); #> [1] TRUE"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Append two rxui models together — rxAppendModel","title":"Append two rxui models together — rxAppendModel","text":"Append two rxui models together","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Append two rxui models together — rxAppendModel","text":"","code":"rxAppendModel(..., common = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Append two rxui models together — rxAppendModel","text":"... models append together common boolean determines need common value bind","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Append two rxui models together — rxAppendModel","text":"New model models appended together","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Append two rxui models together — rxAppendModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Append two rxui models together — rxAppendModel","text":"","code":"# \\donttest{  ocmt <- function() {   ini({     tka <- exp(0.45) # Ka     tcl <- exp(1) # Cl     tv <- exp(3.45); # log V     ## the label(\"Label name\") works with all models     add.sd <- 0.7   })   model({     ka <- tka     cl <- tcl     v <- tv     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl / v * center     cp <- center / v     cp ~ add(add.sd)   }) }  idr <- function() {   ini({     tkin <- log(1)     tkout <- log(1)     tic50 <- log(10)     gamma <- fix(1)     idr.sd <- 1   })   model({     kin <- exp(tkin)     kout <- exp(tkout)     ic50 <- exp(tic50)     d/dt(eff) <- kin - kout*(1-ceff^gamma/(ic50^gamma+ceff^gamma))     eff ~ add(idr.sd)   }) }  rxAppendModel(ocmt %>% model(ceff=cp,append=TRUE), idr) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #>   #>   #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd      tkin     tkout     tic50     gamma  #>  1.568312  2.718282 31.500392  0.700000  0.000000  0.000000  2.302585  1.000000  #>    idr.sd  #>  1.000000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3              eff #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                cmt                dvid* #> 1   cp ~ …  cmt='cp' or cmt=4  dvid='cp' or dvid=1 #> 2  eff ~ … cmt='eff' or cmt=3 dvid='eff' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>         tkin <- 0 #>         tkout <- 0 #>         tic50 <- 2.30258509299405 #>         gamma <- fix(1) #>         idr.sd <- c(0, 1) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>         ceff <- cp #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ic50 <- exp(tic50) #>         d/dt(eff) <- kin - kout * (1 - ceff^gamma/(ic50^gamma +  #>             ceff^gamma)) #>         eff ~ add(idr.sd) #>     }) #> }  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign Control Variable — rxAssignControlValue","title":"Assign Control Variable — rxAssignControlValue","text":"Assign Control Variable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign Control Variable — rxAssignControlValue","text":"","code":"rxAssignControlValue(ui, option, value)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign Control Variable — rxAssignControlValue","text":"ui rxode2 ui function option Option name control modify value Value control modify","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign Control Variable — rxAssignControlValue","text":"Nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign Control Variable — rxAssignControlValue","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign pointer based on model variables — rxAssignPtr","title":"Assign pointer based on model variables — rxAssignPtr","text":"Assign pointer based model variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign pointer based on model variables — rxAssignPtr","text":"","code":"rxAssignPtr(object = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign pointer based on model variables — rxAssignPtr","text":"object rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign pointer based on model variables — rxAssignPtr","text":"nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates a logical matrix for block matrixes. — rxBlockZeros","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"Creates logical matrix block matrixes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"","code":"rxBlockZeros(mat, i)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"mat Matrix Row/column block matrix setup.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"logical matrix returning elements zero.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the C file associated with the rxode2 object — rxC","title":"Return the C file associated with the rxode2 object — rxC","text":"return C code generating rxode2 DLL.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the C file associated with the rxode2 object — rxC","text":"","code":"rxC(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the C file associated with the rxode2 object — rxC","text":"obj rxode2 family objects character string model specification location file model specification.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the C file associated with the rxode2 object — rxC","text":"path library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the C file associated with the rxode2 object — rxC","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":null,"dir":"Reference","previous_headings":"","what":"Use cat when rxode2.verbose is TRUE — rxCat","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"Use cat rxode2.verbose TRUE","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"","code":"rxCat(a, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"... Parameters sent cat","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":null,"dir":"Reference","previous_headings":"","what":"rxChain Chain or add item to solved system of equations — rxChain","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"Add item solved system equations","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"","code":"rxChain(obj1, obj2)  # S3 method for solveRxDll +(obj1, obj2)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"obj1 Solved object. obj2 New object added/piped/chained solved object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"newObject event table, return new solved object new event table.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":null,"dir":"Reference","previous_headings":"","what":"Second command in chaining commands — rxChain2","title":"Second command in chaining commands — rxChain2","text":"s3 method called internally + \\%>\\% operators.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Second command in chaining commands — rxChain2","text":"","code":"rxChain2(obj, solvedObject)  # S3 method for default rxChain2(obj, solvedObject)  # S3 method for EventTable rxChain2(obj, solvedObject)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Second command in chaining commands — rxChain2","text":"obj object added/chained/piped solved object solvedObject solved object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Second command in chaining commands — rxChain2","text":"chained operation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Second command in chaining commands — rxChain2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":null,"dir":"Reference","previous_headings":"","what":"Cleanup anonymous DLLs by unloading them — rxClean","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"cleans rxode2 loaded DLLs","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"","code":"rxClean(wd)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"wd directory cleaned; (DEPRECIATED), longer anything. unloads rxode2 anonymous dlls.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"TRUE successful","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":null,"dir":"Reference","previous_headings":"","what":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"Combine Error Lines create rxode2 expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"","code":"rxCombineErrorLines(   uiModel,   errLines = NULL,   prefixLines = NULL,   paramsLine = NULL,   modelVars = FALSE,   cmtLines = TRUE,   dvidLine = TRUE,   lstExpr = NULL,   useIf = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"uiModel UI model errLines Error lines; missing, get error lines rxGetDistributionSimulationLines() prefixLines Prefix lines, param statement paramsLine Params line, present. modelVars Return model vars instead rxode2 statement cmtLines Include trailing cmt lines dvidLine Include trailing dvid() specification lstExpr list expressions model, NULL.  NULL defaults model expressions accessible uiModel$lstExpr. useIf Use (CMT == X) endpoints","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"quoted expression can evaluated compiled rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"exported allow functions mangle error lines make types estimation methods (needed)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"","code":"# \\donttest{  one.cmt <- function() {   ini({     ## You may label each parameter with a comment     tka <- 0.45 # Log Ka     tcl <- log(c(0, 2.7, 100)) # Log Cl     ## This works with interactive models     ## You may also label the preceding line with label(\"label text\")     tv <- 3.45; label(\"log V\")     ## the label(\"Label name\") works with all models     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     linCmt() ~ add(add.sd)   }) }  f <- rxode2(one.cmt) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()'  # You can get the simulation model easily by rxCombineErrorLines(f) #> rxode2({ #>     params(tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v) #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     rx_yj_ ~ 2 #>     rx_lambda_ ~ 1 #>     rx_low_ ~ 0 #>     rx_hi_ ~ 1 #>     rx_pred_f_ ~ linCmt() #>     rx_pred_ ~ rx_pred_f_ #>     rx_r_ ~ (add.sd)^2 #>     ipredSim <- rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>         rx_hi_) #>     sim <- rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.rxLinCmt, rx_lambda_,  #>         rx_yj_, rx_low_, rx_hi_) #>     cmt(rxLinCmt) #>     dvid(1) #> })  # You can then get the compiled model by simply evaluting the model: r <- eval(rxCombineErrorLines(f)) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # This also works with multile endpoint models: pk.turnover.emax <- function() {   ini({     tktr <- log(1)     tka <- log(1)     tcl <- log(0.1)     tv <- log(10)     ##     eta.ktr ~ 1     eta.ka ~ 1     eta.cl ~ 2     eta.v ~ 1     prop.err <- 0.1     pkadd.err <- 0.1     ##     temax <- logit(0.8)     tec50 <- log(0.5)     tkout <- log(0.05)     te0 <- log(100)     ##     eta.emax ~ .5     eta.ec50  ~ .5     eta.kout ~ .5     eta.e0 ~ .5     ##     pdadd.err <- 10   })   model({     ktr <- exp(tktr + eta.ktr)     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     ##     emax=expit(temax+eta.emax)     ec50 =  exp(tec50 + eta.ec50)     kout = exp(tkout + eta.kout)     e0 = exp(te0 + eta.e0)     ##     DCP = center/v     PD=1-emax*DCP/(ec50+DCP)     ##     effect(0) = e0     kin = e0*kout     ##     d/dt(depot) = -ktr * depot     d/dt(gut) =  ktr * depot -ka * gut     d/dt(center) =  ka * gut - cl / v * center     d/dt(effect) = kin*PD -kout*effect     ##     cp = center / v     cp ~ prop(prop.err) + add(pkadd.err)     effect ~ add(pdadd.err)   }) }  f <- rxode2(pk.turnover.emax) #>   #>   rxCombineErrorLines(f) #> rxode2({ #>     params(tktr, tka, tcl, tv, prop.err, pkadd.err, temax, tec50,  #>         tkout, te0, pdadd.err, eta.ktr, eta.ka, eta.cl, eta.v,  #>         eta.emax, eta.ec50, eta.kout, eta.e0) #>     ktr <- exp(tktr + eta.ktr) #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     emax = expit(temax + eta.emax) #>     ec50 = exp(tec50 + eta.ec50) #>     kout = exp(tkout + eta.kout) #>     e0 = exp(te0 + eta.e0) #>     DCP = center/v #>     PD = 1 - emax * DCP/(ec50 + DCP) #>     effect(0) = e0 #>     kin = e0 * kout #>     d/dt(depot) = -ktr * depot #>     d/dt(gut) = ktr * depot - ka * gut #>     d/dt(center) = ka * gut - cl/v * center #>     d/dt(effect) = kin * PD - kout * effect #>     cp = center/v #>     if (CMT == 5) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ cp #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (pkadd.err)^2 + (rx_pred_f_)^2 * (prop.err)^2 #>         ipredSim <- rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim <- rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.cp, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     if (CMT == 4) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ effect #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (pdadd.err)^2 #>         ipredSim <- rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim <- rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.effect,  #>             rx_lambda_, rx_yj_, rx_low_, rx_hi_) #>     } #>     cmt(cp) #>     dvid(5, 4) #> })  # Note that in the parsed form, you can also get the compiled rxode2 # model with $simulationModel  f$simulationModel #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_be0b5dbb4a0767d8702ddb24bbc603bc model (✔ ready).  #> $state: depot, gut, center, effect #> $stateExtra: cp #> $params: tktr, tka, tcl, tv, prop.err, pkadd.err, temax, tec50, tkout, te0, pdadd.err, eta.ktr, eta.ka, eta.cl, eta.v, eta.emax, eta.ec50, eta.kout, eta.e0, CMT, rxerr.cp, rxerr.effect #> $lhs: ktr, ka, cl, v, emax, ec50, kout, e0, DCP, PD, kin, cp, ipredSim, sim  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":null,"dir":"Reference","previous_headings":"","what":"Compile a model if needed — rxCompile","title":"Compile a model if needed — rxCompile","text":"compilation workhorse creating rxode2 model DLL files.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compile a model if needed — rxCompile","text":"","code":"rxCompile(   model,   dir,   prefix,   force = FALSE,   modName = NULL,   package = NULL,   ... )  # S3 method for rxModelVars rxCompile(   model,   dir = NULL,   prefix = NULL,   force = FALSE,   modName = NULL,   package = NULL,   ... )  # S3 method for character rxCompile(   model,   dir = NULL,   prefix = NULL,   force = FALSE,   modName = NULL,   package = NULL,   ... )  # S3 method for rxDll rxCompile(model, ...)  # S3 method for rxode2 rxCompile(model, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compile a model if needed — rxCompile","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . dir model directory C file stored compiling.   prefix string indicating prefix use C based functions.  missing, calculated based file name, md5 parsed model. force boolean stating (re)compile forced rxode2 detects models already generated. modName string used model name. string used naming various aspects computations, including generating C symbol names, dynamic libraries, etc. Therefore, necessary modName consists simple ASCII alphanumeric characters starting letter. package Package name pre-compiled binaries. ... arguments sent rxTrans() function.","code":"If unspecified, the C code is stored in a temporary directory, then the model is compiled and moved to the current directory. Afterwards the C code is removed.  If specified, the C code is stored in the specified directory and then compiled in that directory.  The C code is not removed after the DLL is created in the same directory.  This can be useful to debug the c-code outputs."},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compile a model if needed — rxCompile","text":"rxDll object following components dll DLL path model model specification .c function call C code correct context DLL using .C() function. .call function call C code correct context DLL using .Call() function. args list arguments used create rxDll object.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Compile a model if needed — rxCompile","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":null,"dir":"Reference","previous_headings":"","what":"Current Condition for rxode2 object — rxCondition","title":"Current Condition for rxode2 object — rxCondition","text":"Current Condition rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Current Condition for rxode2 object — rxCondition","text":"","code":"rxCondition(obj, condition = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Current Condition for rxode2 object — rxCondition","text":"obj rxode2 object condition specified one conditions rxode2 object (determined rxExpandIfElse()), assign rxode2 current condition parameter.  condition one known condition, condition set NULL, implying conditioning currently used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Current Condition for rxode2 object — rxCondition","text":"Current condition rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Current Condition for rxode2 object — rxCondition","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":null,"dir":"Reference","previous_headings":"","what":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"assumes normal ODE equations first equations ODE expanded forward sensitivities type sensitivity (like adjoint)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"","code":"rxControlUpdateSens(rxControl, sensCmt = NULL, ncmt = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"rxControl Input list rxControl type list sensCmt Number sensitivity compartments ncmt Number compartments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"Updated rxControl $atol, $rtol, $ssAtol $ssRtol updated different sensitivities normal ODEs (first) different sensitivity larger compartments (sensitivities).","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"","code":"tmp <- rxControl()  tmp2 <- rxControlUpdateSens(tmp, 3, 6)  tmp2$atol #> [1] 1e-08 1e-08 1e-08 1e-08 1e-08 1e-08 tmp2$rtol #> [1] 1e-06 1e-06 1e-06 1e-06 1e-06 1e-06 tmp2$ssAtol #> [1] 1e-08 1e-08 1e-08 1e-08 1e-08 1e-08 tmp2$ssRtol #> [1] 1e-06 1e-06 1e-06 1e-06 1e-06 1e-06"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":null,"dir":"Reference","previous_headings":"","what":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"run, R_user_dir rxode2's cache present, create cache","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"","code":"rxCreateCache()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":null,"dir":"Reference","previous_headings":"","what":"Add to rxode2's derivative tables — rxD","title":"Add to rxode2's derivative tables — rxD","text":"Add rxode2's derivative tables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add to rxode2's derivative tables — rxD","text":"","code":"rxD(name, derivatives)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add to rxode2's derivative tables — rxD","text":"name Function Name derivatives list functions. function takes number arguments original function.  first function construct derivative respect first argument; second function construct derivitive respect second argument, .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add to rxode2's derivative tables — rxD","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Add to rxode2's derivative tables — rxD","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add to rxode2's derivative tables — rxD","text":"","code":"## Add an arbitrary list of derivative functions ## In this case the fun(x,y) is assumed to be 0.5*x^2+0.5*y^2  rxD(\"fun\", list(   function(x, y) {     return(x)   },   function(x, y) {     return(y)   } ))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":null,"dir":"Reference","previous_headings":"","what":"Delete the DLL for the model — rxDelete","title":"Delete the DLL for the model — rxDelete","text":"function deletes DLL, delete model information object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Delete the DLL for the model — rxDelete","text":"","code":"rxDelete(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Delete the DLL for the model — rxDelete","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Delete the DLL for the model — rxDelete","text":"boolean stating operation successful.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Delete the DLL for the model — rxDelete","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":null,"dir":"Reference","previous_headings":"","what":"Demote the error type — rxDemoteAddErr","title":"Demote the error type — rxDemoteAddErr","text":"Demote error type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Demote the error type — rxDemoteAddErr","text":"","code":"rxDemoteAddErr(errType)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Demote the error type — rxDemoteAddErr","text":"errType Error type factor","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Demote the error type — rxDemoteAddErr","text":"Demoted Error Type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Demote the error type — rxDemoteAddErr","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Demote the error type — rxDemoteAddErr","text":"","code":"rxErrTypeCombine(\"add\") %>%   rxErrTypeCombine(\"prop\") #> $transform #> [1] untransformed #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] add + prop #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] untransformed #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\"  # This removes the internal additive error rxErrTypeCombine(\"add\") %>%   rxErrTypeCombine(\"prop\") %>%   rxDemoteAddErr() #> $transform #> [1] untransformed #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] prop #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] untransformed #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\"  # This is used for logitNorm(NA), the additive portion is stripped"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":null,"dir":"Reference","previous_headings":"","what":"Jacobian and parameter derivatives — rxDfdy","title":"Jacobian and parameter derivatives — rxDfdy","text":"Return Jacobain parameter derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Jacobian and parameter derivatives — rxDfdy","text":"","code":"rxDfdy(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Jacobian and parameter derivatives — rxDfdy","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Jacobian and parameter derivatives — rxDfdy","text":"list jacobian parameters defined rxode2 object.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Jacobian and parameter derivatives — rxDfdy","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the DLL associated with the rxode2 object — rxDll","title":"Return the DLL associated with the rxode2 object — rxDll","text":"return dynamic load library shared object used run C code rxode2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the DLL associated with the rxode2 object — rxDll","text":"","code":"rxDll(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the DLL associated with the rxode2 object — rxDll","text":"obj rxode2 family objects character string model specification location file model specification.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the DLL associated with the rxode2 object — rxDll","text":"path library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the DLL associated with the rxode2 object — rxDll","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":null,"dir":"Reference","previous_headings":"","what":"Load rxode2 object — rxDynLoad","title":"Load rxode2 object — rxDynLoad","text":"Load rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load rxode2 object — rxDynLoad","text":"","code":"rxDynLoad(obj)  rxLoad(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load rxode2 object — rxDynLoad","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load rxode2 object — rxDynLoad","text":"Boolean returning rxode2 library loaded.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Load rxode2 object — rxDynLoad","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":null,"dir":"Reference","previous_headings":"","what":"Unload rxode2 object — rxDynUnload","title":"Unload rxode2 object — rxDynUnload","text":"Unload rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Unload rxode2 object — rxDynUnload","text":"","code":"rxDynUnload(obj)  rxUnload(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Unload rxode2 object — rxDynUnload","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Unload rxode2 object — rxDynUnload","text":"Boolean returning rxode2 library loaded.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Unload rxode2 object — rxDynUnload","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":null,"dir":"Reference","previous_headings":"","what":"Combine transformations and error structures — rxErrTypeCombine","title":"Combine transformations and error structures — rxErrTypeCombine","text":"Combine error information figure transformation applied current endpoint","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Combine transformations and error structures — rxErrTypeCombine","text":"","code":"rxErrTypeCombine(oldErrType, newErrType)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Combine transformations and error structures — rxErrTypeCombine","text":"oldErrType old transformation, default zero representing prior transformation. parameter first allow piping. parameter addTransform missing oldErrType character value, functions swaps oldErrType addTransform assigns oldErrType zero assuming prior distribution. newErrType new distribution \"added\" current transformation.  assumes inputs preferred distribution name, determined rxPreferredDistributionName()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Combine transformations and error structures — rxErrTypeCombine","text":"new transformation factor","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Combine transformations and error structures — rxErrTypeCombine","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Combine transformations and error structures — rxErrTypeCombine","text":"","code":"rxErrTypeCombine(\"probitNorm\") #> $transform #> [1] probit #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] add #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] none #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\"  rxErrTypeCombine(\"probitNorm\") %>%   rxErrTypeCombine(\"boxCox\") #> $transform #> [1] probit + boxCox #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] add #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] none #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand d(f)/d(eta) — rxExpandFEta_","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"Expand d(f)/d(eta)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"","code":"rxExpandFEta_(state, neta, pred, isTheta = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"state state expand neta number etas pred type prediction isTheta logical, expansion actually thetas instead etas","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"String symengine expressions evaluate calculate df/deta","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":null,"dir":"Reference","previous_headings":"","what":"Faster expand.grid — rxExpandGrid","title":"Faster expand.grid — rxExpandGrid","text":"support x y characters right now","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Faster expand.grid — rxExpandGrid","text":"","code":"rxExpandGrid(x, y, type = 0L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Faster expand.grid — rxExpandGrid","text":"x first element (must character) y second element (must character) type Internal type=0L traditional expand grid type=1L Jacobian expand grid (adds symbols)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Faster expand.grid — rxExpandGrid","text":"Expand grid","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Faster expand.grid — rxExpandGrid","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Faster expand.grid — rxExpandGrid","text":"","code":"## rxExpandGrid(letters, letters) #>     Var1 Var2 #> 1      a    a #> 2      b    a #> 3      c    a #> 4      d    a #> 5      e    a #> 6      f    a #> 7      g    a #> 8      h    a #> 9      i    a #> 10     j    a #> 11     k    a #> 12     l    a #> 13     m    a #> 14     n    a #> 15     o    a #> 16     p    a #> 17     q    a #> 18     r    a #> 19     s    a #> 20     t    a #> 21     u    a #> 22     v    a #> 23     w    a #> 24     x    a #> 25     y    a #> 26     z    a #> 27     a    b #> 28     b    b #> 29     c    b #> 30     d    b #> 31     e    b #> 32     f    b #> 33     g    b #> 34     h    b #> 35     i    b #> 36     j    b #> 37     k    b #> 38     l    b #> 39     m    b #> 40     n    b #> 41     o    b #> 42     p    b #> 43     q    b #> 44     r    b #> 45     s    b #> 46     t    b #> 47     u    b #> 48     v    b #> 49     w    b #> 50     x    b #> 51     y    b #> 52     z    b #> 53     a    c #> 54     b    c #> 55     c    c #> 56     d    c #> 57     e    c #> 58     f    c #> 59     g    c #> 60     h    c #> 61     i    c #> 62     j    c #> 63     k    c #> 64     l    c #> 65     m    c #> 66     n    c #> 67     o    c #> 68     p    c #> 69     q    c #> 70     r    c #> 71     s    c #> 72     t    c #> 73     u    c #> 74     v    c #> 75     w    c #> 76     x    c #> 77     y    c #> 78     z    c #> 79     a    d #> 80     b    d #> 81     c    d #> 82     d    d #> 83     e    d #> 84     f    d #> 85     g    d #> 86     h    d #> 87     i    d #> 88     j    d #> 89     k    d #> 90     l    d #> 91     m    d #> 92     n    d #> 93     o    d #> 94     p    d #> 95     q    d #> 96     r    d #> 97     s    d #> 98     t    d #> 99     u    d #> 100    v    d #> 101    w    d #> 102    x    d #> 103    y    d #> 104    z    d #> 105    a    e #> 106    b    e #> 107    c    e #> 108    d    e #> 109    e    e #> 110    f    e #> 111    g    e #> 112    h    e #> 113    i    e #> 114    j    e #> 115    k    e #> 116    l    e #> 117    m    e #> 118    n    e #> 119    o    e #> 120    p    e #> 121    q    e #> 122    r    e #> 123    s    e #> 124    t    e #> 125    u    e #> 126    v    e #> 127    w    e #> 128    x    e #> 129    y    e #> 130    z    e #> 131    a    f #> 132    b    f #> 133    c    f #> 134    d    f #> 135    e    f #> 136    f    f #> 137    g    f #> 138    h    f #> 139    i    f #> 140    j    f #> 141    k    f #> 142    l    f #> 143    m    f #> 144    n    f #> 145    o    f #> 146    p    f #> 147    q    f #> 148    r    f #> 149    s    f #> 150    t    f #> 151    u    f #> 152    v    f #> 153    w    f #> 154    x    f #> 155    y    f #> 156    z    f #> 157    a    g #> 158    b    g #> 159    c    g #> 160    d    g #> 161    e    g #> 162    f    g #> 163    g    g #> 164    h    g #> 165    i    g #> 166    j    g #> 167    k    g #> 168    l    g #> 169    m    g #> 170    n    g #> 171    o    g #> 172    p    g #> 173    q    g #> 174    r    g #> 175    s    g #> 176    t    g #> 177    u    g #> 178    v    g #> 179    w    g #> 180    x    g #> 181    y    g #> 182    z    g #> 183    a    h #> 184    b    h #> 185    c    h #> 186    d    h #> 187    e    h #> 188    f    h #> 189    g    h #> 190    h    h #> 191    i    h #> 192    j    h #> 193    k    h #> 194    l    h #> 195    m    h #> 196    n    h #> 197    o    h #> 198    p    h #> 199    q    h #> 200    r    h #> 201    s    h #> 202    t    h #> 203    u    h #> 204    v    h #> 205    w    h #> 206    x    h #> 207    y    h #> 208    z    h #> 209    a    i #> 210    b    i #> 211    c    i #> 212    d    i #> 213    e    i #> 214    f    i #> 215    g    i #> 216    h    i #> 217    i    i #> 218    j    i #> 219    k    i #> 220    l    i #> 221    m    i #> 222    n    i #> 223    o    i #> 224    p    i #> 225    q    i #> 226    r    i #> 227    s    i #> 228    t    i #> 229    u    i #> 230    v    i #> 231    w    i #> 232    x    i #> 233    y    i #> 234    z    i #> 235    a    j #> 236    b    j #> 237    c    j #> 238    d    j #> 239    e    j #> 240    f    j #> 241    g    j #> 242    h    j #> 243    i    j #> 244    j    j #> 245    k    j #> 246    l    j #> 247    m    j #> 248    n    j #> 249    o    j #> 250    p    j #> 251    q    j #> 252    r    j #> 253    s    j #> 254    t    j #> 255    u    j #> 256    v    j #> 257    w    j #> 258    x    j #> 259    y    j #> 260    z    j #> 261    a    k #> 262    b    k #> 263    c    k #> 264    d    k #> 265    e    k #> 266    f    k #> 267    g    k #> 268    h    k #> 269    i    k #> 270    j    k #> 271    k    k #> 272    l    k #> 273    m    k #> 274    n    k #> 275    o    k #> 276    p    k #> 277    q    k #> 278    r    k #> 279    s    k #> 280    t    k #> 281    u    k #> 282    v    k #> 283    w    k #> 284    x    k #> 285    y    k #> 286    z    k #> 287    a    l #> 288    b    l #> 289    c    l #> 290    d    l #> 291    e    l #> 292    f    l #> 293    g    l #> 294    h    l #> 295    i    l #> 296    j    l #> 297    k    l #> 298    l    l #> 299    m    l #> 300    n    l #> 301    o    l #> 302    p    l #> 303    q    l #> 304    r    l #> 305    s    l #> 306    t    l #> 307    u    l #> 308    v    l #> 309    w    l #> 310    x    l #> 311    y    l #> 312    z    l #> 313    a    m #> 314    b    m #> 315    c    m #> 316    d    m #> 317    e    m #> 318    f    m #> 319    g    m #> 320    h    m #> 321    i    m #> 322    j    m #> 323    k    m #> 324    l    m #> 325    m    m #> 326    n    m #> 327    o    m #> 328    p    m #> 329    q    m #> 330    r    m #> 331    s    m #> 332    t    m #> 333    u    m #> 334    v    m #> 335    w    m #> 336    x    m #> 337    y    m #> 338    z    m #> 339    a    n #> 340    b    n #> 341    c    n #> 342    d    n #> 343    e    n #> 344    f    n #> 345    g    n #> 346    h    n #> 347    i    n #> 348    j    n #> 349    k    n #> 350    l    n #> 351    m    n #> 352    n    n #> 353    o    n #> 354    p    n #> 355    q    n #> 356    r    n #> 357    s    n #> 358    t    n #> 359    u    n #> 360    v    n #> 361    w    n #> 362    x    n #> 363    y    n #> 364    z    n #> 365    a    o #> 366    b    o #> 367    c    o #> 368    d    o #> 369    e    o #> 370    f    o #> 371    g    o #> 372    h    o #> 373    i    o #> 374    j    o #> 375    k    o #> 376    l    o #> 377    m    o #> 378    n    o #> 379    o    o #> 380    p    o #> 381    q    o #> 382    r    o #> 383    s    o #> 384    t    o #> 385    u    o #> 386    v    o #> 387    w    o #> 388    x    o #> 389    y    o #> 390    z    o #> 391    a    p #> 392    b    p #> 393    c    p #> 394    d    p #> 395    e    p #> 396    f    p #> 397    g    p #> 398    h    p #> 399    i    p #> 400    j    p #> 401    k    p #> 402    l    p #> 403    m    p #> 404    n    p #> 405    o    p #> 406    p    p #> 407    q    p #> 408    r    p #> 409    s    p #> 410    t    p #> 411    u    p #> 412    v    p #> 413    w    p #> 414    x    p #> 415    y    p #> 416    z    p #> 417    a    q #> 418    b    q #> 419    c    q #> 420    d    q #> 421    e    q #> 422    f    q #> 423    g    q #> 424    h    q #> 425    i    q #> 426    j    q #> 427    k    q #> 428    l    q #> 429    m    q #> 430    n    q #> 431    o    q #> 432    p    q #> 433    q    q #> 434    r    q #> 435    s    q #> 436    t    q #> 437    u    q #> 438    v    q #> 439    w    q #> 440    x    q #> 441    y    q #> 442    z    q #> 443    a    r #> 444    b    r #> 445    c    r #> 446    d    r #> 447    e    r #> 448    f    r #> 449    g    r #> 450    h    r #> 451    i    r #> 452    j    r #> 453    k    r #> 454    l    r #> 455    m    r #> 456    n    r #> 457    o    r #> 458    p    r #> 459    q    r #> 460    r    r #> 461    s    r #> 462    t    r #> 463    u    r #> 464    v    r #> 465    w    r #> 466    x    r #> 467    y    r #> 468    z    r #> 469    a    s #> 470    b    s #> 471    c    s #> 472    d    s #> 473    e    s #> 474    f    s #> 475    g    s #> 476    h    s #> 477    i    s #> 478    j    s #> 479    k    s #> 480    l    s #> 481    m    s #> 482    n    s #> 483    o    s #> 484    p    s #> 485    q    s #> 486    r    s #> 487    s    s #> 488    t    s #> 489    u    s #> 490    v    s #> 491    w    s #> 492    x    s #> 493    y    s #> 494    z    s #> 495    a    t #> 496    b    t #> 497    c    t #> 498    d    t #> 499    e    t #> 500    f    t #> 501    g    t #> 502    h    t #> 503    i    t #> 504    j    t #> 505    k    t #> 506    l    t #> 507    m    t #> 508    n    t #> 509    o    t #> 510    p    t #> 511    q    t #> 512    r    t #> 513    s    t #> 514    t    t #> 515    u    t #> 516    v    t #> 517    w    t #> 518    x    t #> 519    y    t #> 520    z    t #> 521    a    u #> 522    b    u #> 523    c    u #> 524    d    u #> 525    e    u #> 526    f    u #> 527    g    u #> 528    h    u #> 529    i    u #> 530    j    u #> 531    k    u #> 532    l    u #> 533    m    u #> 534    n    u #> 535    o    u #> 536    p    u #> 537    q    u #> 538    r    u #> 539    s    u #> 540    t    u #> 541    u    u #> 542    v    u #> 543    w    u #> 544    x    u #> 545    y    u #> 546    z    u #> 547    a    v #> 548    b    v #> 549    c    v #> 550    d    v #> 551    e    v #> 552    f    v #> 553    g    v #> 554    h    v #> 555    i    v #> 556    j    v #> 557    k    v #> 558    l    v #> 559    m    v #> 560    n    v #> 561    o    v #> 562    p    v #> 563    q    v #> 564    r    v #> 565    s    v #> 566    t    v #> 567    u    v #> 568    v    v #> 569    w    v #> 570    x    v #> 571    y    v #> 572    z    v #> 573    a    w #> 574    b    w #> 575    c    w #> 576    d    w #> 577    e    w #> 578    f    w #> 579    g    w #> 580    h    w #> 581    i    w #> 582    j    w #> 583    k    w #> 584    l    w #> 585    m    w #> 586    n    w #> 587    o    w #> 588    p    w #> 589    q    w #> 590    r    w #> 591    s    w #> 592    t    w #> 593    u    w #> 594    v    w #> 595    w    w #> 596    x    w #> 597    y    w #> 598    z    w #> 599    a    x #> 600    b    x #> 601    c    x #> 602    d    x #> 603    e    x #> 604    f    x #> 605    g    x #> 606    h    x #> 607    i    x #> 608    j    x #> 609    k    x #> 610    l    x #> 611    m    x #> 612    n    x #> 613    o    x #> 614    p    x #> 615    q    x #> 616    r    x #> 617    s    x #> 618    t    x #> 619    u    x #> 620    v    x #> 621    w    x #> 622    x    x #> 623    y    x #> 624    z    x #> 625    a    y #> 626    b    y #> 627    c    y #> 628    d    y #> 629    e    y #> 630    f    y #> 631    g    y #> 632    h    y #> 633    i    y #> 634    j    y #> 635    k    y #> 636    l    y #> 637    m    y #> 638    n    y #> 639    o    y #> 640    p    y #> 641    q    y #> 642    r    y #> 643    s    y #> 644    t    y #> 645    u    y #> 646    v    y #> 647    w    y #> 648    x    y #> 649    y    y #> 650    z    y #> 651    a    z #> 652    b    z #> 653    c    z #> 654    d    z #> 655    e    z #> 656    f    z #> 657    g    z #> 658    h    z #> 659    i    z #> 660    j    z #> 661    k    z #> 662    l    z #> 663    m    z #> 664    n    z #> 665    o    z #> 666    p    z #> 667    q    z #> 668    r    z #> 669    s    z #> 670    t    z #> 671    u    z #> 672    v    z #> 673    w    z #> 674    x    z #> 675    y    z #> 676    z    z  ## Another fast method; See ## https://stackoverflow.com/questions/10405637/use-outer-instead-of-expand-grid  expand.grid.jc <- function(seq1, seq2) {   cbind(     Var1 = rep.int(seq1, length(seq2)),     Var2 = rep.int(seq2, rep.int(length(seq1), length(seq2)))   ) } # \\donttest{ microbenchmark::microbenchmark(rxExpandGrid(letters, letters), expand.grid.jc(letters, letters)) #> Unit: microseconds #>                              expr   min      lq      mean  median     uq #>    rxExpandGrid(letters, letters) 10.80 11.1305  45.96206 11.3510 11.602 #>  expand.grid.jc(letters, letters) 16.07 16.7410 193.97223 17.8285 18.640 #>       max neval #>  3295.638   100 #>  4231.945   100 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand grid internal function — rxExpandGrid_","title":"Expand grid internal function — rxExpandGrid_","text":"Expand grid internal function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand grid internal function — rxExpandGrid_","text":"","code":"rxExpandGrid_(c1, c2, type)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand grid internal function — rxExpandGrid_","text":"c1 character vector items expanded c2 second character vector items expanded type 0 typical data frame, 1 symengine sensitivity expansion","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand grid internal function — rxExpandGrid_","text":"data frame (type = 0) symengine string (type=1)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"Expand /else clauses multiple different types lines.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"","code":"rxExpandIfElse(model, removeInis = TRUE, removePrint = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"model Model can character, rxode2 model.  needs normalized syntax, (...){} line.  else statement must line closing bracket statement previous line.  else statment must also contain opening bracket, like code else {} removeInis boolean indicating parameter initializations removed model. removePrint boolean indicating printing statements removed model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"named character vector. names vector logical conditions, values lines satisfy logical conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand second order sensitivity — rxExpandSens2_","title":"Expand second order sensitivity — rxExpandSens2_","text":"Expand second order sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand second order sensitivity — rxExpandSens2_","text":"","code":"rxExpandSens2_(state, s1, s2)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand second order sensitivity — rxExpandSens2_","text":"state state expand s1 variables calculate sensitivity s2 variables calculate second order sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand second order sensitivity — rxExpandSens2_","text":"string symengine second order sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand sensitivity — rxExpandSens_","title":"Expand sensitivity — rxExpandSens_","text":"Expand sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand sensitivity — rxExpandSens_","text":"","code":"rxExpandSens_(state, calcSens)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand sensitivity — rxExpandSens_","text":"state state expand calcSens variables calculate sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand sensitivity — rxExpandSens_","text":"symengine string expanded sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":null,"dir":"Reference","previous_headings":"","what":"Apply the fixed population estimated parameters — rxFixPop","title":"Apply the fixed population estimated parameters — rxFixPop","text":"Apply fixed population estimated parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Apply the fixed population estimated parameters — rxFixPop","text":"","code":"rxFixPop(ui, returnNull = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Apply the fixed population estimated parameters — rxFixPop","text":"ui rxode2 ui function returnNull boolean unchanged values return original ui (FALSE) null (TRUE)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Apply the fixed population estimated parameters — rxFixPop","text":"returnNull TRUE, NULL nothing changed, changed model ui.  returnNull FALSE, return ui matter changed .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Apply the fixed population estimated parameters — rxFixPop","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Apply the fixed population estimated parameters — rxFixPop","text":"","code":"One.comp.transit.allo <- function() {  ini({    # Where initial conditions/variables are specified    lktr <- log(1.15)  #log k transit (/h)    lcl  <- log(0.15)  #log Cl (L/hr)    lv   <- log(7)     #log V (L)    ALLC <- fix(0.75)  #allometric exponent cl    ALLV <- fix(1.00)  #allometric exponent v    prop.err <- 0.15   #proportional error (SD/mean)    add.err <- 0.6     #additive error (mg/L)    eta.ktr ~ 0.5    eta.cl ~ 0.1    eta.v ~ 0.1  })  model({    #Allometric scaling on weight    cl <- exp(lcl + eta.cl + ALLC * logWT70)    v  <- exp(lv + eta.v + ALLV * logWT70)    ktr <- exp(lktr + eta.ktr)    # RxODE-style differential equations are supported    d/dt(depot)   = -ktr * depot    d/dt(central) =  ktr * trans - (cl/v) * central    d/dt(trans)   =  ktr * depot - ktr * trans    ## Concentration is calculated    cp = central/v    # And is assumed to follow proportional and additive error    cp ~ prop(prop.err) + add(add.err)  }) }  m <- rxFixPop(One.comp.transit.allo) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' m #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       lktr        lcl         lv   prop.err    add.err  #>  0.1397619 -1.8971200  1.9459101  0.1500000  0.6000000  #>  #> Omega ($omega):  #>         eta.ktr eta.cl eta.v #> eta.ktr     0.5    0.0   0.0 #> eta.cl      0.0    0.1   0.0 #> eta.v       0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2          central #> 3                  3            trans #>  ── μ-referencing ($muRefTable): ──   #>   theta     eta level #> 1   lcl  eta.cl    id #> 2    lv   eta.v    id #> 3  lktr eta.ktr    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         lktr <- 0.139761942375159 #>         label(\"log k transit (/h)\") #>         lcl <- -1.89711998488588 #>         label(\"log Cl (L/hr)\") #>         lv <- 1.94591014905531 #>         label(\"log V (L)\") #>         prop.err <- c(0, 0.15) #>         label(\"proportional error (SD/mean)\") #>         add.err <- c(0, 0.6) #>         label(\"additive error (mg/L)\") #>         eta.ktr ~ 0.5 #>         eta.cl ~ 0.1 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         cl <- exp(lcl + eta.cl + 0.75 * logWT70) #>         v <- exp(lv + eta.v + 1 * logWT70) #>         ktr <- exp(lktr + eta.ktr) #>         d/dt(depot) = -ktr * depot #>         d/dt(central) = ktr * trans - (cl/v) * central #>         d/dt(trans) = ktr * depot - ktr * trans #>         cp = central/v #>         cp ~ prop(prop.err) + add(add.err) #>     }) #> }  # now everything is already fixed, so calling again will do nothing  rxFixPop(m) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       lktr        lcl         lv   prop.err    add.err  #>  0.1397619 -1.8971200  1.9459101  0.1500000  0.6000000  #>  #> Omega ($omega):  #>         eta.ktr eta.cl eta.v #> eta.ktr     0.5    0.0   0.0 #> eta.cl      0.0    0.1   0.0 #> eta.v       0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2          central #> 3                  3            trans #>  ── μ-referencing ($muRefTable): ──   #>   theta     eta level #> 1   lcl  eta.cl    id #> 2    lv   eta.v    id #> 3  lktr eta.ktr    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         lktr <- 0.139761942375159 #>         label(\"log k transit (/h)\") #>         lcl <- -1.89711998488588 #>         label(\"log Cl (L/hr)\") #>         lv <- 1.94591014905531 #>         label(\"log V (L)\") #>         prop.err <- c(0, 0.15) #>         label(\"proportional error (SD/mean)\") #>         add.err <- c(0, 0.6) #>         label(\"additive error (mg/L)\") #>         eta.ktr ~ 0.5 #>         eta.cl ~ 0.1 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         cl <- exp(lcl + eta.cl + 0.75 * logWT70) #>         v <- exp(lv + eta.v + 1 * logWT70) #>         ktr <- exp(lktr + eta.ktr) #>         d/dt(depot) = -ktr * depot #>         d/dt(central) = ktr * trans - (cl/v) * central #>         d/dt(trans) = ktr * depot - ktr * trans #>         cp = central/v #>         cp ~ prop(prop.err) + add(add.err) #>     }) #> }  # if you call it with returnNull=TRUE when no changes have been # performed, the function will return NULL  rxFixPop(m, returnNull=TRUE) #> NULL"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":null,"dir":"Reference","previous_headings":"","what":"Clear memoise cache for rxode2 — rxForget","title":"Clear memoise cache for rxode2 — rxForget","text":"Clear memoise cache rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Clear memoise cache for rxode2 — rxForget","text":"","code":"rxForget()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Clear memoise cache for rxode2 — rxForget","text":"nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Clear memoise cache for rxode2 — rxForget","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":null,"dir":"Reference","previous_headings":"","what":"Add/Create C functions for use in rxode2 — rxFun","title":"Add/Create C functions for use in rxode2 — rxFun","text":"Add/Create C functions use rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add/Create C functions for use in rxode2 — rxFun","text":"","code":"rxFun(name, args, cCode)  rxRmFun(name)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add/Create C functions for use in rxode2 — rxFun","text":"name can either give name user function simple R function wish convert C.  rxode2 convert R function C, name function match function name provided number arguments match R function provided.  Hence, providing R function conversion C, rest arguments implied. args gives arguments user function cCode C-code new function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add/Create C functions for use in rxode2 — rxFun","text":"","code":"# \\donttest{ ## Right now rxode2 is not aware of the function fun ## Therefore it cannot translate it to symengine or ## Compile a model with it.  try(rxode2(\"a=fun(a,b,c)\")) #>   #>   #> Error : syntax errors (see above)  ## Note for this approach to work, it cannot interfere with C ## function names or reserved rxode2 special terms.  Therefore ## f(x) would not work since f is an alias for bioavailability.  fun <- \" double fun(double a, double b, double c) {   return a*a+b*a+c; } \" ## C-code for function  rxFun(\"fun\", c(\"a\", \"b\", \"c\"), fun) ## Added function  ## Now rxode2 knows how to translate this function to symengine  rxToSE(\"fun(a,b,c)\") #> [1] \"fun(a,b,c)\"  ## And will take a central difference when calculating derivatives  rxFromSE(\"Derivative(fun(a,b,c),a)\") #> [1] \"(fun((a)+6.05545445239334e-06,b,c)-fun(a,b,c))/6.05545445239334e-06\"  ## Of course, you could specify the derivative table manually rxD(\"fun\", list(   function(a, b, c) {     paste0(\"2*\", a, \"+\", b)   },   function(a, b, c) {     return(a)   },   function(a, b, c) {     return(\"0.0\")   } ))  rxFromSE(\"Derivative(fun(a,b,c),a)\") #> [1] \"2*a+b\"  # You can also remove the functions by `rxRmFun`  rxRmFun(\"fun\")  # you can also use R functions directly in rxode2   gg <- function(x, y) {   x + y }  f <- rxode2({  z = gg(x, y) }) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’   e <- et(1:10) |> as.data.frame()  e$x <- 1:10 e$y <- 21:30  rxSolve(f, e) #> Warning: not thread safe method, using 1 core #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time     z     x     y #>   <dbl> <dbl> <dbl> <dbl> #> 1     1    22     1    21 #> 2     2    24     2    22 #> 3     3    26     3    23 #> 4     4    28     4    24 #> 5     5    30     5    25 #> 6     6    32     6    26 #> # ℹ 4 more rows  # Note that since it touches R, it can only run single-threaded. # There are also requirements for the function: # # 1. It accepts one value per argument (numeric) # # 2. It returns one numeric value  # If it is a simple function (like gg) you can also convert it to C # using rxFun and load it into rxode2  rxFun(gg) #> → finding duplicate expressions in d(gg)/d(x)... #> → finding duplicate expressions in d(gg)/d(y)... #> converted R function 'gg' to C (will now use in rxode2) #> converted R function 'rx_gg_d_x' to C (will now use in rxode2) #> converted R function 'rx_gg_d_y' to C (will now use in rxode2) #> Added derivative table for 'gg'  rxSolve(f, e) #> compiled with R user function 'gg'; now there is a clashing C user function #> triggered a recompile to use the C user function (they are always preferred) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time     z     x     y #>   <dbl> <dbl> <dbl> <dbl> #> 1     1    22     1    21 #> 2     2    24     2    22 #> 3     3    26     3    23 #> 4     4    28     4    24 #> 5     5    30     5    25 #> 6     6    32     6    26 #> # ℹ 4 more rows  # to stop the recompile simply reassign the function f <- rxode2(f) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  rxSolve(f, e) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time     z     x     y #>   <dbl> <dbl> <dbl> <dbl> #> 1     1    22     1    21 #> 2     2    24     2    22 #> 3     3    26     3    23 #> 4     4    28     4    24 #> 5     5    30     5    25 #> 6     6    32     6    26 #> # ℹ 4 more rows  rxRmFun(\"gg\") rm(gg) rm(f)   # You can also automatically convert a R function to R code (and # calculate first derivatives)  fun <- function(a, b, c) {   a^2+b*a+c }  rxFun(fun) #> → finding duplicate expressions in d(fun)/d(a)... #> → finding duplicate expressions in d(fun)/d(b)... #> → finding duplicate expressions in d(fun)/d(c)... #> converted R function 'fun' to C (will now use in rxode2) #> converted R function 'rx_fun_d_a' to C (will now use in rxode2) #> converted R function 'rx_fun_d_b' to C (will now use in rxode2) #> converted R function 'rx_fun_d_c' to C (will now use in rxode2) #> Added derivative table for 'fun'  # You can see the R code if you want with rxC  message(rxC(\"fun\")) #> double fun(double  a, double  b, double  c) { #>   double _lastValue=NA_REAL; #>   _lastValue = R_pow_di(a,2)+b*a+c; #>   return _lastValue; #> }  # you can also remove both the function and the # derivatives with rxRmFun(\"fun\")  rxRmFun(\"fun\")   # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":null,"dir":"Reference","previous_headings":"","what":"rxGetControl option from ui — rxGetControl","title":"rxGetControl option from ui — rxGetControl","text":"rxGetControl option ui","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxGetControl option from ui — rxGetControl","text":"","code":"rxGetControl(ui, option, default)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxGetControl option from ui — rxGetControl","text":"ui rxode2 ui object option Option get default Default value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxGetControl option from ui — rxGetControl","text":"Option (present) default value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxGetControl option from ui — rxGetControl","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":null,"dir":"Reference","previous_headings":"","what":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"S3 method getting distribution lines rxode2 simulation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"","code":"rxGetDistributionSimulationLines(line)  # S3 method for norm rxGetDistributionSimulationLines(line)  # S3 method for dnorm rxGetDistributionSimulationLines(line)  # S3 method for t rxGetDistributionSimulationLines(line)  # S3 method for cauchy rxGetDistributionSimulationLines(line)  # S3 method for ordinal rxGetDistributionSimulationLines(line)  # S3 method for default rxGetDistributionSimulationLines(line)  # S3 method for rxUi rxGetDistributionSimulationLines(line)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"line Parsed rxode2 model environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"Lines simulation ipred dv. based idea focei parameters defined","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the linear compartment model true function — rxGetLin","title":"Get the linear compartment model true function — rxGetLin","text":"Get linear compartment model true function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the linear compartment model true function — rxGetLin","text":"","code":"rxGetLin(   model,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   verbose = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the linear compartment model true function — rxGetLin","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . linCmtSens method calculate linCmt() solutions verbose TRUE verbose linear compartmental model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the linear compartment model true function — rxGetLin","text":"model linCmt() replaced linCmtA()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the linear compartment model true function — rxGetLin","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get model properties without compiling it. — rxGetModel","title":"Get model properties without compiling it. — rxGetModel","text":"Get model properties without compiling .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get model properties without compiling it. — rxGetModel","text":"","code":"rxGetModel(   model,   calcSens = NULL,   calcJac = NULL,   collapseModel = NULL,   indLin = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get model properties without compiling it. — rxGetModel","text":"model rxode2 specification calcSens boolean indicating rxode2 calculate sensitivities according specified ODEs. calcJac boolean indicating rxode2 calculate Jacobain according specified ODEs. collapseModel boolean indicating rxode2 remove LHS variables calculating sensitivities. indLin Calculate inductive linearization matrices compile inductive linearization support.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get model properties without compiling it. — rxGetModel","text":"rxode2 trans list","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get model properties without compiling it. — rxGetModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Get rxode2 model from object — rxGetrxode2","title":"Get rxode2 model from object — rxGetrxode2","text":"Get rxode2 model object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get rxode2 model from object — rxGetrxode2","text":"","code":"rxGetrxode2(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get rxode2 model from object — rxGetrxode2","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get rxode2 model from object — rxGetrxode2","text":"rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":null,"dir":"Reference","previous_headings":"","what":"Format rxSolve and related objects as html. — rxHtml","title":"Format rxSolve and related objects as html. — rxHtml","text":"Format rxSolve related objects html.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Format rxSolve and related objects as html. — rxHtml","text":"","code":"rxHtml(x, ...)  # S3 method for rxSolve rxHtml(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Format rxSolve and related objects as html. — rxHtml","text":"x rxode2 object ... Extra arguments sent kable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Format rxSolve and related objects as html. — rxHtml","text":"html code rxSolve object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Format rxSolve and related objects as html. — rxHtml","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":null,"dir":"Reference","previous_headings":"","what":"Set the preferred factoring by state — rxIndLinState","title":"Set the preferred factoring by state — rxIndLinState","text":"Set preferred factoring state","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set the preferred factoring by state — rxIndLinState","text":"","code":"rxIndLinState(preferred = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set the preferred factoring by state — rxIndLinState","text":"preferred list state's preferred factorization","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set the preferred factoring by state — rxIndLinState","text":"Nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set the preferred factoring by state — rxIndLinState","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":null,"dir":"Reference","previous_headings":"","what":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"one state ODE separated specifies incorporated matrix exponential.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"","code":"rxIndLinStrategy(strategy = c(\"curState\", \"split\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"strategy strategy inductive linearization matrix building curState Prefer parameterizing terms current state, followed first state observed term. split Split parameterization states term dividing number states term adding matrix term state.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"Nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLin_.html","id":null,"dir":"Reference","previous_headings":"","what":"Inductive linearization solver — rxIndLin_","title":"Inductive linearization solver — rxIndLin_","text":"Inductive linearization solver","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLin_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Inductive linearization solver — rxIndLin_","text":"cSub = Current subject number op rxode2 solving options tp Prior time point/time zero yp Prior state;  vector size = neq; Final state updated tf Final Time InfusionRate = Rates compartment;  vector size = neq Indicator compartment \"\" cache 0 = Cache doIndLin == 0, cache > 0 = nInf-1 rxode2 matrix exponential function IndF rxode2 Inductive Linearization function F","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLin_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Inductive linearization solver — rxIndLin_","text":"Returns status solving 1 = Successful solve -1 = Maximum number iterations reached inductive linearization","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":null,"dir":"Reference","previous_headings":"","what":"Initial Values and State values for a rxode2 object — rxInits","title":"Initial Values and State values for a rxode2 object — rxInits","text":"Returns initial values rxDll object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Initial Values and State values for a rxode2 object — rxInits","text":"","code":"rxInits(   obj,   vec = NULL,   req = NULL,   defaultValue = 0,   noerror = FALSE,   noini = FALSE,   rxLines = FALSE )  rxInit(   obj,   vec = NULL,   req = NULL,   defaultValue = 0,   noerror = FALSE,   noini = FALSE,   rxLines = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Initial Values and State values for a rxode2 object — rxInits","text":"obj rxDll, rxode2, named vector representing default initial arguments vec supplied, named vector model. req Required names, required order ODE solver defaultValue number NA representing default value parameters missing vec, required req. noerror boolean specifying error thrown missing parameter values default = NA","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Initial Values and State values for a rxode2 object — rxInits","text":"Initial values rxDll object","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Initial Values and State values for a rxode2 object — rxInits","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":null,"dir":"Reference","previous_headings":"","what":"Invert matrix using RcppArmadillo. — rxInv","title":"Invert matrix using RcppArmadillo. — rxInv","text":"Invert matrix using RcppArmadillo.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Invert matrix using RcppArmadillo. — rxInv","text":"","code":"rxInv(matrix)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Invert matrix using RcppArmadillo. — rxInv","text":"matrix matrix inverted.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Invert matrix using RcppArmadillo. — rxInv","text":"inverse pseudo inverse matrix.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":null,"dir":"Reference","previous_headings":"","what":"Check the type of an object using Rcpp — rxIs","title":"Check the type of an object using Rcpp — rxIs","text":"Check type object using Rcpp","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check the type of an object using Rcpp — rxIs","text":"","code":"rxIs(obj, cls)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check the type of an object using Rcpp — rxIs","text":"obj Object check cls Type class.  s3 classes lists/environments primitive classes checked. matrix types distinguished numeric.matrix, integer.matrix, logical.matrix, character.matrix well traditional matrix class. Additionally checks event.data.frame data.frame object time,  evid amt. (UPPER, lower Title cases accepted)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check the type of an object using Rcpp — rxIs","text":"boolean indicating object member class.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Check the type of an object using Rcpp — rxIs","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":null,"dir":"Reference","previous_headings":"","what":"Checks if the rxode2 object was built with the current build — rxIsCurrent","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"Checks rxode2 object built current build","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"","code":"rxIsCurrent(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"boolean indicating built current rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":null,"dir":"Reference","previous_headings":"","what":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"Determine DLL associated rxode2 object loaded","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"","code":"rxIsLoaded(obj)  rxDllLoaded(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"Boolean returning rxode2 library loaded.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":null,"dir":"Reference","previous_headings":"","what":"Left handed Variables — rxLhs","title":"Left handed Variables — rxLhs","text":"returns model calculated variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Left handed Variables — rxLhs","text":"","code":"rxLhs(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Left handed Variables — rxLhs","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Left handed Variables — rxLhs","text":"character vector listing calculated parameters","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Left handed Variables — rxLhs","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":null,"dir":"Reference","previous_headings":"","what":"Lock/unlocking of rxode2 dll file — rxLock","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"Lock/unlocking rxode2 dll file","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"","code":"rxLock(obj)  rxUnlock(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the md5 of an rxode2 object or file — rxMd5","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"md5 based model possibly extra c code supplied model.  addition md5 based syntax options, compiled rxode2 library md5, rxode2 version/repository.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"","code":"rxMd5(model, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . ... ignored arguments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"rxode2 object, return named list: file_md5 model's file's md5 parsed_md5  parsed model's file's md5. Otherwise return md5 based arguments provided","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":null,"dir":"Reference","previous_headings":"","what":"All model variables for a rxode2 object — rxModelVars","title":"All model variables for a rxode2 object — rxModelVars","text":"Return known model variables specified rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"All model variables for a rxode2 object — rxModelVars","text":"","code":"rxModelVars(obj)  rxModelVarsS3(obj)  # S3 method for rxUi rxModelVarsS3(obj)  # S3 method for default rxModelVarsS3(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"All model variables for a rxode2 object — rxModelVars","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"All model variables for a rxode2 object — rxModelVars","text":"list rxode2 model properties including: params  character vector names model parameters lhs character vector names model calculated parameters state character vector compartments rxode2 object trans named vector translated model properties including type jacobian specified, C function prefixes, well C functions names called compiled model. md5 named vector gives digest model (file_md5) parsed model (parsed_md5) model  named vector giving input model (model), normalized model (comments standard syntax parsing, normModel), interim code used generate final C file parseModel","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"All model variables for a rxode2 object — rxModelVars","text":"items calculated compilation; built-rxode2 compiled DLL. allow extension, s3 hook added function rxModelVarsS3.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"All model variables for a rxode2 object — rxModelVars","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the rxModels information — rxModels_","title":"Get the rxModels information — rxModels_","text":"Get rxModels  information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the rxModels information — rxModels_","text":"","code":"rxModels_(env = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the rxModels information — rxModels_","text":"env boolean returns environment models stored (TRUE), currently assigned rxode2 model variables (FALSE).","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the rxModels information — rxModels_","text":"internal rxModels information environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the normalized model — rxNorm","title":"Get the normalized model — rxNorm","text":"get syntax preferred model processing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the normalized model — rxNorm","text":"","code":"rxNorm(obj, condition = NULL, removeInis, removeJac, removeSens)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the normalized model — rxNorm","text":"obj rxode2 family objects condition Character string logical condition use subsetting normalized model.  missing, condition set via rxCondition, return whole code conditional settings intact.  condition set rxCondition, use condition. removeInis boolean indicating parameter initialization removed model removeJac boolean indicating Jacobians removed. removeSens boolean indicating sensitivities removed.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the normalized model — rxNorm","text":"Normalized Normal syntax (comments)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the normalized model — rxNorm","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":null,"dir":"Reference","previous_headings":"","what":"Optimize rxode2 for computer evaluation — rxOptExpr","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"optimizes rxode2 code computer evaluation calculating redundant expressions .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"","code":"rxOptExpr(x, msg = \"model\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"x rxode2 model can accessed rxNorm msg name type object rxode2 optimizing message optimizing.  example \"model\" produce following message optimizing model: finding duplicate expressions model...","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"Optimized rxode2 model text.  order type lhs state variables maintained evaluation sped . parameters names maintained, order may modified.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":null,"dir":"Reference","previous_headings":"","what":"Parameters specified by the model — rxParams","title":"Parameters specified by the model — rxParams","text":"returns model's parameters required solve ODE system, can used pipe parameters rxode2 solve","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Parameters specified by the model — rxParams","text":"","code":"rxParams(obj, ...)  # S3 method for rxode2 rxParams(   obj,   constants = TRUE,   ...,   params = NULL,   inits = NULL,   iCov = NULL,   keep = NULL,   thetaMat = NULL,   omega = NULL,   dfSub = NULL,   sigma = NULL,   dfObs = NULL,   nSub = NULL,   nStud = NULL )  # S3 method for rxSolve rxParams(   obj,   constants = TRUE,   ...,   params = NULL,   inits = NULL,   iCov = NULL,   keep = NULL,   thetaMat = NULL,   omega = NULL,   dfSub = NULL,   sigma = NULL,   dfObs = NULL,   nSub = NULL,   nStud = NULL )  # S3 method for rxEt rxParams(   obj,   ...,   params = NULL,   inits = NULL,   iCov = NULL,   keep = NULL,   thetaMat = NULL,   omega = NULL,   dfSub = NULL,   sigma = NULL,   dfObs = NULL,   nSub = NULL,   nStud = NULL )  rxParam(obj, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Parameters specified by the model — rxParams","text":"obj rxode2 family objects ... arguments including scaling factors compartment.  includes S# = numeric scale compartment # dividing compartment amount scale factor, like NONMEM. constants boolean indicting constants included list parameters. Currently rxode2 parses constants variables case wish change without recompiling rxode2 model. params numeric named vector values every parameter ODE system; names must correspond parameter identifiers used ODE specification; inits vector initial values state variables (e.g., amounts compartment), order vector must state variables (e.g., PK/PD compartments); iCov data frame individual non-time varying covariates combine events dataset merge. keep Columns keep either input dataset iCov dataset.  iCov dataset, column kept per line.  input dataset, records added data LOCF (Last Observation Carried forward) imputation performed. thetaMat Named theta matrix. omega Estimate Covariance matrix. omega list, assume block matrix convert full matrix simulations.  omega NA using rxode2 ui model, subject variability described omega matrix set zero. dfSub Degrees freedom sample subject variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. sigma Named sigma covariance Cholesky decomposition covariance matrix.  names columns indicate parameters simulated.  simulated every observation solved system. sigma NA using rxode2 ui model, unexplained variability described sigma matrix set zero. dfObs Degrees freedom sample unexplained variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. nSub Number subject variabilities (ETAs) simulated every realization parameters. nStud Number virtual studies characterize uncertainty estimated parameters.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Parameters specified by the model — rxParams","text":"extracting parameters rxode2 model, character vector listing parameters model.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Parameters specified by the model — rxParams","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":null,"dir":"Reference","previous_headings":"","what":"Prepare Error function for inclusion in rxode2 — rxParseErr","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"Prepare Error function inclusion rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"","code":"rxParseErr(   x,   baseTheta,   ret = \"rx_r_\",   init = NULL,   addProp = c(\"combined2\", \"combined1\") )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"x error function baseTheta Base theta start numbering add(.) prop(.) . ret Internal return type.  changed user... init Initialization vector","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"rxode2 transformed text","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":null,"dir":"Reference","previous_headings":"","what":"Parse PK function for inclusion in rxode2 — rxParsePk","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"Parse PK function inclusion rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"","code":"rxParsePk(x, init = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"x PK function init Initialization vector","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"rxode2 transformed text.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":null,"dir":"Reference","previous_headings":"","what":"Prepare Pred function for inclusion in rxode2 — rxParsePred","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"Prepare Pred function inclusion rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"","code":"rxParsePred(x, init = NULL, err = NULL, addProp = c(\"combined2\", \"combined1\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"x pred function init Initialization vector","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"rxode2 transformed text.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates a package from compiled rxode2 models — rxPkg","title":"Creates a package from compiled rxode2 models — rxPkg","text":"Creates package compiled rxode2 models","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates a package from compiled rxode2 models — rxPkg","text":"","code":"rxPkg(   ...,   package,   wd = getwd(),   action = c(\"install\", \"build\", \"binary\", \"create\"),   license = c(\"gpl3\", \"lgpl\", \"mit\", \"agpl3\"),   name = \"Firstname Lastname\",   fields = list() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates a package from compiled rxode2 models — rxPkg","text":"... Models build package package String package name create wd character string working directory create subdirectory according modName. specified, subdirectory named “modName.d” created populated C file, dynamic loading library, plus various working files. missing, files created (removed) temporary directory, rxode2 DLL model created current directory named rx_????_platform, example rx_129f8f97fb94a87ca49ca8dafe691e1e_i386.dll action Type action take package created license type license package. name Full name author fields named list fields add DESCRIPTION, potentially overriding default values. See use_description() can set personalized defaults using package options.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates a package from compiled rxode2 models — rxPkg","text":"function returns nothing used side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Creates a package from compiled rxode2 models — rxPkg","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate a from a Poisson process — rxPp","title":"Simulate a from a Poisson process — rxPp","text":"Simulate Poisson process","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate a from a Poisson process — rxPp","text":"","code":"rxPp(   n,   lambda,   gamma = 1,   prob = NULL,   t0 = 0,   tmax = Inf,   randomOrder = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate a from a Poisson process — rxPp","text":"n Number time points simulate Poisson process lambda Rate Poisson process gamma Asymmetry rate Poisson process.  gamma=1.0, simulates homogenous Poisson process.  gamma<1.0, Poisson process events early, gamma > 1.0, Poisson process events late process. gamma non-zero, tmax infinite indicate end Poisson process simulated.  pharamcometric cases, end study. Internally uses rate : l(t) = lambdagamma(t/tmax)^(gamma-1) prob specified, probability function one argument, time, gives probability Poisson time t accepted rejection time. t0 starting time Poisson process tmax maximum time Poisson process randomOrder TRUE randomize order Poisson events.  default (FALSE) returns Poisson process order events occurred.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate a from a Poisson process — rxPp","text":"returns vector Poisson process times; dropout >= tmax, rest times = tmax indicate dropout equal tmax.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate a from a Poisson process — rxPp","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate a from a Poisson process — rxPp","text":"","code":"## Sample homogenous Poisson process of rate 1/10 rxPp(10, 1 / 10) #>  [1] 14.16159 22.43681 41.13405 48.71743 51.37508 54.60656 75.32184 84.62128 #>  [9] 85.80475 96.63193  ## Sample inhomogenous Poisson rate of 1/10  rxPp(10, 1 / 10, gamma = 2, tmax = 100) #>  [1] 37.13008 37.33395 38.19765 41.94718 50.45308 51.16211 65.45638 75.56937 #>  [9] 76.00046 83.59277  ## Typically the Poisson process times are in a sequential order, ## using randomOrder gives the Poisson process in random order  rxPp(10, 1 / 10, gamma = 2, tmax = 10, randomOrder = TRUE) #>  [1] 10.000000 10.000000 10.000000 10.000000  6.028535 10.000000 10.000000 #>  [8] 10.000000 10.000000 10.000000  ## This uses an arbitrary function to sample a non-homogenous Poisson process  rxPp(10, 1 / 10, prob = function(x) {   1/(1+abs(x)) }) #>  [1]  14.74011  38.01194  53.91062 235.31848 381.74524 397.80743 504.64689 #>  [8] 721.10404 756.19022 889.86201"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":null,"dir":"Reference","previous_headings":"","what":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"determined internal preferred condition name list .errIdenticalDists","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"","code":"rxPreferredDistributionName(dist)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"dist input distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"Preferred distribution term","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"","code":"rxPreferredDistributionName(\"dt\") #> [1] \"t\"  rxPreferredDistributionName(\"add\") #> [1] \"add\"  # can be vectorized  rxPreferredDistributionName(c(\"add\",\"dt\")) #>   add    dt  #> \"add\"   \"t\""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":null,"dir":"Reference","previous_headings":"","what":"rxode2 progress bar functions — rxProgress","title":"rxode2 progress bar functions — rxProgress","text":"rxProgress sets progress bar","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxode2 progress bar functions — rxProgress","text":"","code":"rxProgress(num, core = 0L)  rxTick()  rxProgressStop(clear = TRUE)  rxProgressAbort(error = \"Aborted calculation\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxode2 progress bar functions — rxProgress","text":"num Tot number operations track core Number cores show.  1, show number cores clear Boolean telling clear progress bar completion (displayed).  default TRUE error rxProgressAbort error displayed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxode2 progress bar functions — rxProgress","text":"return NULL invisibly.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"rxode2 progress bar functions — rxProgress","text":"rxTick progress bar tick rxProgressStop stop progress bar rxProgressAbort shows abort rxProgressStop called.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxode2 progress bar functions — rxProgress","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"rxode2 progress bar functions — rxProgress","text":"","code":"f <- function() {   on.exit({     rxProgressAbort()   })   rxProgress(100)   for (i in 1:100) {     rxTick()     Sys.sleep(1 / 100)   }   rxProgressStop() } # \\donttest{ f() # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":null,"dir":"Reference","previous_headings":"","what":"Prune branches (ie if/else) from rxode2 — rxPrune","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"prunes branches (ie /else) rxode2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"","code":"rxPrune(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"x rxode2 model can accessed rxNorm","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"Pruned rxode2 model text.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":null,"dir":"Reference","previous_headings":"","what":"Reload rxode2 DLL — rxReload","title":"Reload rxode2 DLL — rxReload","text":"Can useful debugging","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reload rxode2 DLL — rxReload","text":"","code":"rxReload()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Reload rxode2 DLL — rxReload","text":"boolean object reloaded","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Reload rxode2 DLL — rxReload","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":null,"dir":"Reference","previous_headings":"","what":"rxRemoveControl options for UI object — rxRemoveControl","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"rxRemoveControl options UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"","code":"rxRemoveControl(ui)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"ui rxode2 ui object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":null,"dir":"Reference","previous_headings":"","what":"Rename items inside of a rxode2 ui model — rxRename","title":"Rename items inside of a rxode2 ui model — rxRename","text":"rxRename() changes names individual variables, lhs, ode states using new_name = old_name syntax","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Rename items inside of a rxode2 ui model — rxRename","text":"","code":"rxRename(.data, ..., envir = parent.frame())  .rxRename(.data, ..., envir = parent.frame())  rename.rxUi(.data, ...)  rename.function(.data, ...)  # S3 method for rxUi rxRename(.data, ...)  # S3 method for `function` rxRename(.data, ...)  # S3 method for default rxRename(.data, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Rename items inside of a rxode2 ui model — rxRename","text":".data rxode2 ui function, named data consistent dplyr::rename() ... rename items envir Environment evaluation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Rename items inside of a rxode2 ui model — rxRename","text":"New model items renamed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Rename items inside of a rxode2 ui model — rxRename","text":"similar dplyr's rename() function.  dplyr loaded, s3 methods work ui objects. Note .rxRename() internal function called renaming likely need call unless writing extension function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Rename items inside of a rxode2 ui model — rxRename","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Rename items inside of a rxode2 ui model — rxRename","text":"","code":"ocmt <- function() {   ini({     tka <- exp(0.45) # Ka     tcl <- exp(1) # Cl     ## This works with interactive models     ## You may also label the preceding line with label(\"label text\")     tv <- exp(3.45) # log V     ## the label(\"Label name\") works with all models     add.sd <- 0.7   })   model({     ka <- tka     cl <- tcl     v <- tv     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  ocmt %>% rxRename(cpParent=cp) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #>  1.568312  2.718282 31.500392  0.700000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cpParent = center/v #>         cpParent ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":null,"dir":"Reference","previous_headings":"","what":"Rep R0 for foce — rxRepR0_","title":"Rep R0 for foce — rxRepR0_","text":"Rep R0 foce","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Rep R0 for foce — rxRepR0_","text":"","code":"rxRepR0_(neta)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Rep R0 for foce — rxRepR0_","text":"neta ETA substitute","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Rep R0 for foce — rxRepR0_","text":"Returns string R code substitute rx_r expression symengine environment .s","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReservedKeywords.html","id":null,"dir":"Reference","previous_headings":"","what":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","title":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","text":"list description rxode2 supported reserved keywords","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReservedKeywords.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","text":"","code":"rxReservedKeywords"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReservedKeywords.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","text":"data frame 3 columns 31 rows Reserved Name Reserved Keyword Name Meaning Reserved Keyword Meaning Alias Keyword Alias","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxResidualError.html","id":null,"dir":"Reference","previous_headings":"","what":"A description of Rode2 supported residual errors — rxResidualError","title":"A description of Rode2 supported residual errors — rxResidualError","text":"description Rode2 supported residual errors","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxResidualError.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A description of Rode2 supported residual errors — rxResidualError","text":"","code":"rxResidualError"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxResidualError.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"A description of Rode2 supported residual errors — rxResidualError","text":"data frame 6 columns 181 rows Error model description type residual error Functional Form additive proportional functional form used Transformation type transformation done DV prediction code Example code residual error type addProp type add+prop residual error default equivalent lhs left handed side specification represents, either response variable, compartment specification","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":null,"dir":"Reference","previous_headings":"","what":"Load a model into a symengine environment — rxS","title":"Load a model into a symengine environment — rxS","text":"Load model symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load a model into a symengine environment — rxS","text":"","code":"rxS(x, doConst = TRUE, promoteLinSens = FALSE, envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load a model into a symengine environment — rxS","text":"x rxode2 object doConst Load constants environment well. promoteLinSens Promote solved linear compartment systems sensitivity-based solutions. envir default NULL; Environment put symengine variables .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load a model into a symengine environment — rxS","text":"rxode2/symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Load a model into a symengine environment — rxS","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":null,"dir":"Reference","previous_headings":"","what":"rxSetControl options for UI object — rxSetControl","title":"rxSetControl options for UI object — rxSetControl","text":"rxSetControl options UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxSetControl options for UI object — rxSetControl","text":"","code":"rxSetControl(ui, control)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxSetControl options for UI object — rxSetControl","text":"ui rxode2 ui object control Default value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxSetControl options for UI object — rxSetControl","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxSetControl options for UI object — rxSetControl","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign covariates for piping — rxSetCovariateNamesForPiping","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"Assign covariates piping","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"","code":"rxSetCovariateNamesForPiping(covariates = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"covariates NULL (covariates), list covariates. nlmixr uses function set covariates pipe nlmixr fit.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"","code":"# First set the name of known covariates # Note this is case sensitive  rxSetCovariateNamesForPiping(c(\"WT\",\"HT\", \"TC\"))  one.compartment <- function() {  ini({    tka <- 0.45 ; label(\"Log Ka\")    tcl <- 1 ; label(\"Log Cl\")    tv <- 3.45 ; label(\"Log V\")    eta.ka ~ 0.6    eta.cl ~ 0.3    eta.v ~ 0.1    add.err <- 0.7  })  model({    ka <- exp(tka + eta.ka)    cl <- exp(tcl + eta.cl)    v <- exp(tv + eta.v)    d / dt(depot) <- -ka * depot    d/dt(depot) <- -ka * depot    d / dt(center) <- ka * depot - cl / v * center    cp <- center / v    cp ~ add(add.err)  }) }  # now TC is detected as a covariate instead of a population parameter  one.compartment %>%   model({ka <- exp(tka + eta.ka + TC * cov_C)}) #>   #>   #> ℹ add covariate `TC` (known covariate) #> ℹ add population parameter `cov_C` and set estimate to 1 #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>     tka     tcl      tv add.err   cov_C  #>    0.45    1.00    3.45    0.70    1.00  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tka eta.ka    id   TC*cov_C #> 2   tcl eta.cl    id            #> 3    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Log Ka\") #>         tcl <- 1 #>         label(\"Log Cl\") #>         tv <- 3.45 #>         label(\"Log V\") #>         add.err <- c(0, 0.7) #>         cov_C <- 1 #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka + TC * cov_C) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.err) #>     }) #> }  # You can turn it off by simply adding it back  rxSetCovariateNamesForPiping()  one.compartment %>%   model({ka <- exp(tka + eta.ka + TC * cov_C)}) #>   #>   #> ℹ add population parameter `TC` and set estimate to 1 #> ℹ add population parameter `cov_C` and set estimate to 1 #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>     tka     tcl      tv add.err      TC   cov_C  #>    0.45    1.00    3.45    0.70    1.00    1.00  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Log Ka\") #>         tcl <- 1 #>         label(\"Log Cl\") #>         tv <- 3.45 #>         label(\"Log V\") #>         add.err <- c(0, 0.7) #>         TC <- 1 #>         cov_C <- 1 #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka + TC * cov_C) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.err) #>     }) #> }  # The covariates you set with `rxSetCovariateNamesForPiping()` # are turned off every time you solve (or fit in nlmixr)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":null,"dir":"Reference","previous_headings":"","what":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"Set variables model piping automatic covarite selection","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"","code":"rxSetPipingAuto(   thetamodelVars = rex::rex(or(\"tv\", \"t\", \"pop\", \"POP\", \"Pop\", \"TV\", \"T\", \"cov\", \"err\",     \"eff\")),   covariateExceptions = rex::rex(start, or(\"wt\", \"sex\", \"crcl\", \"kout\"), end),   etaParts = c(\"eta\", \"ETA\", \"Eta\", \"ppv\", \"PPV\", \"Ppv\", \"iiv\", \"Iiv\", \"bsv\", \"Bsv\",     \"BSV\", \"bpv\", \"Bpv\", \"BPV\", \"psv\", \"PSV\", \"Psv\") )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"thetamodelVars prefixes theta model variables regular expression covariateExceptions regular expression covariates always covariates etaParts list eta prefixes/post-fixes identify variable subject variability","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"called startup set defaults, though can change wish piping can work differently individual setup","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":null,"dir":"Reference","previous_headings":"","what":"Defunct setting of product — rxSetProd","title":"Defunct setting of product — rxSetProd","text":"Defunct setting product","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Defunct setting of product — rxSetProd","text":"","code":"rxSetProd(type = c(\"long double\", \"double\", \"logify\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Defunct setting of product — rxSetProd","text":"type used type product","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Defunct setting of product — rxSetProd","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":null,"dir":"Reference","previous_headings":"","what":"Set timing for progress bar — rxSetProgressBar","title":"Set timing for progress bar — rxSetProgressBar","text":"Set timing progress bar","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set timing for progress bar — rxSetProgressBar","text":"","code":"rxSetProgressBar(seconds = 1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set timing for progress bar — rxSetProgressBar","text":"seconds sets number seconds need elapse drawing next segment progress bar.  zero turns progress bar.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set timing for progress bar — rxSetProgressBar","text":"nothing, used side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set timing for progress bar — rxSetProgressBar","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":null,"dir":"Reference","previous_headings":"","what":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"Silence rxode2's C/C++ messages","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"","code":"rxSetSilentErr(silent)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"silent can 0L \"noisy\"  1L \"silent\"","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"TRUE; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":null,"dir":"Reference","previous_headings":"","what":"Defunct setting of sum — rxSetSum","title":"Defunct setting of sum — rxSetSum","text":"Defunct setting sum","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Defunct setting of sum — rxSetSum","text":"","code":"rxSetSum(type = c(\"pairwise\", \"fsum\", \"kahan\", \"neumaier\", \"c\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Defunct setting of sum — rxSetSum","text":"type used type product","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Defunct setting of sum — rxSetSum","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Setup the initial conditions. — rxSetupIni","title":"Setup the initial conditions. — rxSetupIni","text":"Setup initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Setup the initial conditions. — rxSetupIni","text":"","code":"rxSetupIni(obj, inits = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Setup the initial conditions. — rxSetupIni","text":"obj rxode2 object inits numeric vector initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Setup the initial conditions. — rxSetupIni","text":"initial conditions setup","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Setup the initial conditions. — rxSetupIni","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":null,"dir":"Reference","previous_headings":"","what":"Setup the initial conditions. — rxSetupScale","title":"Setup the initial conditions. — rxSetupScale","text":"Setup initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Setup the initial conditions. — rxSetupScale","text":"","code":"rxSetupScale(obj, scale = NULL, extraArgs = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Setup the initial conditions. — rxSetupScale","text":"obj rxode2 object scale numeric vector scales extraArgs list extra args parse initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Setup the initial conditions. — rxSetupScale","text":"setup scale changing compartment values","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Setup the initial conditions. — rxSetupScale","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":null,"dir":"Reference","previous_headings":"","what":"Use Shiny to help develop an rxode2 model — rxShiny","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"Use Shiny help develop rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"","code":"rxShiny(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   data = data.frame() )  # S3 method for rxSolve rxShiny(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   data = data.frame() )  # S3 method for default rxShiny(   object = NULL,   params = NULL,   events = NULL,   inits = NULL,   ...,   data = data.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"object rxode2 family objects. supplied 2-compartment indirect effect model used.  supplied, use model associated rxode2 object model exploration. params Initial parameters model events Event information (currently ignored) inits Initial estimates model ... arguments passed rxShiny.  Currently anything. data data like plot.  data time variable well compartment calculated variable matches rxode2 model, data added plot specific compartment calculated variable.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"Nothing; Starts shiny server","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"Zufar Mulyukov Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"Simulate Parameters Theta/Omega specification","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"","code":"rxSimThetaOmega(   params = NULL,   omega = NULL,   omegaDf = NULL,   omegaLower = as.numeric(c(R_NegInf)),   omegaUpper = as.numeric(c(R_PosInf)),   omegaIsChol = FALSE,   omegaSeparation = \"auto\",   omegaXform = 1L,   nSub = 1L,   thetaMat = NULL,   thetaLower = as.numeric(c(R_NegInf)),   thetaUpper = as.numeric(c(R_PosInf)),   thetaDf = NULL,   thetaIsChol = FALSE,   nStud = 1L,   sigma = NULL,   sigmaLower = as.numeric(c(R_NegInf)),   sigmaUpper = as.numeric(c(R_PosInf)),   sigmaDf = NULL,   sigmaIsChol = FALSE,   sigmaSeparation = \"auto\",   sigmaXform = 1L,   nCoresRV = 1L,   nObs = 1L,   dfSub = 0,   dfObs = 0,   simSubjects = TRUE,   simVariability = as.logical(c(NA_LOGICAL)) )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"params Named Vector rxode2 model parameters omega Estimate Covariance matrix. omega list, assume block matrix convert full matrix simulations.  omega NA using rxode2 ui model, subject variability described omega matrix set zero. omegaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. omegaLower Lower bounds simulated ETAs (default -Inf) omegaUpper Upper bounds simulated ETAs (default Inf) omegaIsChol Indicates omega supplied Cholesky decomposed matrix instead traditional symmetric matrix. omegaSeparation Omega separation strategy Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. omegaXform taking omega values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. nSub Number subject variabilities (ETAs) simulated every realization parameters. thetaMat Named theta matrix. thetaLower Lower bounds simulated population parameter variability (default -Inf) thetaUpper Upper bounds simulated population unexplained variability (default Inf) thetaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. thetaIsChol Indicates theta supplied Cholesky decomposed matrix instead traditional symmetric matrix. nStud Number virtual studies characterize uncertainty estimated parameters. sigma Named sigma covariance Cholesky decomposition covariance matrix.  names columns indicate parameters simulated.  simulated every observation solved system. sigma NA using rxode2 ui model, unexplained variability described sigma matrix set zero. sigmaLower Lower bounds simulated unexplained variability (default -Inf) sigmaUpper Upper bounds simulated unexplained variability (default Inf) sigmaDf Degrees freedom sigma t-distribution.  default equivalent Inf, normal distribution. sigmaIsChol Boolean indicating sigma Cholesky decomposition instead symmetric covariance sigmaSeparation separation strategy sigma; Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. sigmaXform taking sigma values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. nCoresRV Number cores used simulation sigma variables.  default 1. reproduce results need run platform number cores. reason set one, regardless number cores used threaded ODE solving. nObs Number observations simulate (sigma matrix) dfSub Degrees freedom sample subject variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. dfObs Degrees freedom sample unexplained variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. simSubjects boolean indicated rxode2 simulate subjects studies (TRUE, default) studies (FALSE) simVariability determines variability simulated. NA (default) determined solver.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"data frame simulated subjects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Options, Solving & Simulation of an ODE/solved system — rxSolve","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"uses rxode2 family objects, file, model specification solve ODE system.  many options solved rxode2 model, first required object, events -times optional params inits.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"","code":"rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   scale = NULL,   method = c(\"liblsoda\", \"lsoda\", \"dop853\", \"indLin\"),   sigdig = NULL,   atol = 1e-08,   rtol = 1e-06,   maxsteps = 70000L,   hmin = 0,   hmax = NA_real_,   hmaxSd = 0,   hini = 0,   maxordn = 12L,   maxords = 5L,   ...,   cores,   covsInterpolation = c(\"locf\", \"linear\", \"nocb\", \"midpoint\"),   addCov = TRUE,   sigma = NULL,   sigmaDf = NULL,   sigmaLower = -Inf,   sigmaUpper = Inf,   nCoresRV = 1L,   sigmaIsChol = FALSE,   sigmaSeparation = c(\"auto\", \"lkj\", \"separation\"),   sigmaXform = c(\"identity\", \"variance\", \"log\", \"nlmixrSqrt\", \"nlmixrLog\",     \"nlmixrIdentity\"),   nDisplayProgress = 10000L,   amountUnits = NA_character_,   timeUnits = \"hours\",   theta = NULL,   thetaLower = -Inf,   thetaUpper = Inf,   eta = NULL,   addDosing = FALSE,   stateTrim = Inf,   updateObject = FALSE,   omega = NULL,   omegaDf = NULL,   omegaIsChol = FALSE,   omegaSeparation = c(\"auto\", \"lkj\", \"separation\"),   omegaXform = c(\"variance\", \"identity\", \"log\", \"nlmixrSqrt\", \"nlmixrLog\",     \"nlmixrIdentity\"),   omegaLower = -Inf,   omegaUpper = Inf,   nSub = 1L,   thetaMat = NULL,   thetaDf = NULL,   thetaIsChol = FALSE,   nStud = 1L,   dfSub = 0,   dfObs = 0,   returnType = c(\"rxSolve\", \"matrix\", \"data.frame\", \"data.frame.TBS\", \"data.table\",     \"tbl\", \"tibble\"),   seed = NULL,   nsim = NULL,   minSS = 10L,   maxSS = 1000L,   infSSstep = 12,   strictSS = TRUE,   istateReset = TRUE,   subsetNonmem = TRUE,   maxAtolRtolFactor = 0.1,   from = NULL,   to = NULL,   by = NULL,   length.out = NULL,   iCov = NULL,   keep = NULL,   indLinPhiTol = 1e-07,   indLinPhiM = 0L,   indLinMatExpType = c(\"expokit\", \"Al-Mohy\", \"arma\"),   indLinMatExpOrder = 6L,   drop = NULL,   idFactor = TRUE,   mxhnil = 0,   hmxi = 0,   warnIdSort = TRUE,   warnDrop = TRUE,   ssAtol = 1e-08,   ssRtol = 1e-06,   safeZero = TRUE,   sumType = c(\"pairwise\", \"fsum\", \"kahan\", \"neumaier\", \"c\"),   prodType = c(\"long double\", \"double\", \"logify\"),   sensType = c(\"advan\", \"autodiff\", \"forward\", \"central\"),   linDiff = c(tlag = 1.5e-05, f = 1.5e-05, rate = 1.5e-05, dur = 1.5e-05, tlag2 =     1.5e-05, f2 = 1.5e-05, rate2 = 1.5e-05, dur2 = 1.5e-05),   linDiffCentral = c(tlag = TRUE, f = TRUE, rate = TRUE, dur = TRUE, tlag2 = TRUE, f2 =     TRUE, rate2 = TRUE, dur2 = TRUE),   resample = NULL,   resampleID = TRUE,   maxwhile = 1e+05,   atolSens = 1e-08,   rtolSens = 1e-06,   ssAtolSens = 1e-08,   ssRtolSens = 1e-06,   simVariability = NA,   nLlikAlloc = NULL,   useStdPow = FALSE,   naTimeHandle = c(\"ignore\", \"warn\", \"error\"),   addlKeepsCov = FALSE,   addlDropSs = TRUE,   ssAtDoseTime = TRUE,   ss2cancelAllPending = FALSE,   envir = parent.frame() )  # S3 method for `function` rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for rxUi rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for rxode2tos rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for nlmixr2FitData rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for nlmixr2FitCore rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for default rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for rxSolve update(object, ...)  # S3 method for rxode2 predict(object, ...)  # S3 method for `function` predict(object, ...)  # S3 method for rxUi predict(object, ...)  # S3 method for rxSolve predict(object, ...)  # S3 method for rxEt predict(object, ...)  # S3 method for rxParams predict(object, ...)  # S3 method for rxode2 simulate(object, nsim = 1L, seed = NULL, ...)  # S3 method for rxSolve simulate(object, nsim = 1L, seed = NULL, ...)  # S3 method for rxParams simulate(object, nsim = 1L, seed = NULL, ...)  # S3 method for rxSolve solve(a, b, ...)  # S3 method for rxUi solve(a, b, ...)  # S3 method for `function` solve(a, b, ...)  # S3 method for rxode2 solve(a, b, ...)  # S3 method for rxParams solve(a, b, ...)  # S3 method for rxEt solve(a, b, ...)  rxControl(   ...,   params = NULL,   events = NULL,   inits = NULL,   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"object either rxode2 family objects, file-name rxode2 model specification, string rxode2 model specification. params numeric named vector values every parameter ODE system; names must correspond parameter identifiers used ODE specification; events eventTable object describing input (e.g., doses) dynamic system observation sampling time points (see eventTable()); inits vector initial values state variables (e.g., amounts compartment), order vector must state variables (e.g., PK/PD compartments); scale numeric named vector scaling ode parameters system.  names must correspond parameter identifiers ODE specification. ODE variables divided scaling factor.  example scale=c(center=2) divide center ODE variable 2. method method solving ODEs.  Currently supports: \"liblsoda\" thread safe lsoda.  supports parallel thread-based solving, ignores user Jacobian specification. \"lsoda\" -- LSODA solver.  support parallel thread-based solving, allows user Jacobian specification. \"dop853\" -- DOP853 solver.  support parallel thread-based solving user Jacobian specification \"indLin\" -- Solving inductive linearization.  rxode2 dll must setup specially use solving routine. sigdig Specifies \"significant digits\" ode solving requests.  specified controls relative absolute tolerances ODE solvers.  default tolerance 0.5*10^(-sigdig-2) regular ODEs. sensitivity equations default 0.5*10\\^(-sigdig-1.5) (sensitivity changes applicable liblsoda).  also controls atol/rtol steady state solutions. ssAtol/ssRtol 0.5*10\\^(-sigdig) sensitivities 0.5*10\\^(-sigdig+0.625).  default unspecified (NULL) uses standard atol/rtol. atol numeric absolute tolerance (1e-8 default) used ODE solver determine good solution achieved;  also used solved linear model check prior doses add anything solution. rtol numeric relative tolerance (1e-6 default) used ODE solver determine good solution achieved. also used solved linear model check prior doses add anything solution. maxsteps maximum number (internally defined) steps allowed one call solver. (5000 default) hmin minimum absolute step size allowed. default value 0. hmax maximum absolute step size allowed.  hmax=NA (default), uses average difference + hmaxSd*sd times sampling events. hmaxSd user specified parameter defaults zero.  hmax=NULL rxode2 uses maximum difference times sampling events.  value 0 equivalent infinite maximum absolute step size. hmaxSd number standard deviations time difference add hmax. default 0 hini step size attempted first step. default value determined solver (hini = 0) maxordn maximum order allowed nonstiff (Adams) method.  default 12.  can 1 12. maxords maximum order allowed stiff (BDF) method.  default value 5.  can 1 5. ... arguments including scaling factors compartment.  includes S# = numeric scale compartment # dividing compartment amount scale factor, like NONMEM. cores Number cores used parallel ODE solving.  equivalent calling setRxThreads() covsInterpolation specifies interpolation method time-varying covariates. solving ODEs often samples times outside sampling time specified events. happens, time varying covariates interpolated.  Currently can : \"linear\" interpolation, interpolates covariate solving line observed covariates extrapolating new covariate value. \"constant\" -- Last observation carried forward (default). \"NOCB\" -- Next Observation Carried Backward.  method NONMEM uses. \"midpoint\" Last observation carried forward midpoint; Next observation carried backward midpoint. addCov boolean indicating covariates added output matrix data frame. default disabled. sigma Named sigma covariance Cholesky decomposition covariance matrix.  names columns indicate parameters simulated.  simulated every observation solved system. sigma NA using rxode2 ui model, unexplained variability described sigma matrix set zero. sigmaDf Degrees freedom sigma t-distribution.  default equivalent Inf, normal distribution. sigmaLower Lower bounds simulated unexplained variability (default -Inf) sigmaUpper Upper bounds simulated unexplained variability (default Inf) nCoresRV Number cores used simulation sigma variables.  default 1. reproduce results need run platform number cores. reason set one, regardless number cores used threaded ODE solving. sigmaIsChol Boolean indicating sigma Cholesky decomposition instead symmetric covariance sigmaSeparation separation strategy sigma; Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. sigmaXform taking sigma values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. nDisplayProgress integer indicating minimum number c-based solves progress bar shown.  default 10,000. amountUnits supplies dose units data frame supplied instead event table.  importing data rxode2 event table. timeUnits supplies time units data frame supplied instead event table.  importing data rxode2 event table. theta vector parameters named THETA\\[#\\] added parameters thetaLower Lower bounds simulated population parameter variability (default -Inf) thetaUpper Upper bounds simulated population unexplained variability (default Inf) eta vector parameters named ETA\\[#\\] added parameters addDosing Boolean indicating solve add rxode2 EVID related columns.  also include dosing information estimates doses.  default, rxode2 includes estimates observations. (default FALSE). addDosing NULL, include EVID=0 solve exclude model-times EVID=2. addDosing NA classic rxode2 EVID events returned. addDosing TRUE add event information NONMEM-style format; subsetNonmem=FALSE rxode2 also include extra event types (EVID) ending infusion modeled times: EVID=-1 modeled rate infusions turned (matches rate=-1) EVID=-2 modeled duration infusions turned (matches rate=-2) EVID=-10 specified rate infusions turned (matches rate>0) EVID=-20 specified dur infusions turned (matches dur>0) EVID=101,102,103,... Modeled time 101 first model time, 102 second etc. stateTrim amounts/concentrations one states value, trim value. default Inf.  Also trims -stateTrim large negative amounts/concentrations.  want trim range say c(0, 2000000) may specify 2 values lower upper range make sure state values reasonable range. updateObject internally used flag update rxode2 solved object (supplying rxode2 solved object) well returning new object.  probably modify FALSE default unless willing unexpected results. omega Estimate Covariance matrix. omega list, assume block matrix convert full matrix simulations.  omega NA using rxode2 ui model, subject variability described omega matrix set zero. omegaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. omegaIsChol Indicates omega supplied Cholesky decomposed matrix instead traditional symmetric matrix. omegaSeparation Omega separation strategy Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. omegaXform taking omega values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. omegaLower Lower bounds simulated ETAs (default -Inf) omegaUpper Upper bounds simulated ETAs (default Inf) nSub Number subject variabilities (ETAs) simulated every realization parameters. thetaMat Named theta matrix. thetaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. thetaIsChol Indicates theta supplied Cholesky decomposed matrix instead traditional symmetric matrix. nStud Number virtual studies characterize uncertainty estimated parameters. dfSub Degrees freedom sample subject variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. dfObs Degrees freedom sample unexplained variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. returnType tells type object returned.  currently supported types : \"rxSolve\" (default) return reactive data frame can change easily change different pieces solve update data frame.  currently standard solving method rxode2,  used rxSolve(object, ...), solve(object,...), \"data.frame\" -- returns plain, non-reactive data frame; Currently slightly faster returnType=\"matrix\" \"matrix\" -- returns plain matrix column names attached solved object.  used object$run well object$solve \"data.table\" -- returns data.table; data.table created reference (ie setDt()), fast. \"tbl\" \"tibble\" returns tibble format. seed object specifying random number generator initialized nsim represents number simulations.  rxode2, supply single subject event tables (created [eventTable()]) minSS Minimum number iterations steady-state dose maxSS Maximum number iterations steady-state dose infSSstep Step size determining constant infusion reached steady state.  default large value, 12. strictSS Boolean indicating strict steady-state required. strict steady-state (TRUE) required least minSS doses administered total number steady states doses continue maxSS reached, atol rtol every compartment reached.  However, ODE solving problems occur minSS reached whole subject considered invalid solve. strictSS FALSE long minSS reached last good solve ODE solving problems occur considered steady state, even though either atol, rtol maxSS achieved. istateReset TRUE, reset ISTATE variable 1 lsoda liblsoda doses, like deSolve; FALSE, reset ISTATE variable doses. subsetNonmem subset NONMEM compatible EVIDs .  default TRUE. maxAtolRtolFactor maximum atol/rtol FOCEi routines may adjust .  default 0.1 observations event table, start observations value. default zero. observations event table, end observations value. default 24 + maximum dose time. observations event table, amount increment observations . length.number observations create observations event table. default 200. iCov data frame individual non-time varying covariates combine events dataset merge. keep Columns keep either input dataset iCov dataset.  iCov dataset, column kept per line.  input dataset, records added data LOCF (Last Observation Carried forward) imputation performed. indLinPhiTol requested accuracy tolerance exponential matrix. indLinPhiM maximum size Krylov basis indLinMatExpType matrix exponential type use rxode2.  Currently following supported: Al-Mohy Uses exponential matrix method Al-Mohy Higham (2009) arma Use exponential matrix RcppArmadillo expokit Use exponential matrix Roger B. Sidje (1998) indLinMatExpOrder integer, order approximation used, Al-Mohy expokit values. best value depends machine precision (slightly matrix). use 6 default. drop Columns drop output idFactor boolean indicates original ID values maintained. changes default sequentially ordered ID factor original ID values original dataset.  default enabled. mxhnil maximum number messages printed (per problem) warning T + H = T step (H = step size).  must positive result non-default value.  default value 0 (infinite). hmxi inverse maximum absolute value H used. hmxi = 0.0 allowed corresponds infinite hmax1 (default).  hminandhmximay changed time, take effect next change ofHis considered. option considered withmethod=\"liblsoda\"`. warnIdSort Warn ID present rxode2 assumes order parameters/iCov order parameters input dataset. warnDrop Warn column(s) supposed dropped, present. ssAtol Steady state atol convergence factor.  Can vector based state. ssRtol Steady state rtol convergence factor.  Can vector based state. safeZero Use safe zero divide log routines.  default turned may turn wish. sumType Sum type use sum() rxode2 code blocks. pairwise uses pairwise sum (fast, default) fsum uses PreciseSum package's fsum function (accurate) kahan uses Kahan correction neumaier uses Neumaier correction c uses correction: default/native summing prodType Product use prod() rxode2 blocks long double converts long double, performs multiplication converts back. double uses standard double scale multiplication. sensType Sensitivity type linCmt() model: advan Use direct advan solutions autodiff Use autodiff advan solutions forward Use forward difference solutions central Use central differences linDiff gives linear difference amount types linear compartment model parameters sensitivities calculated. named components numeric vector : \"lag\" Central compartment lag \"f\" Central compartment bioavailability \"rate\" Central compartment modeled rate \"dur\" Central compartment modeled duration \"lag2\" Depot compartment lag \"f2\" Depot compartment bioavailability \"rate2\" Depot compartment modeled rate \"dur2\" Depot compartment modeled duration linDiffCentral gives parameters use central differences linear compartment model parameters.  components linDiff resample character vector model variables resample input dataset; sampling done replacement. NULL FALSE resampling done.  TRUE resampling done covariates input dataset resampleID boolean representing resampling done individual basis TRUE (ie. whole patient selected) covariate resampled independent subject identifier FALSE.  resampleID=TRUE correlations parameters retained, resampleID=FALSE ignores patient covariate correaltions. Hence default resampleID=TRUE. maxwhile represents maximum times loop evaluated exiting.  default 100000 atolSens Sensitivity atol, can different atol liblsoda.  allows less accurate solve gradients (desired) rtolSens Sensitivity rtol, can different rtol liblsoda.  allows less accurate solve gradients (desired) ssAtolSens Sensitivity absolute tolerance (atol) calculating steady state achieved sensitivity compartments. ssRtolSens Sensitivity relative tolerance (rtol) calculating steady state achieved sensitivity compartments. simVariability determines variability simulated. NA (default) determined solver. nLlikAlloc number log likelihood endpoints used model.  allows independent log likelihood per endpoint focei nlmixr2.  likely set, though hurt anything (just may take memory larger allocations). useStdPow uses C's pow exponentiation instead R's R_pow R_pow_di.  default FALSE naTimeHandle Determines time handling happens time becomes NA: current options : ignore ignores NA time input passes . warn (default) produce warning end solve, continues solving passing NA time error stop solve parallel solved ODE (otherwise stopping can crash R) addlKeepsCov determines additional dosing items repeats dose (FALSE) keeps covariates record dose (TRUE) addlDropSs steady state doses addl specification steady state flag dropped repeated doses (TRUE) retained (FALSE) ssAtDoseTime Boolean TRUE back calculates steady concentration actual time dose, otherwise FALSE doses shifted ss2cancelAllPending TRUE SS=2 event type cancels pending doses like SS=1.  FALSE pending doses canceled SS=2 (infusions started SS=2 occurred canceled, though). envir environment look R user functions (defaults parent environment) using solve(), equivalent object argument.  specify object later argument list overwrites parameter. b using solve(), equivalent params argument.  specify params named argument, overwrites output","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"“rxSolve” solve object stores solved value special data.frame type determined returnType. default many rows sampled time points many columns system variables (defined ODEs additional assignments rxode2 model code).  also stores information call allow dynamic updating solved object. operations object similar data-frame, expand $ [[\"\"]] access operators assignment operators resolve based different parameter values, initial conditions, solver parameters, events (updating time variable). can call eventTable() methods solved object update event table resolve system equations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"rest document focus different ODE solving methods, followed core solving method's options, rxode2 event handling options, rxode2's numerical stability options, rxode2's output options, finally internal rxode2 options compatibility options.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"\"New Scaling Squaring Algorithm Matrix Exponential\", Awad H. Al-Mohy Nicholas J. Higham, August 2009 Roger B. Sidje (1998).  EXPOKIT: Software package computing matrix exponentials.  ACM - Transactions Mathematical Software 24(1), 130-156. Hindmarsh, . C. ODEPACK, Systematized Collection ODE Solvers. Scientific Computing, R. S. Stepleman et al. (Eds.), North-Holland, Amsterdam, 1983, pp. 55-64. Petzold, L. R. Automatic Selection Methods Solving Stiff Nonstiff Systems Ordinary Differential Equations. Siam J. Sci. Stat. Comput. 4 (1983), pp. 136-148. Hairer, E., Norsett, S. P., Wanner, G. Solving ordinary differential equations , nonstiff problems. 2nd edition, Springer Series Computational Mathematics, Springer-Verlag (1993).","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"Matthew Fidler, Melissa Hallow  Wenping Wang","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveFree.html","id":null,"dir":"Reference","previous_headings":"","what":"Free the C solving/parsing information. — rxSolveFree","title":"Free the C solving/parsing information. — rxSolveFree","text":"Take ODE C system free .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveFree.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Free the C solving/parsing information. — rxSolveFree","text":"","code":"rxSolveFree()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveFree.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Free the C solving/parsing information. — rxSolveFree","text":"logical indicating memory successfully freed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":null,"dir":"Reference","previous_headings":"","what":"See if the memory is installed for a solve — rxSolveSetup","title":"See if the memory is installed for a solve — rxSolveSetup","text":"See memory installed solve","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"See if the memory is installed for a solve — rxSolveSetup","text":"","code":"rxSolveSetup()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"See if the memory is installed for a solve — rxSolveSetup","text":"boolean saying memnory currently free rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"See if the memory is installed for a solve — rxSolveSetup","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":null,"dir":"Reference","previous_headings":"","what":"This function splits a function based on + or - terms — rxSplitPlusQ","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"uses parser disturb terms within functions.  example:","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"","code":"rxSplitPlusQ(x, level = 0, mult = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"x Quoted R expression splitting level Internal level parsing mult boolean split based * / expressions instead. default turned .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"character vector split expressions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"aexp(b+c)+dlog(e-f)-g*f return c(\"* exp(b + c)\", \"d * log(e - f)\", \"- g * f\")","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":null,"dir":"Reference","previous_headings":"","what":"State variables — rxState","title":"State variables — rxState","text":"returns model's compartments states.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"State variables — rxState","text":"","code":"rxState(obj = NULL, state = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"State variables — rxState","text":"obj rxode2 family objects state string indicating state compartment like lookup.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"State variables — rxState","text":"state missing, return character vector states. state string, return compartment number named state.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"State variables — rxState","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Recast model in terms of sum/prod — rxSumProdModel","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"Recast model terms sum/prod","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"","code":"rxSumProdModel(model, expand = FALSE, sum = TRUE, prod = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"model rxode2 model expand Boolean indicating expression expanded. sum Use sum(...) prod Use prod(...)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"model string prod(.) sum(.) operations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":null,"dir":"Reference","previous_headings":"","what":"Get list of supported functions — rxSupportedFuns","title":"Get list of supported functions — rxSupportedFuns","text":"Get list supported functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get list of supported functions — rxSupportedFuns","text":"","code":"rxSupportedFuns()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get list of supported functions — rxSupportedFuns","text":"list supported functions rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get list of supported functions — rxSupportedFuns","text":"","code":"rxSupportedFuns() #>   [1] \"gammafn\"              \"lgammafn\"             \"lgamma\"               #>   [4] \"loggamma\"             \"digamma\"              \"trigamma\"             #>   [7] \"tetragamma\"           \"pentagamma\"           \"cospi\"                #>  [10] \"sinpi\"                \"tanpi\"                \"log1p\"                #>  [13] \"expm1\"                \"factorial\"            \"lfactorial\"           #>  [16] \"lgamma1p\"             \"log10\"                \"log2\"                 #>  [19] \"log1pexp\"             \"phi\"                  \"pnorm\"                #>  [22] \"normcdf\"              \"qnorm\"                \"fabs\"                 #>  [25] \"pow\"                  \"R_pow\"                \"R_pow_di\"             #>  [28] \"Rx_pow_di\"            \"Rx_pow\"               \"lbeta\"                #>  [31] \"abs\"                  \"acos\"                 \"acosh\"                #>  [34] \"asin\"                 \"asinh\"                \"atan\"                 #>  [37] \"atan2\"                \"atanh\"                \"beta\"                 #>  [40] \"cos\"                  \"cosh\"                 \"erf\"                  #>  [43] \"erfc\"                 \"exp\"                  \"gamma\"                #>  [46] \"linCmtA\"              \"linCmtC\"              \"linCmtB\"              #>  [49] \"log\"                  \"polygamma\"            \"rxTBS\"                #>  [52] \"rxTBSi\"               \"rxTBSd\"               \"rxTBSd2\"              #>  [55] \"sin\"                  \"sinh\"                 \"sqrt\"                 #>  [58] \"tan\"                  \"tanh\"                 \"gammap\"               #>  [61] \"floor\"                \"round\"                \"ceil\"                 #>  [64] \"trunc\"                \"bessel_i\"             \"bessel_j\"             #>  [67] \"bessel_k\"             \"bessel_y\"             \"logspace_add\"         #>  [70] \"logspace_sub\"         \"fmax2\"                \"fmin2\"                #>  [73] \"sign\"                 \"fsign\"                \"fprec\"                #>  [76] \"fround\"               \"ftrunc\"               \"transit\"              #>  [79] \"gammaq\"               \"gammapDer\"            \"gammapInv\"            #>  [82] \"gammapInva\"           \"gammaqInv\"            \"gammaqInva\"           #>  [85] \"lowergamma\"           \"uppergamma\"           \"max\"                  #>  [88] \"min\"                  \"logit\"                \"expit\"                #>  [91] \"probit\"               \"probitInv\"            \"tlast\"                #>  [94] \"tfirst\"               \"lag\"                  \"lead\"                 #>  [97] \"dose\"                 \"podo\"                 \"dabs\"                 #> [100] \"dabs2\"                \"abs1\"                 \"dabs1\"                #> [103] \"erfinv\"               \"abs0\"                 \"dosenum\"              #> [106] \"first\"                \"last\"                 \"diff\"                 #> [109] \"is.nan\"               \"is.na\"                \"is.finite\"            #> [112] \"is.infinite\"          \"llikPois\"             \"llikPoisDlambda\"      #> [115] \"llikBinom\"            \"llikBinomDprob\"       \"llikNbinom\"           #> [118] \"llikNbinomDprob\"      \"llikNbinomMu\"         \"llikNbinomMuDmu\"      #> [121] \"llikBeta\"             \"llikBetaDshape1\"      \"llikBetaDshape2\"      #> [124] \"llikT\"                \"llikTDdf\"             \"llikTDmean\"           #> [127] \"llikTDsd\"             \"llikChisq\"            \"llikChisqDdf\"         #> [130] \"llikExp\"              \"llikExpDrate\"         \"llikF\"                #> [133] \"llikFDdf1\"            \"llikFDdf2\"            \"llikGeom\"             #> [136] \"llikGeomDprob\"        \"llikUnif\"             \"llikUnifDalpha\"       #> [139] \"llikUnifDbeta\"        \"llikWeibull\"          \"llikWeibullDshape\"    #> [142] \"llikWeibullDscale\"    \"llikGamma\"            \"llikGammaDshape\"      #> [145] \"llikGammaDrate\"       \"llikCauchy\"           \"llikCauchyDlocation\"  #> [148] \"llikCauchyDscale\"     \"llikNorm\"             \"llikNormDmean\"        #> [151] \"llikNormDsd\"          \"llikXPois\"            \"llikXPoisDlambda\"     #> [154] \"llikXBinom\"           \"llikXBinomDprob\"      \"llikXNbinomMu\"        #> [157] \"llikXNbinomMuDmu\"     \"llikXNbinom\"          \"llikXNbinomDprob\"     #> [160] \"llikXBeta\"            \"llikXBetaDshape1\"     \"llikXBetaDshape2\"     #> [163] \"llikXT\"               \"llikXTDdf\"            \"llikXTDmean\"          #> [166] \"llikXTDsd\"            \"llikXChisq\"           \"llikXChisqDdf\"        #> [169] \"llikXExp\"             \"llikXExpDrate\"        \"llikXF\"               #> [172] \"llikXFDdf1\"           \"llikXFDdf2\"           \"llikXGeom\"            #> [175] \"llikXGeomDprob\"       \"llikXUnif\"            \"llikXUnifDalpha\"      #> [178] \"llikXUnifDbeta\"       \"llikXWeibull\"         \"llikXWeibullDshape\"   #> [181] \"llikXWeibullDscale\"   \"llikXGamma\"           \"llikXGammaDshape\"     #> [184] \"llikXGammaDrate\"      \"llikXCauchy\"          \"llikXCauchyDlocation\" #> [187] \"llikXCauchyDscale\"    \"llikXNorm\"            \"llikXNormDmean\"       #> [190] \"llikXNormDsd\"         \"linCmt\"               \"rnorm\"                #> [193] \"rxnorm\"               \"rxbinom\"              \"rbinom\"               #> [196] \"rxcauchy\"             \"rcauchy\"              \"rchisq\"               #> [199] \"rxchisq\"              \"rexp\"                 \"rxexp\"                #> [202] \"rbeta\"                \"rxbeta\"               \"rgeom\"                #> [205] \"rxgeom\"               \"rxpois\"               \"rpois\"                #> [208] \"rxt\"                  \"rt\""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":null,"dir":"Reference","previous_headings":"","what":"Respect suppress messages — rxSuppressMsg","title":"Respect suppress messages — rxSuppressMsg","text":"turns silent REprintf C suppressMessages() turned . makes REprintf act like messages R, can suppressed suppressMessages()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Respect suppress messages — rxSuppressMsg","text":"","code":"rxSuppressMsg()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Respect suppress messages — rxSuppressMsg","text":"Nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Respect suppress messages — rxSuppressMsg","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Respect suppress messages — rxSuppressMsg","text":"","code":"# rxSupressMsg() is called with rxode2()  # Note the errors are output to the console  try(rxode2(\"d/dt(matt)=/3\"), silent = TRUE) #>   #>    # When using suppressMessages, the output is suppressed  suppressMessages(try(rxode2(\"d/dt(matt)=/3\"), silent = TRUE))  # In rxode2, we use REprintf so that interrupted threads do not crash R # if there is a user interrupt. This isn't captured by R's messages, but # This interface allows the `suppressMessages()` to suppress the C printing # as well  # If you  want to suppress messages from rxode2 in other packages, you can use # this function"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":null,"dir":"Reference","previous_headings":"","what":"Get Omega^-1 and derivatives — rxSymInvChol","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"Get Omega^-1 derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"","code":"rxSymInvChol(   invObjOrMatrix,   theta = NULL,   type = \"cholOmegaInv\",   thetaNumber = 0L )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"invObjOrMatrix Object inverse-type calculations.  matrix, setup object inversion rxSymInvCholCreate() default arguments return reactive s3 object.  Otherwise, use inversion object calculate requested derivative/inverse. theta Thetas used calculation.  missing (NULL), special s3 class created returned access Omega^1 objects needed cache based theta used. type type object.  Currently following types supported: cholOmegaInv gives Cholesky decomposition Omega Inverse matrix. omegaInv gives Omega Inverse matrix. d(omegaInv) gives d(Omega^-1) withe respect theta parameter specified thetaNumber. d(D) gives d(diagonal(Omega^-1)) respect theta parameter specified thetaNumber parameter thetaNumber types d(omegaInv) d(D), theta number derivative taken .  must positive 1 number thetas defining Omega matrix.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"Matrix based parameters environment matrixes calculated variables omega, omegaInv, dOmega, dOmegaInv.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"Creates object calculating Omega/Omega^-1 derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"","code":"rxSymInvCholCreate(   mat,   diag.xform = c(\"sqrt\", \"log\", \"identity\"),   create.env = TRUE,   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"mat Initial Omega matrix diag.xform transformation diagonal elements OMEGA. chol(Omega^-1) create.env -- Create environment calculate inverses. (default TRUE) envir -- Environment evaluate function, bu default parent frame.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"rxSymInv object rxSymInv environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholN.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","title":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","text":"Return dimension built-derivatives/inverses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholN.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","text":"","code":"rxSymInvCholN()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholN.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","text":"dimension built-derivatives/inverses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":null,"dir":"Reference","previous_headings":"","what":"Sync options with rxode2 variables — rxSyncOptions","title":"Sync options with rxode2 variables — rxSyncOptions","text":"Accessing rxode2 options via getOption slows solving.  allows options synced variables.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Sync options with rxode2 variables — rxSyncOptions","text":"","code":"rxSyncOptions(setDefaults = c(\"none\", \"permissive\", \"strict\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Sync options with rxode2 variables — rxSyncOptions","text":"setDefaults setup rxode2's default solving options following options: \"none\" leave options alone \"permissive\" permissive option set similar R language specifications. \"strict\" strict option set similar original rxode2(). requires semicolons end lines equals assignment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Sync options with rxode2 variables — rxSyncOptions","text":"nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Sync options with rxode2 variables — rxSyncOptions","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyntaxFunctions.html","id":null,"dir":"Reference","previous_headings":"","what":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","title":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","text":"list description Rode supported syntax functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyntaxFunctions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","text":"","code":"rxSyntaxFunctions"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyntaxFunctions.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","text":"data frame 3 columns 98 rows Function Reserved function Name Description Description function Aliases Function Aliases","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTempDir.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the rxode2 temporary directory — rxTempDir","title":"Get the rxode2 temporary directory — rxTempDir","text":"Get rxode2 temporary directory","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTempDir.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the rxode2 temporary directory — rxTempDir","text":"","code":"rxTempDir()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTempDir.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the rxode2 temporary directory — rxTempDir","text":"rxode2 temporary directory.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":null,"dir":"Reference","previous_headings":"","what":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"rxTheme ggplot2 theme rxode2 plots","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"","code":"rxTheme(   base_size = 11,   base_family = \"\",   base_line_size = base_size/22,   base_rect_size = base_size/22,   grid = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"base_size base font size, given pts. base_family base font family base_line_size base size line elements base_rect_size base size rect elements grid Boolean indicating grid (TRUE) (FALSE). also character indicating x y.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"ggplot2 theme used rxode2","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":null,"dir":"Reference","previous_headings":"","what":"rxode2 to symengine environment — rxToSE","title":"rxode2 to symengine environment — rxToSE","text":"rxode2 symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxode2 to symengine environment — rxToSE","text":"","code":"rxToSE(   x,   envir = NULL,   progress = FALSE,   promoteLinSens = TRUE,   parent = parent.frame() )  .rxToSE(x, envir = NULL, progress = FALSE)  rxFromSE(   x,   unknownDerivatives = c(\"forward\", \"central\", \"error\"),   parent = parent.frame() )  .rxFromSE(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxode2 to symengine environment — rxToSE","text":"x expression envir default NULL; Environment put symengine variables . progress shows progress bar true. promoteLinSens Promote solved linear compartment systems sensitivity-based solutions. parent parent environment look R-based user functions unknownDerivatives handling derivatives unknown functions, translator translate different types numeric derivatives.  currently supported methods :","code":"- `forward` for forward differences - `central` for central differences - `error` for throwing an error for unknown derivatives"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxode2 to symengine environment — rxToSE","text":"rxode2 symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxode2 to symengine environment — rxToSE","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":null,"dir":"Reference","previous_headings":"","what":"Translate the model to C code if needed — rxTrans","title":"Translate the model to C code if needed — rxTrans","text":"function translates model C code, needed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Translate the model to C code if needed — rxTrans","text":"","code":"rxTrans(   model,   modelPrefix = \"\",   md5 = \"\",   modName = NULL,   modVars = FALSE,   ... )  # S3 method for default rxTrans(   model,   modelPrefix = \"\",   md5 = \"\",   modName = NULL,   modVars = FALSE,   ... )  # S3 method for character rxTrans(   model,   modelPrefix = \"\",   md5 = \"\",   modName = NULL,   modVars = FALSE,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Translate the model to C code if needed — rxTrans","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . modelPrefix Prefix model functions compiled make sure multiple rxode2 objects can coexist R session. md5 md5 model parsing, used embed md5 DLL, provide functions like rxModelVars(). modName string used model name. string used naming various aspects computations, including generating C symbol names, dynamic libraries, etc. Therefore, necessary modName consists simple ASCII alphanumeric characters starting letter. modVars returns model variables instead named vector translated properties. ... Ignored parameters.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Translate the model to C code if needed — rxTrans","text":"named vector translated model properties including type jacobian specified, C function prefixes, well C functions names called compiled model.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Translate the model to C code if needed — rxTrans","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":null,"dir":"Reference","previous_headings":"","what":"Compress/Decompress rxode2 ui — rxUiDecompress","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"Compress/Decompress rxode2 ui","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"","code":"rxUiDecompress(ui)  rxUiCompress(ui)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"ui rxode2 ui object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"compressed decompressed rxui object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"","code":"one.cmt <- function() {   ini({     ## You may label each parameter with a comment     tka <- 0.45 # Log Ka     tcl <- log(c(0, 2.7, 100)) # Log Cl     ## This works with interactive models     ## You may also label the preceding line with label(\"label text\")     tv <- 3.45; label(\"log V\")     ## the label(\"Label name\") works with all models     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     linCmt() ~ add(add.sd) | tmp   }) }  f <- rxode2(one.cmt) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' print(class(f)) #> [1] \"rxUi\" \"raw\"  print(is.environment(f)) #> [1] FALSE  f  <- rxUiDecompress(f) print(class(f)) #> [1] \"rxUi\" print(is.environment(f)) #> [1] TRUE  f  <- rxUiCompress(f) print(class(f)) #> [1] \"rxUi\" \"raw\"  print(is.environment(f)) #> [1] FALSE"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":null,"dir":"Reference","previous_headings":"","what":"S3 for getting information from UI model — rxUiGet.cmtLines","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"S3 getting information UI model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"","code":"# S3 method for cmtLines rxUiGet(x, ...)  # S3 method for dvidLine rxUiGet(x, ...)  # S3 method for paramsLine rxUiGet(x, ...)  # S3 method for simulationSigma rxUiGet(x, ...)  # S3 method for simulationModel rxUiGet(x, ...)  # S3 method for symengineModelNoPrune rxUiGet(x, ...)  # S3 method for symengineModelPrune rxUiGet(x, ...)  # S3 method for simulationIniModel rxUiGet(x, ...)  rxUiGet(x, ...)  # S3 method for params rxUiGet(x, ...)  # S3 method for theta rxUiGet(x, ...)  # S3 method for lstChr rxUiGet(x, ...)  # S3 method for omega rxUiGet(x, ...)  # S3 method for funTxt rxUiGet(x, ...)  # S3 method for allCovs rxUiGet(x, ...)  # S3 method for muRefTable rxUiGet(x, ...)  # S3 method for multipleEndpoint rxUiGet(x, ...)  # S3 method for funPrint rxUiGet(x, ...)  # S3 method for fun rxUiGet(x, ...)  # S3 method for md5 rxUiGet(x, ...)  # S3 method for ini rxUiGet(x, ...)  # S3 method for iniFun rxUiGet(x, ...)  # S3 method for modelFun rxUiGet(x, ...)  # S3 method for model rxUiGet(x, ...)  # S3 method for modelDesc rxUiGet(x, ...)  # S3 method for thetaLower rxUiGet(x, ...)  # S3 method for thetaUpper rxUiGet(x, ...)  # S3 method for lhsVar rxUiGet(x, ...)  # S3 method for varLhs rxUiGet(x, ...)  # S3 method for lhsEta rxUiGet(x, ...)  # S3 method for lhsTheta rxUiGet(x, ...)  # S3 method for lhsCov rxUiGet(x, ...)  # S3 method for etaLhs rxUiGet(x, ...)  # S3 method for thetaLhs rxUiGet(x, ...)  # S3 method for covLhs rxUiGet(x, ...)  # S3 method for default rxUiGet(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"x list (UIenvironment, exact).  UI environment parsed function rxode2.  exact boolean says exact match required. ... arguments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"value requested UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":null,"dir":"Reference","previous_headings":"","what":"Unloads all rxode2 compiled DLLs — rxUnloadAll","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"Unloads rxode2 compiled DLLs","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"","code":"rxUnloadAll()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"List rxode2 dlls still loaded boolean rxode2 dlls unloaded","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"","code":"print(rxUnloadAll()) #> NULL"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":null,"dir":"Reference","previous_headings":"","what":"Use model object in your package — rxUse","title":"Use model object in your package — rxUse","text":"Use model object package","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Use model object in your package — rxUse","text":"","code":"rxUse(obj, overwrite = TRUE, compress = \"bzip2\", internal = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Use model object in your package — rxUse","text":"obj model save. overwrite default, use_data() overwrite existing files. really want , set TRUE. compress Choose type compression used save(). one \"gzip\", \"bzip2\", \"xz\". internal run internally.  default FALSE","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Use model object in your package — rxUse","text":"Nothing; used side effects called user","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":null,"dir":"Reference","previous_headings":"","what":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"Validate rxode2 allows easy validation/qualification nlmixr running testing suite system.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"","code":"rxValidate(type = NULL, skipOnCran = TRUE)  rxTest(type = NULL, skipOnCran = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"type Type test filter test type, expression, evaluate contents, respecting skipOnCran skipOnCran TRUE skip test CRAN.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":null,"dir":"Reference","previous_headings":"","what":"Version and repository for this dparser package. — rxVersion","title":"Version and repository for this dparser package. — rxVersion","text":"Version repository dparser package.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Version and repository for this dparser package. — rxVersion","text":"","code":"rxVersion(   extra = \"\",   echo = FALSE,   version = sessionInfo()$otherPkgs$rxode2$Version )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Version and repository for this dparser package. — rxVersion","text":"extra extra text display logo echo Boolean echo text logo, default FALSE version Version display/return","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Version and repository for this dparser package. — rxVersion","text":"character vector version repository.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Version and repository for this dparser package. — rxVersion","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate beta variable from threefry generator — rxbeta","title":"Simulate beta variable from threefry generator — rxbeta","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate beta variable from threefry generator — rxbeta","text":"","code":"rxbeta(shape1, shape2, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate beta variable from threefry generator — rxbeta","text":"shape1, shape2 non-negative parameters Beta distribution. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate beta variable from threefry generator — rxbeta","text":"beta random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate beta variable from threefry generator — rxbeta","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate beta variable from threefry generator — rxbeta","text":"","code":"# \\donttest{  ## Use threefry engine  rxbeta(0.5, 0.5, n = 10) # with rxbeta you have to explicitly state n #>  [1] 0.477852330 0.303485792 0.039549351 0.172127808 0.029142179 0.228528849 #>  [7] 0.002045024 0.123955210 0.122974648 0.003892215 rxbeta(5, 1, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 0.9837921 0.9619446 0.9528713 0.8186334 0.8958941 0.9547040 0.9377216 #>  [8] 0.8771444 0.9122368 0.9036474  rxbeta(1, 3) #> [1] 0.4918356   ## This example uses `rxbeta` directly in the model  rx <- function() {   model({     a <- rxbeta(2, 2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Binomial variable from threefry generator — rxbinom","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"","code":"rxbinom(size, prob, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"size number trials (zero ). prob probability success trial. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"binomial random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"","code":"# \\donttest{ ## Use threefry engine  rxbinom(10, 0.9, n = 10) # with rxbinom you have to explicitly state n #>  [1] 10  9  9 10  9  9  9 10 10 10 rxbinom(3, 0.5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 0 3 0 2 2 1 2 1 1 3  rxbinom(4, 0.7) #> [1] 2   ## This example uses `rxbinom` directly in the model  rx <- function() {   model({     a <- rxbinom(1, 0.5)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Cauchy variable from threefry generator — rxcauchy","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"","code":"rxcauchy(location = 0, scale = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"location, scale location scale parameters. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"Cauchy random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"","code":"# \\donttest{  ## Use threefry engine  rxcauchy(0, 1, n = 10) # with rxcauchy you have to explicitly state n #>  [1] -4.058532e-01 -3.735548e-01  2.292820e+00  1.715716e+01 -3.247341e+04 #>  [6] -1.381140e+00  6.254309e-02  1.008986e+02  1.192805e+01  5.395418e-01 rxcauchy(0.5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  1.35507665 -0.09529856  0.15545628  0.52832597  2.91300334  0.67517336 #>  [7]  2.74087058  0.92316592  1.35864663 -0.36322597  rxcauchy(3) #> [1] 6.705508   ## This example uses `rxcauchy` directly in the model  rx <- function() {   model({     a <- rxcauchy(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate chi-squared variable from threefry generator — rxchisq","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"","code":"rxchisq(df, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"df degrees freedom (non-negative, can non-integer). n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"chi squared random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"","code":"# \\donttest{  ## Use threefry engine  rxchisq(0.5, n = 10) # with rxchisq you have to explicitly state n #>  [1] 8.310255e-01 5.056546e-04 1.283680e-02 1.156829e-01 1.791516e-01 #>  [6] 1.262242e+00 7.069398e-01 7.941145e-01 2.610348e-07 7.152674e+00 rxchisq(5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 2.285712 4.831273 3.575691 8.457173 6.136148 4.831445 3.387999 1.390385 #>  [9] 4.401743 4.226414  rxchisq(1) #> [1] 0.4391989   ## This example uses `rxchisq` directly in the model  rx <- function() {   model({     a <- rxchisq(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate exponential variable from threefry generator — rxexp","title":"Simulate exponential variable from threefry generator — rxexp","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate exponential variable from threefry generator — rxexp","text":"","code":"rxexp(rate, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate exponential variable from threefry generator — rxexp","text":"rate vector rates. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate exponential variable from threefry generator — rxexp","text":"exponential random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate exponential variable from threefry generator — rxexp","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate exponential variable from threefry generator — rxexp","text":"","code":"# \\donttest{  ## Use threefry engine  rxexp(0.5, n = 10) # with rxexp you have to explicitly state n #>  [1] 2.3809833 1.2842193 2.9579776 0.2992783 2.2362850 1.1166611 0.1232554 #>  [8] 0.3223169 3.1258548 0.1552611 rxexp(5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 0.18422563 0.04552514 0.03749666 0.07330912 0.13716012 0.77328162 #>  [7] 0.40892297 0.07167627 0.24915418 0.04749041  rxexp(1) #> [1] 1.054455   ## This example uses `rxexp` directly in the model  rx <- function() {   model({     a <- rxexp(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate F variable from threefry generator — rxf","title":"Simulate F variable from threefry generator — rxf","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate F variable from threefry generator — rxf","text":"","code":"rxf(df1, df2, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate F variable from threefry generator — rxf","text":"df1, df2 degrees freedom.  Inf allowed. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate F variable from threefry generator — rxf","text":"f random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate F variable from threefry generator — rxf","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate F variable from threefry generator — rxf","text":"","code":"# \\donttest{  ## Use threefry engine  rxf(0.5, 0.5, n = 10) # with rxf you have to explicitly state n #>  [1] 8.504451e+00 3.630793e+00 8.678236e-03 3.880547e+00 2.243042e+06 #>  [6] 2.801723e-01 4.225989e-01 1.000520e-01 3.304096e-01 1.149793e+00 rxf(5, 1, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  2.2338907 11.2163069  0.1898719  4.8782950  2.2084744  7.4676924 #>  [7]  1.2271939  7.7742898  0.1681263  4.3235439  rxf(1, 3) #> [1] 0.01338054   ## This example uses `rxf` directly in the model  rx <- function() {   model({     a <- rxf(2, 2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate gamma variable from threefry generator — rxgamma","title":"Simulate gamma variable from threefry generator — rxgamma","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate gamma variable from threefry generator — rxgamma","text":"","code":"rxgamma(shape, rate = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate gamma variable from threefry generator — rxgamma","text":"shape shape gamma random variable rate alternative way specify scale. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate gamma variable from threefry generator — rxgamma","text":"gamma random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate gamma variable from threefry generator — rxgamma","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate gamma variable from threefry generator — rxgamma","text":"","code":"# \\donttest{  ## Use threefry engine  rxgamma(0.5, n = 10) # with rxgamma you have to explicitly state n #>  [1] 1.355794e-06 1.184154e-01 9.402169e-03 2.871131e-01 5.795578e-01 #>  [6] 1.362098e+00 2.274397e-01 2.389341e-01 2.283278e-02 2.453927e-01 rxgamma(5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 6.081028 4.553699 4.216456 3.053786 2.837575 3.992559 2.974116 4.653716 #>  [9] 5.765096 3.788091  rxgamma(1) #> [1] 0.4450636   ## This example uses `rxbeta` directly in the model  rx <- function() {   model({     a <- rxgamma(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate geometric variable from threefry generator — rxgeom","title":"Simulate geometric variable from threefry generator — rxgeom","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate geometric variable from threefry generator — rxgeom","text":"","code":"rxgeom(prob, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate geometric variable from threefry generator — rxgeom","text":"prob probability success trial. 0 < prob <= 1. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate geometric variable from threefry generator — rxgeom","text":"geometric random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate geometric variable from threefry generator — rxgeom","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate geometric variable from threefry generator — rxgeom","text":"","code":"# \\donttest{  ## Use threefry engine  rxgeom(0.5, n = 10) # with rxgeom you have to explicitly state n #>  [1] 1 1 0 0 0 1 3 2 0 0 rxgeom(0.25, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  0  8  0 13  1  5  6  1 11  6  rxgeom(0.75) #> [1] 0   ## This example uses `rxgeom` directly in the model  rx <- function() {   model({     a <- rxgeom(0.24)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Binomial variable from threefry generator — rxnbinom","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"","code":"rxnbinom(size, prob, n = 1L, ncores = 1L)  rxnbinomMu(size, mu, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"size target number successful trials, dispersion     parameter (shape parameter gamma mixing distribution).     Must strictly positive, need integer. prob probability success trial. 0 < prob <= 1. n number observations.  length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks mu alternative parametrization via mean: see ‘Details’.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"negative binomial random deviates. Note rxbinom2 uses mu parameterization rxbinom uses prob parameterization (mu=size/(prob+size))","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"","code":"# \\donttest{ ## Use threefry engine  rxnbinom(10, 0.9, n = 10) # with rxbinom you have to explicitly state n #>  [1]  9 10  9  8 10 10  8  9  9 10 rxnbinom(3, 0.5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 3 2 1 3 3 1 3 3 2 2  rxnbinom(4, 0.7) #> [1] 4  # use mu parameter rxnbinomMu(40, 40, n=10) #>  [1] 20 24 20 21 21 22 20 21 20 12  ## This example uses `rxbinom` directly in the model  rx <- function() {   model({     a <- rxnbinom(10, 0.5)   }) }  et <- et(1, id = 1:100)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  rx <- function() {   model({     a <- rxnbinomMu(10, 40)   }) }  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate random normal variable from threefry generator — rxnormV","title":"Simulate random normal variable from threefry generator — rxnormV","text":"Simulate random normal variable threefry generator","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate random normal variable from threefry generator — rxnormV","text":"","code":"rxnormV(mean = 0, sd = 1, n = 1L, ncores = 1L)  rxnorm(mean = 0, sd = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate random normal variable from threefry generator — rxnormV","text":"mean vector means. sd vector standard deviations. n number observations ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate random normal variable from threefry generator — rxnormV","text":"normal random number deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate random normal variable from threefry generator — rxnormV","text":"","code":"# \\donttest{ ## Use threefry engine  rxnorm(n = 10) # with rxnorm you have to explicitly state n #>  [1]  0.37076053 -0.96034099  0.02749010 -0.82762453  0.58984915 -1.80511477 #>  [7]  0.50379739 -0.67643890  1.45510439  0.04165521 rxnorm(n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] -0.51162593 -1.13782819  0.06566640 -0.12404287  1.79306612  0.83099897 #>  [7] -0.44485847  0.06223717  1.88310497  0.22993201  rxnorm(2, 3) ## The first 2 arguments are the mean and standard deviation #> [1] -0.9022625   ## This example uses `rxnorm` directly in the model  rx <- function() {   model({     a <- rxnorm()   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":null,"dir":"Reference","previous_headings":"","what":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"Set function body rxUi object retaining object information (like data)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"","code":"rxode2(x, envir = environment(x)) <- value  # S3 method for `function` rxode2(x, envir = environment(x)) <- value  # S3 method for default rxode2(x, envir = environment(x)) <- value  rxode(x, envir = environment(x)) <- value  RxODE(x, envir = environment(x)) <- value"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"x rxUi object envir environment assignment ocurs value value assigned","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"rxode2 ui/function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"","code":"one.compartment <- function() {   ini({     tka <- log(1.57); label(\"Ka\")     tcl <- log(2.72); label(\"Cl\")     tv <- log(31.5); label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  two.compartment <- function() {   ini({     lka <- 0.45 ; label(\"Absorption rate (Ka)\")     lcl <- 1 ; label(\"Clearance (CL)\")     lvc  <- 3 ; label(\"Central volume of distribution (V)\")     lvp  <- 5 ; label(\"Peripheral volume of distribution (Vp)\")     lq  <- 0.1 ; label(\"Intercompartmental clearance (Q)\")     propSd <- 0.5 ; label(\"Proportional residual error (fraction)\")   })   model({     ka <- exp(lka)     cl <- exp(lcl)     vc <- exp(lvc)     vp <- exp(lvp)     q  <- exp(lq)     kel <- cl/vc     k12 <- q/vc     k21 <- q/vp     d/dt(depot) <- -ka*depot     d/dt(central) <-  ka*depot - kel*central - k12*central + k21*peripheral1     d/dt(peripheral1) <- k12*central - k21*peripheral1     cp <- central / vc     cp ~ prop(propSd)   }) }  ui <- rxode2(one.compartment) #>   #>    rxode2(ui) <- two.compartment"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Create an ODE-based model specification — rxode2","title":"Create an ODE-based model specification — rxode2","text":"Create dynamic ODE-based model object suitably translation fast C code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Create an ODE-based model specification — rxode2","text":"","code":"rxode2(   model,   modName = basename(wd),   wd = getwd(),   filename = NULL,   extraC = NULL,   debug = FALSE,   calcJac = NULL,   calcSens = NULL,   collapseModel = FALSE,   package = NULL,   ...,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   indLin = FALSE,   verbose = FALSE,   fullPrint = getOption(\"rxode2.fullPrint\", FALSE),   envir = parent.frame() )  RxODE(   model,   modName = basename(wd),   wd = getwd(),   filename = NULL,   extraC = NULL,   debug = FALSE,   calcJac = NULL,   calcSens = NULL,   collapseModel = FALSE,   package = NULL,   ...,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   indLin = FALSE,   verbose = FALSE,   fullPrint = getOption(\"rxode2.fullPrint\", FALSE),   envir = parent.frame() )  rxode(   model,   modName = basename(wd),   wd = getwd(),   filename = NULL,   extraC = NULL,   debug = FALSE,   calcJac = NULL,   calcSens = NULL,   collapseModel = FALSE,   package = NULL,   ...,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   indLin = FALSE,   verbose = FALSE,   fullPrint = getOption(\"rxode2.fullPrint\", FALSE),   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Create an ODE-based model specification — rxode2","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . modName string used model name. string used naming various aspects computations, including generating C symbol names, dynamic libraries, etc. Therefore, necessary modName consists simple ASCII alphanumeric characters starting letter. wd character string working directory create subdirectory according modName. specified, subdirectory named “modName.d” created populated C file, dynamic loading library, plus various working files. missing, files created (removed) temporary directory, rxode2 DLL model created current directory named rx_????_platform, example rx_129f8f97fb94a87ca49ca8dafe691e1e_i386.dll filename file name connection object ODE-based model specification resides. one model filename may specified. extraC Extra c code include model.  can useful specify functions model.  C functions usually take double precision arguments, return double precision values. debug boolean indicating executable compiled verbose debugging information turned . calcJac boolean indicating rxode2 calculate Jacobain according specified ODEs. calcSens boolean indicating rxode2 calculate sensitivities according specified ODEs. collapseModel boolean indicating rxode2 remove LHS variables calculating sensitivities. package Package name pre-compiled binaries. ... ignored arguments. linCmtSens method calculate linCmt() solutions indLin Calculate inductive linearization matrices compile inductive linearization support. verbose TRUE verbose linear compartmental model fullPrint using printf within model, TRUE print every step (except /indLin), otherwise FALSE print calculating d/dt envir environment look R user functions (defaults parent environment)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Create an ODE-based model specification — rxode2","text":"object (environment) class rxode2 (see Chambers Temple Lang (2001)) consisting following list strings functions:","code":"* `model` a character string holding the source model specification. * `get.modelVars`a function that returns a list with 3 character     vectors, `params`, `state`, and `lhs` of variable names used in the model     specification. These will be output when the model is computed (i.e., the ODE solved by integration).    * `solve`{this function solves (integrates) the ODE. This       is done by passing the code to [rxSolve()].       This is as if you called `rxSolve(rxode2object, ...)`,       but returns a matrix instead of a rxSolve object.        `params`: a numeric named vector with values for every parameter       in the ODE system; the names must correspond to the parameter       identifiers used in the ODE specification;        `events`: an `eventTable` object describing the       input (e.g., doses) to the dynamic system and observation       sampling time points (see  [eventTable()]);        `inits`: a vector of initial values of the state variables       (e.g., amounts in each compartment), and the order in this vector       must be the same as the state variables (e.g., PK/PD compartments);         `stiff`: a logical (`TRUE` by default) indicating whether       the ODE system is stiff or not.        For stiff ODE systems (`stiff = TRUE`), `rxode2` uses       the LSODA (Livermore Solver for Ordinary Differential Equations)       Fortran package, which implements an automatic method switching       for stiff and non-stiff problems along the integration interval,       authored by Hindmarsh and Petzold (2003).        For non-stiff systems (`stiff = FALSE`), `rxode2` uses `DOP853`,       an explicit Runge-Kutta method of order 8(5, 3) of Dormand and Prince       as implemented in C by Hairer and Wanner (1993).        `trans_abs`: a logical (`FALSE` by default) indicating       whether to fit a transit absorption term       (TODO: need further documentation and example);        `atol`: a numeric absolute tolerance (1e-08 by default);        `rtol`: a numeric relative tolerance (1e-06 by default).        The output of \\dQuote{solve} is a matrix with as many rows as there       are sampled time points and as many columns as system variables       (as defined by the ODEs and additional assignments in the rxode2 model           code).}    * `isValid` a function that (naively) checks for model validity,       namely that the C object code reflects the latest model       specification.   * `version` a string with the version of the `rxode2`       object (not the package).   * `dynLoad` a function with one `force = FALSE` argument       that dynamically loads the object code if needed.   * `dynUnload` a function with no argument that unloads       the model object code.   * `delete` removes all created model files, including C and DLL files.       The model object is no longer valid and should be removed, e.g.,       `rm(m1)`.   * `run` deprecated, use `solve`.   * `get.index` deprecated.   * `getObj` internal (not user callable) function."},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Create an ODE-based model specification — rxode2","text":"Rx name rxode2 meant suggest abbreviation Rx medical prescription, thus suggest package emphasis pharmacometrics modeling, including pharmacokinetics (PK), pharmacodynamics (PD), disease progression, drug-disease modeling, etc. ODE-based model specification may coded inside four places: Inside rxode2({}) block statements:      Inside rxode2(\"\") string statement:      file name loaded rxode2:    model function can parsed rxode2:    block statement, character string text file internal rxode2 compilation manager translates ODE system C, compiles loads R session. call rxode2 produces object class rxode2 consists list-like structure (environment) various member functions. last type model (model function), call rxode2 creates parsed rxode2 ui can translated rxode2 compilation model.         type function required nlmixr2 estimation can extended modified model piping. reason focused documentation. basic model specification consists one statements optionally terminated semi-colons ; optional comments (comments delimited # end--line). block statements set statements delimited curly braces, { ... }. Statements can either assignments, conditional /else /else, loops (can exited break), special statements, printing statements (debugging/testing). Assignment statements can : simple assignments, left hand identifier (.e., variable) special time-derivative assignments, left hand specifies change amount corresponding state variable (compartment) respect time e.g., d/dt(depot): special initial-condition assignments left hand specifies compartment initial condition specified, e.g. depot(0) = 0 special model event changes including bioavailability (f(depot)=1), lag time (alag(depot)=0), modeled rate (rate(depot)=2) modeled duration (dur(depot)=2). example model features event specification modeled infusions rxode2 data specification found rxode2 events vignette. special change point syntax, model times. model times specified mtime(var)=time special Jacobian-derivative assignments, left hand specifies change compartment ode respect variable. example, d/dt(y) = dy, Jacobian compartment can specified df(y)/dy(dy) = 1. may advantage obtaining solution specifying Jacobian stiff ODE systems. However, stiff systems tried LSODA, actually slightly slowed solving. Note assignment can done =, <- ~. assigning ~ operator, simple assignments time-derivative assignments output. Note rxode2 model functions assignment ~ can also overloaded residual distribution specification. Special statements can : Compartment declaration statements, can change default dosing compartment assumed compartment number(s) well add extra compartment names end (useful multiple-endpoint nlmixr models); specified cmt(compartmentName) Parameter declaration statements, can make sure input parameters certain order instead ordering parameters order parsed. useful keeping parameter order using 2 different ODE models. specified param(par1, par2,...) example model shown :   Expressions assignment statements can numeric logical. Numeric expressions can include following numeric operators +, -, *, /, ^ mathematical functions defined C R math libraries (e.g., fabs, exp, log, sin, abs). may also access R’s functions R math libraries, like lgammafn log gamma function. rxode2 syntax case-sensitive, .e., ABC different abc, Abc, ABc, etc.","code":"library(rxode2) mod <- rxode2({   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; }) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_36fd6c4312c660ddf5390102e91a4bb3_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^ mod <- rxode2(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \") ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_4fdd80e5dd5fd33d1951b83e06f7ad84_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^ writeLines(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \", \"modelFile.rxode2\") mod <- rxode2(filename='modelFile.rxode2') unlink(\"modelFile.rxode2\") mod <- function() {   model({     # simple assignment     C2 <- centr/V2      # time-derivative assignment     d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;   }) }  mod <- rxode2(mod) # or simply mod() if the model is at the end of the function  # These model functions often have residual components and initial # (`ini({})`) conditions attached as well. For example the # theophylline model can be written as:  one.compartment <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  # after parsing the model mod <- one.compartment() mod$simulationModel ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_bbee652505e8bddf28c7e688baef12d9_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^  ## rxode2 2.1.2.9000 model named rx_bbee652505e8bddf28c7e688baef12d9 model (ready).  ## x$state: depot, center ## x$stateExtra: cp ## x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp ## x$lhs: ka, cl, v, cp, ipredSim, sim # or  mod$simulationIniModel ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_fe01e26500a95dd1ba4d8972bc71bc7a_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^  ## rxode2 2.1.2.9000 model named rx_fe01e26500a95dd1ba4d8972bc71bc7a model (ready).  ## x$state: depot, center ## x$stateExtra: cp ## x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp ## x$lhs: ka, cl, v, cp, ipredSim, sim # simple assignment    C2 <- centr/V2     # time-derivative assignment    d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"identifiers","dir":"Reference","previous_headings":"","what":"Identifiers","title":"Create an ODE-based model specification — rxode2","text":"Like R, Identifiers (variable names) may consist one alphanumeric, underscore _ period . characters, first character digit underscore _. Identifiers model specification can refer : State variables dynamic system (e.g., compartments pharmacokinetics model). Implied input variable, t (time), tlast (last time point), podo (oral dose, undocumented case absorption transit models). Special constants like pi R’s predefined constants. Model parameters (e.g., ka rate absorption, CL clearance, etc.) Others, created assignments part model specification; referred LHS (left-hand side) variable. Currently, rxode2 modeling language recognizes system state variables “parameters”, thus, values need passed R ODE model (e.g., age) either passed params argument integrator function rxSolve() supplied event data-set. certain variable names rxode2 event tables. avoid confusion, following event table-related items assigned, used state can accessed rxode2 code: cmt dvid addl ss rate id However following variables used model specification: evid ii Sometimes rxode2 generates variables fed back rxode2. Similarly, nlmixr2 generates variables used nlmixr estimation simulation. variables start either rx nlmixr prefixes. avoid problems, suggested use variables starting either rx nlmixr prefixes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"logical-operators","dir":"Reference","previous_headings":"","what":"Logical Operators","title":"Create an ODE-based model specification — rxode2","text":"Logical operators support standard R operators ==, != >= <= > <. Like R can () () statements, ifelse() expressions. Additionally can standard assignment. instance, following valid:   Notice can also use character expressions comparisons. convenience comes cost since character comparisons slower numeric expressions. Unlike R, .numeric .integer logical statements needed, cause syntax error try use function.","code":"cov1 = covm*(sexf == \"female\") + covm*(sexf != \"female\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"supported-functions","dir":"Reference","previous_headings":"","what":"Supported functions","title":"Create an ODE-based model specification — rxode2","text":"supported functions rxode2 can seen rxSupportedFuns(). brief description built-functions following table: Note lag(cmt) = equivalent alag(cmt) = = lag(wt)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"reserved-keywords","dir":"Reference","previous_headings":"","what":"Reserved keywords","title":"Create an ODE-based model specification — rxode2","text":"reserved keywords rxode2 model. following table: Note rxFlag always output 11 calc_lhs since final variables calculated, though can tweak test certain parts rxode2 using flag.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"residual-functions-when-using-rxode-functions","dir":"Reference","previous_headings":"","what":"Residual functions when using rxode2 functions","title":"Create an ODE-based model specification — rxode2","text":"addition ~ hiding output certain types output, also used specify residual output endpoint input rxode2 model function (includes residual model({}) block). specifications form:   Indicating variable var variable represents individual central tendencies model also represents compartment specification data-set. can also change compartment name using | syntax, :   case var represents central tendency cmt represents compartment dvid specification.","code":"var ~ add(add.sd) var ~ add(add.sd) | cmt"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"transformations","dir":"Reference","previous_headings":"","what":"Transformations","title":"Create an ODE-based model specification — rxode2","text":"normal related distributions, can apply transformation sides using keywords/functions apply transformations. default likelihood transformations calculated untransformed scale. bounded variables like logit-normal probit-normal low high values defaulted 0 1 missing. models wish proportional model one transformation can replace standard deviation NA allow transformations, lnorm(), probitNorm() logitNorm() can combined variance stabilizing yeoJohnson() transformation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"normal-and-t-related-distributions","dir":"Reference","previous_headings":"","what":"Normal and t-related distributions","title":"Create an ODE-based model specification — rxode2","text":"normal t-related distributions, wanted keep ability use skewed distributions additive proportional t/cauchy-space, distributions specified differently comparison supported distributions within nlmixr2: Note normal t-related distributions nlmixr2 calculate cwres npde normal assumption help assess goodness fit model. Also note +dnorm() mostly testing purposes slow estimation procedure nlmixr2. suggest adding (except explicit testing). multiple endpoint models mix non-normal normal distributions, whole problem shifted log-likelihood method estimation nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"notes-on-additive-proportional-models","dir":"Reference","previous_headings":"","what":"Notes on additive + proportional models","title":"Create an ODE-based model specification — rxode2","text":"two different ways specify additive proportional models, call combined1 combined2, way Monolix calls two distributions (avoid software differences naming). first, combined1, assumes additive proportional differences standard deviation scale, : y=f+(+b* f^c)*err second, combined2, assumes additive proportional differences combined variance scale: y=f+[sqrt(^2+b^2 *f^(2c))]*err default nlmixr2/rxode2 otherwise specified combined2 since mirrors adding 2 normal distributions statistics add variances (standard deviations). However, combined1 can describe data possibly even better combined2 possible options rxode2/nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"distributions-of-known-likelihoods","dir":"Reference","previous_headings":"","what":"Distributions of known likelihoods","title":"Create an ODE-based model specification — rxode2","text":"residuals related normal, t-distribution cauchy, often residual specification form:   compartment specification left handed side specification. generalized likelihood can specify:","code":"cmt ~ dbeta(alpha, beta) ll(cmt) ~ llik specification"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"ordinal-likelihoods","dir":"Reference","previous_headings":"","what":"Ordinal likelihoods","title":"Create an ODE-based model specification — rxode2","text":"Finally, ordinal likelihoods/simulations can specified 2 ways. first :   err represents compartment p0 probability specific category: model ensure sum p values less 1. Additionally can write arbitrary number categories ordinal model described . seems little p0 probability category 1 sometimes scores non-whole numbers. can modeled follows:   numeric categories specified explicitly, probabilities remain :","code":"err ~ c(p0, p1, p2) err ~ c(p0=0, p1=1, p2=2, 3)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"general-table-of-supported-residual-distributions","dir":"Reference","previous_headings":"","what":"General table of supported residual distributions","title":"Create an ODE-based model specification — rxode2","text":"general supported following table (available rxode2::rxResidualError)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"creating-rxode-models","dir":"Reference","previous_headings":"","what":"Creating rxode2 models","title":"Create an ODE-based model specification — rxode2","text":"NA","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Create an ODE-based model specification — rxode2","text":"Chamber, J. M. Temple Lang, D. (2001) Object Oriented Programming R. R News, Vol. 1, . 3, September 2001. https://cran.r-project.org/doc/Rnews/Rnews_2001-3.pdf. Hindmarsh, . C. ODEPACK, Systematized Collection ODE Solvers. Scientific Computing, R. S. Stepleman et al. (Eds.), North-Holland, Amsterdam, 1983, pp. 55-64. Petzold, L. R. Automatic Selection Methods Solving Stiff Nonstiff Systems Ordinary Differential Equations. Siam J. Sci. Stat. Comput. 4 (1983), pp. 136-148. Hairer, E., Norsett, S. P., Wanner, G. Solving ordinary differential equations , nonstiff problems. 2nd edition, Springer Series Computational Mathematics, Springer-Verlag (1993). Plevyak, J. dparser, https://dparser.sourceforge.net/. Web. 12 Oct. 2015.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Create an ODE-based model specification — rxode2","text":"Melissa Hallow, Wenping Wang Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Create an ODE-based model specification — rxode2","text":"","code":"# \\donttest{  mod <- function() {   ini({     KA   <- .291     CL   <- 18.6     V2   <- 40.2     Q    <- 10.5     V3   <- 297.0     Kin  <- 1.0     Kout <- 1.0     EC50 <- 200.0   })   model({     # A 4-compartment model, 3 PK and a PD (effect) compartment     # (notice state variable names 'depot', 'centr', 'peri', 'eff')     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot;     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  <-                    Q*C2 - Q*C3;     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff;     eff(0)      <- 1   }) }  m1 <- rxode2(mod) #>   #>   print(m1) #>  ── rxode2-based free-form 4-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.291  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2            centr #> 3                  3             peri #> 4                  4              eff #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         KA <- 0.291 #>         CL <- 18.6 #>         V2 <- 40.2 #>         Q <- 10.5 #>         V3 <- 297 #>         Kin <- 1 #>         Kout <- 1 #>         EC50 <- 200 #>     }) #>     model({ #>         C2 <- centr/V2 #>         C3 <- peri/V3 #>         d/dt(depot) <- -KA * depot #>         d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3 #>         d/dt(peri) <- Q * C2 - Q * C3 #>         d/dt(eff) <- Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>         eff(0) <- 1 #>     }) #> }  # Step 2 - Create the model input as an EventTable, # including dosing and observation (sampling) events  # QD (once daily) dosing for 5 days.  qd <- et(amountUnits = \"ug\", timeUnits = \"hours\") %>%   et(amt = 10000, addl = 4, ii = 24)  # Sample the system hourly during the first day, every 8 hours # then after qd <- qd %>% et(0:24) %>%   et(from = 24 + 8, to = 5 * 24, by = 8)  # Step 3 - solve the system  qd.cp <- rxSolve(m1, qd) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  head(qd.cp) #>    time       C2        C3     depot    centr      peri      eff #> 1 0 [h]  0.00000 0.0000000 10000.000    0.000    0.0000 1.000000 #> 2 1 [h] 43.99334 0.9113641  7475.157 1768.532  270.6751 1.083968 #> 3 2 [h] 54.50866 2.6510696  5587.797 2191.248  787.3677 1.179529 #> 4 3 [h] 51.65163 4.4243597  4176.966 2076.396 1314.0348 1.227523 #> 5 4 [h] 44.37513 5.9432612  3122.347 1783.880 1765.1486 1.233503 #> 6 5 [h] 36.46382 7.1389804  2334.004 1465.845 2120.2772 1.214084  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate ordinal value — rxord","title":"Simulate ordinal value — rxord","text":"Simulate ordinal value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate ordinal value — rxord","text":"","code":"rxord(...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate ordinal value — rxord","text":"... probabilities simulated.  sum number one.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate ordinal value — rxord","text":"number 1 (number probabilities specified + 1)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate ordinal value — rxord","text":"values entered 'rxord' simulation simulate probability falling group. falls outside specified probabilities, simulate group (number probabilities specified + 1)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate ordinal value — rxord","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate ordinal value — rxord","text":"","code":"# This will give values 1, and 2 rxord(0.5) #> [1] 2 rxord(0.5) #> [1] 2 rxord(0.5) #> [1] 2 rxord(0.5) #> [1] 2  # This will give values 1, 2 and 3 rxord(0.3, 0.3) #> [1] 3 rxord(0.3, 0.3) #> [1] 3 rxord(0.3, 0.3) #> [1] 1"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate random Poisson variable from threefry generator — rxpois","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"","code":"rxpois(lambda, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"lambda vector (non-negative) means. n number random values return. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"poission random number deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"","code":"# \\donttest{ ## Use threefry engine  rxpois(lambda = 3, n = 10) # with rxpois you have to explicitly state n #>  [1] 3 2 3 3 4 4 2 1 5 3 rxpois(lambda = 3, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 5 5 4 5 3 4 2 4 4 4  rxpois(4) ## The first arguments are the lambda parameter #> [1] 5   ## This example uses `rxpois` directly in the model  rx <- function() {   model({     a <- rxpois(3)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate student t variable from threefry generator — rxt","title":"Simulate student t variable from threefry generator — rxt","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate student t variable from threefry generator — rxt","text":"","code":"rxt(df, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate student t variable from threefry generator — rxt","text":"df degrees freedom (\\(> 0\\), maybe non-integer).  df       = Inf allowed. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate student t variable from threefry generator — rxt","text":"t-distribution random numbers","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate student t variable from threefry generator — rxt","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate student t variable from threefry generator — rxt","text":"","code":"# \\donttest{  ## Use threefry engine rxt(df = 3, n = 10) # with rxt you have to explicitly state n #>  [1]  0.56020706  0.03692413 -0.14483694 -1.17839233 -0.07091994  0.59518023 #>  [7] -0.86772970 -0.13114717 -0.82988663 -0.44191039 rxt(df = 3, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  0.03932487 -0.52306475  0.43368480 -0.34707505  0.95056248 -0.59519017 #>  [7]  0.88227899  2.54392958 -0.24328189 -1.64972949  rxt(4) ## The first argument is the df parameter #> [1] 0.2307802   ## This example uses `rxt` directly in the model  rx <- function() {    model({     a <- rxt(3)    }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate uniform variable from threefry generator — rxunif","title":"Simulate uniform variable from threefry generator — rxunif","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate uniform variable from threefry generator — rxunif","text":"","code":"rxunif(min = 0, max = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate uniform variable from threefry generator — rxunif","text":"min, max lower upper limits distribution.  Must finite. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate uniform variable from threefry generator — rxunif","text":"uniform random numbers","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate uniform variable from threefry generator — rxunif","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate uniform variable from threefry generator — rxunif","text":"","code":"# \\donttest{  ## Use threefry engine  rxunif(min = 0, max = 4, n = 10) # with rxunif you have to explicitly state n #>  [1] 2.67123413 1.53680622 0.03693407 3.10817092 1.82090596 0.87096816 #>  [7] 3.00018567 0.74465379 0.44367887 1.04829595 rxunif(min = 0, max = 4, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 1.3636154 2.2496600 2.1135472 3.6623704 2.0958273 3.2123092 1.1121945 #>  [8] 3.2064501 0.7037566 3.7979522  rxunif() #> [1] 0.5753584   ## This example uses `rxunif` directly in the model  rx <- function() {   model({     a <- rxunif(0, 3)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Weibull variable from threefry generator — rxweibull","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"","code":"rxweibull(shape, scale = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"shape, scale shape scale parameters, latter defaulting 1. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"Weibull random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"","code":"# \\donttest{  ## Use threefry engine  # with rxweibull you have to explicitly state n rxweibull(shape = 1, scale = 4, n = 10) #>  [1] 6.44811941 9.31743179 3.71729577 3.93298872 1.93424653 0.50995473 #>  [7] 0.08795812 3.51517749 3.87991413 1.65760268  # You can parallelize the simulation using openMP rxweibull(shape = 1, scale = 4, n = 10, ncores = 2) #>  [1]  8.43975112  0.60066178  3.19996975  8.95933943  0.90401027  1.12725152 #>  [7]  0.02614905  2.35284437  2.74451590 14.14117829  rxweibull(3) #> [1] 1.146769   ## This example uses `rxweibull` directly in the model  rx <- function() {   model({     a <- rxweibull(1, 3)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":null,"dir":"Reference","previous_headings":"","what":"Dosing/Amt geom/stat — stat_amt","title":"Dosing/Amt geom/stat — stat_amt","text":"dosing geom shows vertical lines dose occurs","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Dosing/Amt geom/stat — stat_amt","text":"","code":"stat_amt(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   ... )  geom_amt(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Dosing/Amt geom/stat — stat_amt","text":"mapping Set aesthetic mappings created aes(). specified inherit.aes = TRUE (default), combined default mapping top level plot. must supply mapping plot mapping. data data displayed layer. three options: NULL, default, data inherited plot data specified call ggplot(). data.frame, object, override plot data. objects fortified produce data frame. See fortify() variables created. function called single argument, plot data. return value must data.frame, used layer data. function can created formula (e.g. ~ head(.x, 10)). position Position adjustment, either string naming adjustment (e.g. \"jitter\" use position_jitter), result call position adjustment function. Use latter need change settings adjustment. show.legend logical. layer included legends? NA, default, includes aesthetics mapped. FALSE never includes, TRUE always includes. can also named logical vector finely select aesthetics display. inherit.aes FALSE, overrides default aesthetics, rather combining . useful helper functions define data aesthetics inherit behaviour default plot specification, e.g. borders(). ... arguments passed layer(). often aesthetics, used set aesthetic fixed value, like colour = \"red\" size = 3. may also parameters paired geom/stat.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Dosing/Amt geom/stat — stat_amt","text":"returns stat_amt context ggplot2 plot","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Dosing/Amt geom/stat — stat_amt","text":"Requires following aesthetics: x representing x values, usually time amt representing dosing values;  missing zero dose given","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Dosing/Amt geom/stat — stat_amt","text":"","code":"# \\donttest{ library(rxode2) library(units) #> udunits database from /usr/share/xml/udunits/udunits2.xml   ## Model from RxODE tutorial mod1 <- function() {   ini({     KA <- 2.94E-01     CL <- 1.86E+01     V2 <- 4.02E+01     Q  <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200   })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <-  KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff   }) }  ## These are making the more complex regimens of the rxode2 tutorial  ## bid for 5 days bid <- et(timeUnits=\"hr\") %>%   et(amt=10000,ii=12,until=set_units(5, \"days\"))  ## qd for 5 days qd <- et(timeUnits=\"hr\") %>%       et(amt=20000,ii=24,until=set_units(5, \"days\"))  ## bid for 5 days followed by qd for 5 days  et <- seq(bid,qd) %>% et(seq(0,11*24,length.out=100))  bidQd <- rxSolve(mod1, et, addDosing=TRUE) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # by default dotted and under-stated plot(bidQd, C2) + geom_amt(aes(amt=amt))   # of course you can make it a bit more visible  plot(bidQd, C2) + geom_amt(aes(amt=amt), col=\"red\", lty=1, linewidth=1.2)   # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":null,"dir":"Reference","previous_headings":"","what":"Censoring geom/stat — stat_cens","title":"Censoring geom/stat — stat_cens","text":"censoring geom shows left right censoring specified nlmixr input data-set fit","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Censoring geom/stat — stat_cens","text":"","code":"stat_cens(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   width = 0.01,   ... )  geom_cens(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   width = 0.01,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Censoring geom/stat — stat_cens","text":"mapping Set aesthetic mappings created aes(). specified inherit.aes = TRUE (default), combined default mapping top level plot. must supply mapping plot mapping. data data displayed layer. three options: NULL, default, data inherited plot data specified call ggplot(). data.frame, object, override plot data. objects fortified produce data frame. See fortify() variables created. function called single argument, plot data. return value must data.frame, used layer data. function can created formula (e.g. ~ head(.x, 10)). position Position adjustment, either string naming adjustment (e.g. \"jitter\" use position_jitter), result call position adjustment function. Use latter need change settings adjustment. show.legend logical. layer included legends? NA, default, includes aesthetics mapped. FALSE never includes, TRUE always includes. can also named logical vector finely select aesthetics display. inherit.aes FALSE, overrides default aesthetics, rather combining . useful helper functions define data aesthetics inherit behaviour default plot specification, e.g. borders(). width represents width (\\ censoring box ... arguments passed layer(). often aesthetics, used set aesthetic fixed value, like colour = \"red\" size = 3. may also parameters paired geom/stat.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Censoring geom/stat — stat_cens","text":"returns ggplot2 stat","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Censoring geom/stat — stat_cens","text":"Requires following aesthetics: x Represents independent variable, often time scale y represents dependent variable CENS censoring information; (-1 right censored, 0 censoring 1 left censoring) LIMIT represents corresponding limit () add boxes representing areas fit censored.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":null,"dir":"Reference","previous_headings":"","what":"Summary of rxDll object — summary.rxDll","title":"Summary of rxDll object — summary.rxDll","text":"gives expanded information rxDll object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Summary of rxDll object — summary.rxDll","text":"","code":"# S3 method for rxDll summary(object, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Summary of rxDll object — summary.rxDll","text":"object RxDll object ... arguments.  Includes noprint, logical telling object print rxDll object first. default FALSE","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Summary of rxDll object — summary.rxDll","text":"object returned","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Summary of rxDll object — summary.rxDll","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Print expanded information about the rxode2 object. — summary.rxode2","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"prints expanded information rxode2 object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"","code":"# S3 method for rxode2 summary(object, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"object rxode2 object ... Ignored parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"object returned","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":null,"dir":"Reference","previous_headings":"","what":"Update for rxUi — update.rxUi","title":"Update for rxUi — update.rxUi","text":"Update rxUi","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update for rxUi — update.rxUi","text":"","code":"# S3 method for rxUi update(object, ..., envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update for rxUi — update.rxUi","text":"object rxode2 UI object ... Lines update envir Environment evaluating ini() style calls","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update for rxUi — update.rxUi","text":"new rxode2 updated UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":null,"dir":"Reference","previous_headings":"","what":"uppergamma: upper incomplete gamma function — uppergamma","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"tgamma boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"","code":"uppergamma(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"numeric '' parameter upper incomplete gamma z numeric 'z' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"uppergamma results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"uppergamma function given : \\(uppergamma(, z) = \\int_{z}^{\\infty}t^{-1}\\cdot e^{-t} dt\\)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"","code":"uppergamma(1, 3) #> [1] 0.04978707  uppergamma(1:3, 3) #> [1] 0.04978707 0.19914827 0.84638016  uppergamma(1, 1:3) #> [1] 0.36787944 0.13533528 0.04978707"},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":null,"dir":"Reference","previous_headings":"","what":"Set random effects and residual error to zero — zeroRe","title":"Set random effects and residual error to zero — zeroRe","text":"Set random effects residual error zero","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set random effects and residual error to zero — zeroRe","text":"","code":"zeroRe(object, which = c(\"omega\", \"sigma\"), fix = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set random effects and residual error to zero — zeroRe","text":"object model modify types parameters set zero fix parameters fixed zero value?","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set random effects and residual error to zero — zeroRe","text":"object parameters set zero","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set random effects and residual error to zero — zeroRe","text":"Bill Denney","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Set random effects and residual error to zero — zeroRe","text":"","code":"one.compartment <- function() {   ini({     tka <- log(1.57); label(\"Ka\")     tcl <- log(2.72); label(\"Cl\")     tv <- log(31.5); label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) } zeroRe(one.compartment) #>   #>   #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.4510756 1.0006319 3.4499875 0.0000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka      0      0     0 #> eta.cl      0      0     0 #> eta.v       0      0     0 #> attr(,\"lotriFix\") #>        eta.ka eta.cl eta.v #> eta.ka   TRUE  FALSE FALSE #> eta.cl  FALSE   TRUE FALSE #> eta.v   FALSE  FALSE  TRUE #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.451075619360217 #>         label(\"Ka\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- fix(0, 0) #>         eta.ka ~ fix(0) #>         eta.cl ~ fix(0) #>         eta.v ~ fix(0) #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-1-3","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.1.3","text":"Make sure object uncompressed rxode2 ui solving rxSolve (See #661) Fix #670 using last simulated observation residual trailing doses.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-1-3","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.1.3","text":"Create function see rxode2 solve loaded memory (rxode2::rxSolveSetup()) Create new function fixes rxode2 population values model (drops initial estimates); rxFixPop()","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"other-changes-2-1-3","dir":"Changelog","previous_headings":"","what":"Other changes","title":"rxode2 2.1.3","text":"Pendantic -remap (requested CRAN)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-212","dir":"Changelog","previous_headings":"","what":"rxode2 2.1.2","title":"rxode2 2.1.2","text":"CRAN release: 2024-01-30","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"other-changes-2-1-2","dir":"Changelog","previous_headings":"","what":"Other changes","title":"rxode2 2.1.2","text":"rxUi compression now defaults fast compression Fixes String literal formatting issues identified CRAN (#643) Removes linear compartment solutions gradients intel c++ compiler (since crash compiler).","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-210","dir":"Changelog","previous_headings":"","what":"rxode2 2.1.0","title":"rxode2 2.1.0","text":"CRAN release: 2023-12-11","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"breaking-changes-2-1-0","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"rxode2 2.1.0","text":"Steady state lag times longer shifted lag time solved steady state default. addition steady state original time dosing also back-calculated. want old behavior can bring back option ssAtDoseTime=FALSE. “dop853” now uses hmax/h0 values rxControl() rxSolve(). may change ODE solving using “dop853” specified (xgxr available), x axis longer assumed hours","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-1-0","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.1.0","text":"User defined functions can now R functions. many R functions can converted C rxFun() (can see C code afterwards rxC(\"funName\")) Parallel solving models require sorting (like modeled lag times, modeled duration etc) now solve parallel instead downgrading single threaded solving Steady state infusions duration infusions greater inter-dose interval now supported. Added $symengineModelNoPrune $symengineModelPrune loading models rxode2 rxS() plotting creating confidence intervals multiple endpoint models simulated rxode2 ui model, can plot/summarize endpoint sim. (ie. confint(model,   \"sim\") plot(model, sim)). want summarize subset endpoints, can focus endpoint pre-pending endpoint sim. example wanted plot/summarize endpoint eff use sim.eff. (ie confint(model, \"sim.eff\") plot(model, sim.eff)) Added model$simulationIniModel prepend initial conditions ini({}) block classic rxode2({}) model. Now model$simulationModel model$simulationIniModel save use initialization values compiled model, solve original ui model. Allow ini(model) <- NULL drop ini block .ini(NULL) gives ini({}) (Issue #523) Add function modelExtract() extract model lines allow modifying changing model piping simply assigning modified lines model(ui) <- newModifiedLines Add Algebraic mu-referencing detection (mu2) allows express mu-referenced covariates : Instead previously required (log.WT.div.70.5 calculated data) mu expressions. ui now information allow transformation data internally transformation old mu-referencing style run optimization. Allow steady state infusions duration infusion greater inter-dose interval solved. Solves now possibly print information issuing “solve system” error function rxSetPipingAuto() now exported change way affect piping individual setup Allow covariates specified model piping, mod   %>% model(=var+3, cov=\"var\") add \"var\" covariate. calculating confidence intervals rxode2 simulated objects can now use stratify simulation summary. example can now stratify gender race : confint(sim,   \"sim\", =c(\"race\", \"gender\")) calculating intervals rxode2 simulated objects can now use ci=FALSE calculates default intervals without bands percentiles; can also choose match secondary bands limits levels use ci=0.99 instance new function introduced meanProbs() calculates mean expected confidence bands either normal t distribution related new function introduced calculates mean confidence bands Bernoulli/Binomial distribution (binomProbs()) calculating intervals rxode2 simulated objects can also use mean=TRUE use mean first level confidence using meanProbs(). confidence interval can override n used confidence interval using n=#. can also change prediction interval instead using pred=TRUE. Also calculating intervals rxode2 simulated object can also use mean=\"binom\" use binomial distributional information (ci) first level confidence using binomProbs(). confidence interval can override n used confidence interval using n=#. can also change prediction interval instead using pred=TRUE. pred=TRUE can override number predicted samples m=# plotting confint derived intervals rxode2 simulation, can now subset based simulated value like plot(ci, Cc) plot variable Cc summarized even also summarized eff (instance). rxode2 ui compressed ui object, can modify ini block $ini <- modify model block $model <-. equivalent ini(model) <- model(model) <-, respectively. Otherwise, object added user defined components function (ie $meta). object uncompressed, simply assigns environment instead (just like ). printing meta information happens lotri compatible matrix, use lotri express instead default R expression. Allow character vectors converted expressions piping (#552) rxAppendModel() now take arbitrary number models append together; also better handling models duplicate parameters models without ini() blocks (#617 / #573 / #575). keep now also keep attributes input data (special handling levels); means broader variety classes kept carrying information (example ordered factors, data frame columns unit information, etc) Piping arguments append ini() model() aligned perform similarly. Therefore ini(append=) now can take expressions instead simply strings model(append=) can also take strings. Also model piping now can specify integer line number modified just like ini() . Also model(append=FALSE) changed model(append=NULL). behavior don’t specify argument, behavior changed align ini() piping. Hence model(append=TRUE) append model(append=FALSE) now pre-pend model. model(append=NULL) modify lines like behavior ini(append=NULL). default model(line) modifying line -place still applies. breaking change, code perform . Labels can now dropped ini(param=label(NULL)). Also parameters can dropped idiom model(param=NULL) ini(param=NULL) changes parameter covariate align idiom dropping parameters rxRename refactored run faster","code":"cl <- exp(tcl + eta.cl + wt_cl * log(WT/70.5)) cl <- exp(tcl + eta.cl + wt_cl * log.WT.div.70.5)"},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"internal-new-features-2-1-0","dir":"Changelog","previous_headings":"","what":"Internal new features","title":"rxode2 2.1.0","text":"Add .model() list expressions, implies model(ui) <-   ui$lstExpr assign model components. also robustly work character vectors Simulated objects rxSolve now can access model variables $rxModelVars Simulation models UI now use rxerr.endpoint instead err.endpoint sigma residual error. align convention internally generated variables start rx nlmixr Sorting uses timsort now, upgraded latest version Morwenn","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-1-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.1.0","text":"Simulating/solving functions/ui now prefers params omega sigma model (#632) Piping add constants initial estimates constants specified model({}) block (like k <- 1), ini block Bug fix geom_amt() aes transformation x Bug fix covariate updates may affect multiple compartment models (like issue #581)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"maintenance-fixes-2-1-0","dir":"Changelog","previous_headings":"","what":"Maintenance fixes","title":"rxode2 2.1.0","text":"Modify plot code work development xgxr","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2014","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.14","title":"rxode2 2.0.14","text":"CRAN release: 2023-10-07 CRAN requested FORTRAN kind changed portable; commented code, simply removed comment. Bug-fix geom_amt(); also now uses linewidth least ggplot2 3.4.0 documentation cleaned rxode2 2.0.13","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2013","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.13","title":"rxode2 2.0.13","text":"CRAN release: 2023-04-22","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-0-13","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.0.13","text":"bug fixed zeroRe() function works correlated omega values. bug fixed rename() function works initial conditions compartments (cmt(0))","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-0-13","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.0.13","text":"new function zeroRe() allows simple setting omega /sigma values zero model (#456) Diagonal zeros omega sigma matrices treated zeros model. corresponding omega sigma matrices drop columns/rows diagonals zero create new omega sigma matrix simulation. idiom NONMEM uses simulation matrices. Add ability pipe model estimates another model parentModel %>% ini(modelWithNewEsts) Add ability append model statements piping using %>%   model(x=3, append=d/dt(depot)), still supports appending append=TRUE pre-pending append=NA (default replace lines append=FALSE) rxSolve’s keep argument now maintain character factor classes input data class (#190) Parameter labels may now modified via ini(param = label(\"text\")) (#351). Parameter order may modified via append argument ini() piping model. example, ini(param = 1, append = 0) ini(param = label(\"text\"), append = \"param2\") (#352).","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"internal-changes-2-0-13","dir":"Changelog","previous_headings":"","what":"Internal changes","title":"rxode2 2.0.13","text":"lower/upper bounds outside required bounds, adjustment displayed. initial values piped break model’s boundary condition reset boundary unbounded message boundary reset. Added .rxUi() function convert following objects rxUi objects: rxode2, rxModelVars, function. Converting nlmixr2 fits rxUi placed s3 method corresponding package. assertRxUi(x) now uses .rxUi() can extended outside rxode2/nlmixr2. rxode2 now supports addl ss doses Moved rxDerived rxode2parse (re-exported ). Added test transit compartment solving absence dosing transit compartment (fixed rxode2parse solving tested ) Using ini() without arguments rxode2 type function return ini() block. Also added method ini(mod) <-   iniBlock modify ini block wish. iniBlock expression. Using model() without arguments rxode2 type function return model() block. Also added new method model(mod) <- modelBlock Added new method rxode2(mod) <- modFunction allows replacing function new function maintaining meta information ui (like information comes nonmem2rx models). modFunction body new function, new function, new rxode2 ui. rxode2 ui objects now $sticky item inside internal (compressed) environment. $sticky tells variables keep “significant” change ui piping sort model change. respected model piping, modifying model ini(mod)<-, model(mod)<-, rxode2(mod)<-. significant change change model block, change number estimates, change value estimates. Estimate bounds, weather estimate fixed estimate label changes considered significant. Added .ini() method convert various formats ini expression. used internally ini(mod)<-. want assign something new can convert ini expression, add method .ini(). Added .model() method convert various formats model expression. used internally model(mod)<-. want assign something new can convert model expression, add method .model().","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2011","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.11","title":"rxode2 2.0.11","text":"CRAN release: 2022-11-01 Give meaningful error ‘rxode2’ ui models error expressions Break ABI requirement roxde2() rxode2parse() new rxode2parse fix sprintf exclusion shown CRAN.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2010","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.10","title":"rxode2 2.0.10","text":"CRAN release: 2022-10-20 Time invariant covariates can now contain ‘NA’ values. column ‘NA’ entire id, now ‘rxode2’ warns id column instead just id. fix CRAN issues ‘nlmixr2est’, make version dependency explicit.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-209","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.9","title":"rxode2 2.0.9","text":"CRAN release: 2022-10-19 Remove log likelihoods ‘rxode2’ reduce compilation time increase maintainability ‘rxode2’. transferred ‘rxode2ll’ (requested CRAN). Remove parsing ‘rxode2’ solved linear compartment code move ‘rxode2parse’ reduce compilation time (requested CRAN). Remove random number generation ‘rxode2’ move ‘rxode2random’ reduce compilation time (requested CRAN). Remove event table translation generation ‘rxode2’ move ‘rxode2et’ reduce compilation time (requested CRAN). Change rxode2 ui object compressed, serialized object default. reduce C stack size problem occurs many environments R. Warn ignoring items simulations Export method change ‘rxode2’ solve methods internal integers Bug fix time invariant covariates identified time variant covariate individual’s time starts 0.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-208","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.8","title":"rxode2 2.0.8","text":"CRAN release: 2022-09-23","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"breaking-changes-2-0-8","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"rxode2 2.0.8","text":"rxgamma now allows rate input. aligns internal rxode2 version rxgamma clarifies used. also aligned llikGamma function used generalized likelihood estimation. ui cauchy simulations now follow ui normal t distributions, means can combine transformations. cauchy t distribution one degree freedom. ui dnorm() norm() longer equivalent add(). Now allows use loglik llikNorm() instead standard nlmixr2 style focei likelihood. done adding dnorm() end line. also means dnorm() now doesn’t take arguments. Vandercorput normal removed (non-random number generator)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-0-8","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.0.8","text":"Allow models nlmixr2 form without ini({}) block Allow model piping omega matrix f %>% ini(omegaMatrix) Standard models created rxode2() can piped model function Families log-likelihood added rxode2 mixed likelihood nonlinear mixed effects models may specified run. memory footprint rxode2 solving reduced Piping now allow named strings (issue #249)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-0-8","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.0.8","text":"rxode2’s symengine convert sqrt(2) M_SQRT_2 M_SQRT2. fixed; noticeable nlmixr2 log-likelihood estimation methods rxode2 treats DV non-covariate etTran (last time duplicate model). noticeable nlmixr2 log-likelihood estimation methods.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-0-8-1","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.0.8","text":"new flag (rxFlag) created tell rxode2 solving process . useful debugging. outputting variable always 11 calculating left handed equations. using conjunction printf() methods, double variable formatted \"%f\". additional option fullPrint added rxode2() allows rprintf() used almost rxode2() steps (inductive linearization matrix exponential exception ) instead just integration ddt step. defaults FALSE.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-207","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.7","title":"rxode2 2.0.7","text":"CRAN release: 2022-05-17 Removed accidental ^S news requested CRAN. Bug fix complicated mu-referencing. Change rxode2 md5 depend C/C++/Fortran code headers R files. way binary compatibility nlmixr2est rxode2, new version nlmixr2est need submitted CRAN.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-206","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.6","title":"rxode2 2.0.6","text":"CRAN release: 2022-05-09","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"solving-controls-2-0-6","dir":"Changelog","previous_headings":"Breaking changes","what":"Solving controls","title":"rxode2 2.0.6","text":"options rxControl rxSolve strict. camelCase now always used. Old options like add.cov transit_abs longer supported, addCov supported. new option, sigdig added rxControl(), controls common significant figure options like atol, rtol, ssAtol, ssRtol, single option.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"simulations-2-0-6","dir":"Changelog","previous_headings":"Breaking changes","what":"Simulations","title":"rxode2 2.0.6","text":"simulations, $simulationSigma now assumes diagonal matrix. sigma values assumed standard normal, uncorrelated endpoints. Simulation uncertainty still draw identity diagonal matrix Parallel solving now seeds simulation per individual based initial seed plus simulation id. makes simulation reproducible regardless number cores running simulation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"other-breaking-changes-2-0-6","dir":"Changelog","previous_headings":"Breaking changes","what":"Other breaking changes","title":"rxode2 2.0.6","text":"Solved objects now access underlying rxode model $rxode2 instead $rxode Since change names, rxode2, rxode RxODE perform function. Options changed RxODE.syntax rxode2.syntax. Assigning states rxode2.syntax.assign.state (RxODE.syntax.assign.state) longer supported. Enforcing “pure” assignment syntax = syntax longer supported rxode2.syntax.assign longer supported (RxODE.syntax.assign). Since R supports ** exponentiation operator, pure syntax without ** can longer enabled. Hence rxode2.syntax.star.pow (RxODE.syntax.star.pow) longer effect. “pure” syntax requires semicolon can longer enabled. Therefore rxode2.syntax.require.semicolon (RxODE.syntax.require.semicolon) longer effect. syntax state(0) can longer turned . rxode2.syntax.allow.ini0 (RxODE.syntax.allow.ini0) removed. Variable dots variable state names like state.name works R. Therefore, “pure” syntax excluding . values variables enforced rxode2.syntax.allow.dots (RxODE.syntax.allow.dots). mnemonic et(rate=model) et(dur=model) mnemonics removed. rate needs set -1 -2 manually instead. function rxode2Test() removed favor using testthat directly. Transit compartments need use new evid, evid=7. said, transitAbs option longer supported. ID columns input parameter data frames sorted merged original dataset ; underlying assumption ID order now checked outside rxode2(). Note event data frame still sorted.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"additional-features-2-0-6","dir":"Changelog","previous_headings":"","what":"Additional features","title":"rxode2 2.0.6","text":"UI functions nlmixr ported work rxode2 directly. rxModelVars({}) now supported. may now combine 2 models rxode2 rxAppendModel(). fact, long first value rxode2 evaluated ui model, can use c/rbind bind 2 models together. may now append model lines piping using %>% model(lines,   append=TRUE) can also pre-pend lines %>% model(lines,   append=NA) may now rename model variables, states defined parameters %>% rxRename(new=old) dplyr loaded: %>%   rename(new=old) can fix parameters %>% ini(tcl=fix) %>% ini(fix(tcl)) well unfix parameters %>% ini(tcl=unfix) %>% ini(unfix(tcl))","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"internal-changes-2-0-6","dir":"Changelog","previous_headings":"","what":"Internal changes","title":"rxode2 2.0.6","text":"Strict R headers enforced places Since many changes incompatible, version renamed rxode2 rxode2() printout longer uses rules centered headings make display better larger variety systems.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-0-6","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.0.6","text":"tad() related time features reset start infusion (opposed starting beginning end infusion)","code":""}]
+[{"path":"https://nlmixr2.github.io/rxode2/articles/Integrating-User-Defined-Functions-into-rxode2.html","id":"user-defined-functions","dir":"Articles","previous_headings":"","what":"User Defined Functions","title":"Integrating User Defined Functions into rxode2","text":"defining models may wished write small R function make function integrate rxode2 somehow. article discusses 2 ways : R-based user function can loaded simple function certain circumstances translated C run efficiently C function define integrate code","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()`"},{"path":"https://nlmixr2.github.io/rxode2/articles/Integrating-User-Defined-Functions-into-rxode2.html","id":"r-based-user-functions","dir":"Articles","previous_headings":"","what":"R based user functions","title":"Integrating User Defined Functions into rxode2","text":"R-based user function convenient include ODE, slower done written C , C++ compiled language. requested github appropriate example; However, use simple example simply illustrate concepts. Now ODE compiled R functions called solving ODE. Since calling R, forces parallization turned since R single-threaded. also takes time solve since shuttling back forth R C. Lets see simple function performs:  terribly bad, even though shuffling R C. can make better converting functions C:  C version almost twice fast R version. may noticed conversion also created C versions first derivative. done automatically gives just C versions function, C versions derivatives registers rxode2. allows C versions work rxode2 nlmixr2 models. function setup advance allow type conversion. general derivatives calculated return() statement user defined function. means simply let R return last value instead explictly calling return(). Many people prefer method coding. Even return function, function converted C. github issue, used function convert derivatives: still helpful functions early returns, nlmixr2 models requiring derivatives calculated non-optimized finite differences occurs. gets internals rxode2 nlmixr2 can see easily calculating derivatives: Whereas originally defined function newAbs() use new derivatives calculated well: circumstances, conversion C possible, though can still use R function. requirements R functions integrated rxode2 system: function must set number arguments, variable arguments like f(…) currently allowed. function given argument single number, function return single number requirements met can use R function rxode2. Additional requirements conversion C include: functions use within R function must understood available rxode2. Practically speaking fun2() refers fun1(), fun1() must changed C code available rxode2 changing function fun2() C. functions can include /else assignments simple return statements (either returning value value line ). Special R control structures functions (like lapply) present. function refer package functions mentioned, return() statement present, derivative C functions rxode2’s derivative table updated.","code":"newAbs <- function(x) {   if (x < 0) {     -x   } else {     x   } }  f <- rxode2({   a <- newAbs(time) }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ e <- et(-10, 10, length.out=40) mb1 <- microbenchmark::microbenchmark(withoutC=suppressWarnings(rxSolve(f,e)))  library(ggplot2) autoplot(mb1) + rxTheme() # Create C functions automatically with `rxFun()` rxFun(newAbs) #> → finding duplicate expressions in d(newAbs)/d(x)... #> [====|====|====|====|====|====|====|====|====|====] 0:00:00 #> → optimizing duplicate expressions in d(newAbs)/d(x)... #> [====|====|====|====|====|====|====|====|====|====] 0:00:00 #> converted R function 'newAbs' to C (will now use in rxode2) #> converted R function 'rx_newAbs_d_x' to C (will now use in rxode2) #> Added derivative table for 'newAbs' # Recompile to use the C functions # Note it would recompile anyway if you didn't do this step, # it just makes sure that it doesn't recompile every step in # the benchmark f <- rxode2({   a <- newAbs(time) }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ mb2 <- microbenchmark::microbenchmark(withC=rxSolve(f,e, cores=1))  mb <- rbind(mb1, mb2) autoplot(mb) + rxTheme() + xgxr::xgx_scale_y_log10() #> Scale for y is already present. #> Adding another scale for y, which will replace the existing scale. print(mb) #> Unit: milliseconds #>      expr      min       lq     mean   median       uq       max neval #>  withoutC 6.867416 6.970227 7.706955 7.084244 7.519517 12.342958   100 #>     withC 1.961846 2.025228 2.333243 2.122846 2.614116  6.562157   100 # Light f_R <- function(actRad, k_0, a_k) {   photfac <- a_k * actRad + k_0   if (photfac > 1) {     photfac = 1   }   return(photfac) }  rxFun(f_R) #> function contains return statement; derivatives not calculated #> converted R function 'f_R' to C (will now use in rxode2) rxFromSE(\"Derivative(f_R(actRad, k_0, a_k),k_0)\") #> [1] \"(f_R(actRad,(k_0)+6.05545445239334e-06,a_k)-f_R(actRad,k_0,a_k))/6.05545445239334e-06\" rxFromSE(\"Derivative(newAbs(x),x)\") #> [1] \"rx_newAbs_d_x(x)\""},{"path":"https://nlmixr2.github.io/rxode2/articles/Integrating-User-Defined-Functions-into-rxode2.html","id":"c-based-functions","dir":"Articles","previous_headings":"","what":"C based functions","title":"Integrating User Defined Functions into rxode2","text":"can add C functions directly rxode2 well using rxFun(): wanted also use C functions expressions derivatives using rxD() function: Removing function rxRmFun() also remove derivative table:","code":"fun <- \"  double fun(double a, double b, double c) {    return a*a+b*a+c;  } \" ## C-code for function  rxFun(\"fun\", c(\"a\", \"b\", \"c\"), fun) rxD(\"fun\", list(    function(a, b, c) { # derivative of arg1: a      paste0(\"2*\", a, \"+\", b)    },    function(a, b, c) { # derivative of arg2: b      return(a)    },    function(a, b, c) { # derivative of arg3: c      return(\"0.0\")    } )) rxRmFun(\"fun\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"modifying-model","dir":"Articles","previous_headings":"","what":"Modifying model","title":"Modifying Models","text":"tasks might want overall model: Change line model Add line model Rename parameters model Combine different models Create functions add certain model features model go model piping functions can use modify models even add functions modify models. cover model modification functions nlmixr2lib","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"modifying-a-model-line","dir":"Articles","previous_headings":"Modifying model","what":"Modifying a model line","title":"Modifying Models","text":"opinion, modifying lines model likely common task modifying model. may wish modify model subject variability add covariate effects. begin course need base model modify. Let’s start simple PK example, using single-dose theophylline dataset generously provided Dr. Robert . Upton University California, San Francisco: believed enough absorption support subject variability can change line drop subject modifying single line. simply type line want model piping expression: expected, line modified. Also can notice initial estimate subject variability dropped since longer part model. reason wanted add back model can modify model add back: modification, eta.ka automatically assumed subject variability parameter. Also since eta.ka mu-referenced rxode2 points . automatic detection eta.ka name follows convention. Parameters starting ending following names assumed subject variability parameters: eta (NONMEM convention) ppv (per patient variability) psv (per subject variability) iiv (inter-individual variability) bsv (subject variability) bpv (patient variability) functioning correctly can change covariate can add type initial estimate later: seen , eta.ka model assumed data-input parameter covariate instead estimated parameter. can also note WT automatically recognized covariate covWt automatically recognized covariate parameter. general covariates typical/population parameters automatically converted estimated parameters based parameter name starting (ending ): tv (typical value) t (also typical value) pop (population parameter) err (error parameter) eff (effect parameter) cov (covariate parameters) notable exceptions parameters like (wt, sex crcl) assumed covariates. don’t want automatic variable conversion, can also use auto=FALSE: case additional parameters (eta.ka, WT, covWt) assumed parameters dataset.","code":"one.compartment <- function() {   ini({     tka <- 0.45; label(\"Ka\")     tcl <- 1; label(\"Cl\")     tv <- 3.45; label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) } mod <- one.compartment |>   model(ka <- exp(tka)) #> ! remove between subject variability `eta.ka` print(mod) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tcl eta.cl    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } mod2 <- mod |>   model(ka <- tka * exp(eta.ka)) #> ℹ add between subject variability `eta.ka` and set estimate to 1 print(mod2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.ka #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.ka    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tcl eta.cl    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.ka ~ 1 #>     }) #>     model({ #>         ka <- tka * exp(eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } mod2 <- mod |>   model(ka <- tka * exp(eta.ka) + WT * covWt, cov=\"eta.ka\") #> ℹ add covariate `eta.ka` (as requested by cov option) #> ℹ add covariate `WT` #> ℹ add population parameter `covWt` and set estimate to 1 print(mod2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  covWt  #>   0.45   1.00   3.45   0.70   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tcl eta.cl    id            #> 2    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         covWt <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- tka * exp(eta.ka) + WT * covWt #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } mod3 <- mod |>   model(ka <- tka * exp(eta.ka) + WT * covWt, auto=FALSE) #> ℹ add covariate `eta.ka` #> ℹ add covariate `WT` #> ℹ add covariate `covWt` print(mod3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tcl eta.cl    id            #> 2    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- tka * exp(eta.ka) + WT * covWt #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"note-on-automatic-detection-of-variables","dir":"Articles","previous_headings":"Modifying model","what":"Note on automatic detection of variables","title":"Modifying Models","text":"automatic detection variables convenient many models may suit style; like can always change using options(): option disabled, variables assumed covariates promote population parameters ini block last example option enabled none variables starting t added model course use turn back : can use withr::with_options(list(rxode2.autoVarPiping=FALSE), ...) turn option temporarily. don’t like defaults changing variables change well rxSetPipingAuto() example wanted variables starting ending te can change : requested population parameters starting te added ini block. want reset defaults simply call rxSetPipingAuto() without arguments:","code":"options(rxode2.autoVarPiping=FALSE) mod7 <- mod3 |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE)  print(mod7) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tcl eta.cl    id            #> 2    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- tka * exp(eta.ka) + WT * covWt #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } options(rxode2.autoVarPiping=TRUE) mod8 <- mod |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 #> ℹ promote `tkin` to population parameter with initial estimate 1 #> ℹ promote `tkout` to population parameter with initial estimate 1 print(mod8) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  temax    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } rxSetPipingAuto(thetamodelVars = rex::rex(\"te\"))  mod9 <- mod |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 print(mod9) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  temax    te0  tec50  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } rxSetPipingAuto() mod10 <- mod |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 #> ℹ promote `tkin` to population parameter with initial estimate 1 #> ℹ promote `tkout` to population parameter with initial estimate 1"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"adding-model-lines","dir":"Articles","previous_headings":"Modifying model","what":"Adding model lines","title":"Modifying Models","text":"three ways insert lines rxode2/nlmixr2 model. can add lines end model, expression beginning model controlled append option. Let’s assume two different assays run compound noticed different variability. can modify model adding lines end model using append=TRUE: Perhaps instead may want add indirect response model addition concentrations, can choose add : append=lhsVar lhsVar left handed variable want insert new lines: last type insertion may wish add lines beginning model using append=FALSE:","code":"mod4 <- mod |>   model({     cp2 <- cp     cp2 ~ lnorm(lnorm.sd)   }, append=TRUE) #> ℹ add residual parameter `lnorm.sd` and set estimate to 1 print(mod4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>      tka      tcl       tv   add.sd lnorm.sd  #>     0.45     1.00     3.45     0.70     1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v #> eta.cl    0.3   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                cmt                dvid* #> 1   cp ~ …  cmt='cp' or cmt=3  dvid='cp' or dvid=1 #> 2  cp2 ~ … cmt='cp2' or cmt=4 dvid='cp2' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tcl eta.cl    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         lnorm.sd <- c(0, 1) #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>         cp2 <- cp #>         cp2 ~ lnorm(lnorm.sd) #>     }) #> } mod5 <- mod |>   model({     PD <- 1-emax*cp/(ec50+cp)     ##     effect(0) <- e0     kin <- e0*kout     d/dt(effect) <- kin*PD -kout*effect   }, append=d/dt(center)) mod6 <- mod5 |>   model({     emax <- exp(temax)     e0 <- exp(te0 + eta.e0)     ec50 <- exp(tec50)     kin <- exp(tkin)     kout <- exp(tkout)   }, append=FALSE) #> ℹ promote `temax` to population parameter with initial estimate 1 #> ℹ promote `te0` to population parameter with initial estimate 1 #> ℹ promote `eta.e0` to between subject variability with initial estimate 1 #> ℹ promote `tec50` to population parameter with initial estimate 1 #> ℹ promote `tkin` to population parameter with initial estimate 1 #> ℹ promote `tkout` to population parameter with initial estimate 1 print(mod6) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  temax    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3           effect #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         PD <- 1 - emax * cp/(ec50 + cp) #>         effect(0) <- e0 #>         kin <- e0 * kout #>         d/dt(effect) <- kin * PD - kout * effect #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"remove-lines-in-the-model","dir":"Articles","previous_headings":"Modifying model","what":"Remove lines in the model","title":"Modifying Models","text":"lines model can removed one 2 ways either use -param param <- NULL model piping:","code":"mod7 <- mod6 |>   model(-emax) #> ! remove population parameter `temax` print(mod7) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3           effect #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         PD <- 1 - emax * cp/(ec50 + cp) #>         effect(0) <- e0 #>         kin <- e0 * kout #>         d/dt(effect) <- kin * PD - kout * effect #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } # Equivalently  mod8 <- mod6 |>   model(emax <- NULL) #> ! remove population parameter `temax` print(mod8) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd    te0  tec50   tkin  tkout  #>   0.45   1.00   3.45   0.70   1.00   1.00   1.00   1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3           effect #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tkout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         PD <- 1 - emax * cp/(ec50 + cp) #>         effect(0) <- e0 #>         kin <- e0 * kout #>         d/dt(effect) <- kin * PD - kout * effect #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"rename-parameters-in-a-model","dir":"Articles","previous_headings":"Modifying model","what":"Rename parameters in a model","title":"Modifying Models","text":"may want rename parameters model, easy rxRename(). dplyr loaded can even replace rename(). semantics similar two functions, assigning newVar=oldVar. example: can see every instance variable named model renamed inside model ini block. completeness can see dplyr verb (since S3 method):","code":"mod11 <- mod10 |>   rxRename(drug1kout=kout, tv.drug1kout=tkout)  print(mod11) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>          tka          tcl           tv       add.sd        temax          te0  #>         0.45         1.00         3.45         0.70         1.00         1.00  #>        tec50         tkin tv.drug1kout  #>         1.00         1.00         1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tv.drug1kout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         drug1kout <- exp(tv.drug1kout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union mod12 <- mod10 |>   rename(drug1kout=kout, tv.drug1kout=tkout)  print(mod12) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>          tka          tcl           tv       add.sd        temax          te0  #>         0.45         1.00         3.45         0.70         1.00         1.00  #>        tec50         tkin tv.drug1kout  #>         1.00         1.00         1.00  #>  #> Omega ($omega):  #>        eta.cl eta.v eta.e0 #> eta.cl    0.3   0.0      0 #> eta.v     0.0   0.1      0 #> eta.e0    0.0   0.0      1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   te0 eta.e0    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Ka\") #>         tcl <- 1 #>         label(\"Cl\") #>         tv <- 3.45 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         temax <- 1 #>         te0 <- 1 #>         tec50 <- 1 #>         tkin <- 1 #>         tv.drug1kout <- 1 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>         eta.e0 ~ 1 #>     }) #>     model({ #>         emax <- exp(temax) #>         e0 <- exp(te0 + eta.e0) #>         ec50 <- exp(tec50) #>         kin <- exp(tkin) #>         drug1kout <- exp(tv.drug1kout) #>         ka <- exp(tka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"combine-different-models","dir":"Articles","previous_headings":"Modifying model","what":"Combine different models","title":"Modifying Models","text":"can also combine different models rxAppendModel(). general need variables common combine. generally want models link . example pk pd model done renaming cp first model ceff second model: get error try combine models without variables common: want combine models without respecting variables common, can use common=FALSE:","code":"ocmt <- function() {   ini({     tka <- exp(0.45) # Ka     tcl <- exp(1) # Cl     tv <- exp(3.45); # log V     ## the label(\"Label name\") works with all models     add.sd <- 0.7   })   model({     ka <- tka     cl <- tcl     v <- tv     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl / v * center     cp <- center / v     cp ~ add(add.sd)   }) }  idr <- function() {   ini({     tkin <- log(1)     tkout <- log(1)     tic50 <- log(10)     gamma <- fix(1)     idr.sd <- 1   })   model({     kin <- exp(tkin)     kout <- exp(tkout)     ic50 <- exp(tic50)     d/dt(eff) <- kin - kout*(1-ceff^gamma/(ic50^gamma+ceff^gamma))     eff ~ add(idr.sd)   }) }  rxAppendModel(ocmt %>% rxRename(ceff=cp), idr) #> ℹ parameter labels from comments will be replaced by 'label()' #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd      tkin     tkout     tic50     gamma  #>  1.568312  2.718282 31.500392  0.700000  0.000000  0.000000  2.302585  1.000000  #>    idr.sd  #>  1.000000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3              eff #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                 cmt                 dvid* #> 1 ceff ~ … cmt='ceff' or cmt=4 dvid='ceff' or dvid=1 #> 2  eff ~ …  cmt='eff' or cmt=3  dvid='eff' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>         tkin <- 0 #>         tkout <- 0 #>         tic50 <- 2.30258509299405 #>         gamma <- fix(1) #>         idr.sd <- c(0, 1) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         ceff <- center/v #>         ceff ~ add(add.sd) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ic50 <- exp(tic50) #>         d/dt(eff) <- kin - kout * (1 - ceff^gamma/(ic50^gamma +  #>             ceff^gamma)) #>         eff ~ add(idr.sd) #>     }) #> } try(rxAppendModel(ocmt, idr)) #> ℹ parameter labels from comments will be replaced by 'label()' #> Error : not all the models have variables in common (use `common=FALSE` to allow this) mod2 <- rxAppendModel(ocmt, idr, common=FALSE) |>   model(ceff=cp, append=ic50) # here we add the translation after the #> ℹ parameter labels from comments will be replaced by 'label()' # ic50 line to make it reasonable  print(mod2) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd      tkin     tkout     tic50     gamma  #>  1.568312  2.718282 31.500392  0.700000  0.000000  0.000000  2.302585  1.000000  #>    idr.sd  #>  1.000000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3              eff #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                cmt                dvid* #> 1   cp ~ …  cmt='cp' or cmt=4  dvid='cp' or dvid=1 #> 2  eff ~ … cmt='eff' or cmt=3 dvid='eff' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>         tkin <- 0 #>         tkout <- 0 #>         tic50 <- 2.30258509299405 #>         gamma <- fix(1) #>         idr.sd <- c(0, 1) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ic50 <- exp(tic50) #>         ceff <- cp #>         d/dt(eff) <- kin - kout * (1 - ceff^gamma/(ic50^gamma +  #>             ceff^gamma)) #>         eff ~ add(idr.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"creating-more-complex-model-modification-functions","dir":"Articles","previous_headings":"Modifying model","what":"Creating more complex model modification functions","title":"Modifying Models","text":"pretty flexible, may want even , helper functions help create functions . discuss extract model function update . Lets start model: Lets assume moment want remove eta cl. First probably want get model lines. can modelExtract(): Now want worry cl line, can subset : Now wish change line “cl <- exp(tcl+eta.cl)” Note functions modify ini({}) block. may modify ini block first make valid rxode2/nlmixr2 model. particular case, using model piping easier, simply demonstrates two different way extract model information way add information final model. methods can tricky using model parsed correctly. means make sure parameters endpoints follow correct rules","code":"f <- function() {   ini({     tka <- 0.45     tcl <- 1     tv <- 3.45     eta.ka ~ 0.6     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl)     v <- exp(tv + eta.v)     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl/v * center     cp <- center/v   }) } totLines <- modelExtract(f, endpoint=NA) # endpoints should be included  print(totLines) #> [1] \"ka <- exp(tka + eta.ka)\"                    #> [2] \"cl <- exp(tcl)\"                             #> [3] \"v <- exp(tv + eta.v)\"                       #> [4] \"d/dt(depot) <- -ka * depot\"                 #> [5] \"d/dt(center) <- ka * depot - cl/v * center\" #> [6] \"cp <- center/v\" clLine <- modelExtract(f, cl, lines=TRUE) line <- attr(clLine, \"lines\") totLines[line] <- \"cl <- exp(tcl+eta.cl)\"  # For now lets remove the entire `ini` block (so you don't have to # worry about syncing parameters).  #  ini(f) <- NULL  model(f) <- totLines #> ℹ parameter labels from comments are typically ignored in non-interactive mode #> ℹ Need to run with the source intact to parse comments print(f) #> function ()  #> { #>     ini({ #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>     }) #> } #> <environment: 0x55ff06d78570>"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"modifying-initial-estimates","dir":"Articles","previous_headings":"","what":"Modifying initial estimates","title":"Modifying Models","text":"common items want initial estimates : Fix/Unfix parameter Change initial condition values bounds Change initial condition type Change labels transformations Reorder parameters may wish create functions; discuss .","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"fixing-or-unfixing-a-parameter","dir":"Articles","previous_headings":"Modifying initial estimates","what":"Fixing or unfixing a parameter","title":"Modifying Models","text":"can fix model estimates two ways. first fix value whatever model function, done piping model parameter name (like tka) setting equal fix (%>% ini(tka=fix)). full example: can also fix parameter different value wish; similar can specify value fix inside fix pseudo-function follows: %>% ini(tka=fix(0.1)). fully worked example :","code":"f <- function() {   ini({     tka <- 0.45     tcl <- 1     tv <- 3.45     add.sd <- c(0, 0.7)     eta.ka ~ 0.6     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl)     v <- exp(tv + eta.v)     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl/v * center     cp <- center/v     cp ~ add(add.sd)   }) }  f2 <- f |>   ini(tka=fix) #> ℹ fix `tka` to `0.45` print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.45) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f <- function() {   ini({     tka <- 0.45     tcl <- 1     tv <- 3.45     add.sd <- c(0, 0.7)     eta.ka ~ 0.6     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl)     v <- exp(tv + eta.v)     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl/v * center     cp <- center/v     cp ~ add(add.sd)   }) }  f2 <- f |>   ini(tka=fix(0.1)) #> ℹ fix `tka` to `0.45` #> ℹ change initial estimate of `tka` to `0.1` print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"unfixing-parameters","dir":"Articles","previous_headings":"Modifying initial estimates","what":"Unfixing parameters","title":"Modifying Models","text":"unfix parameters similarly fixing. Instead using fix keyword, use unfix keyword. unfix parameter (keeping value) pipe model using (|> ini(tka=unfix)). Starting fixed model fully worked example : can also unfix change initial estimate ini(parameter=unfix(newEst)):","code":"print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f2 |> ini(tka=unfix) #> ℹ unfix `tka` keeping initial estimate `0.1` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } print(f2) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- fix(0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f2 |>   ini(tka=unfix(10)) #> ℹ unfix `tka` keeping initial estimate `0.1` #> ℹ change initial estimate of `tka` to `10` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>  10.00   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 10 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"multiple-parameter-assignment","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Multiple parameter assignment","title":"Modifying Models","text":"can also assign multiple parameters providing : vector/list multiple lines piped ini() block Using covariance matrix case vector can specify pipe model. example: can also added multiple lines commas separating estimates: also use matrix specify covariance:","code":"ini1 <- c(tka=0.1, tcl=1, tv=3)  f4 <- f |> ini(ini1) #> ℹ change initial estimate of `tka` to `0.1` #> ℹ change initial estimate of `tcl` to `1` #> ℹ change initial estimate of `tv` to `3` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.1    1.0    3.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # or equivalently  ini1 <- list(tka=0.1, tcl=1, tv=3)  f4a <- f |> ini(ini1) #> ℹ change initial estimate of `tka` to `0.1` #> ℹ change initial estimate of `tcl` to `1` #> ℹ change initial estimate of `tv` to `3` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.1    1.0    3.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # commas separating values: f4 <- f |> ini(tka=0.1, tcl=1, tv=3) #> ℹ change initial estimate of `tka` to `0.1` #> ℹ change initial estimate of `tcl` to `1` #> ℹ change initial estimate of `tv` to `3` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.1    1.0    3.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # multiple lines in {} f4 <- f |>   ini({     tka <- 0.2     tcl <- 2     tv <- 6   }) #> ℹ change initial estimate of `tka` to `0.2` #> ℹ change initial estimate of `tcl` to `2` #> ℹ change initial estimate of `tv` to `6` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>    0.2    2.0    6.0    0.7  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.2 #>         tcl <- 2 #>         tv <- 6 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } ome <- lotri(eta.ka + eta.v ~ c(0.6,                                 0.01, 10.1))  f4 <- f |> ini(ome) #> ℹ change initial estimate of `eta.ka` to `0.6` #> ℹ add covariance between `eta.v` and `eta.ka` with initial estimate `0.01` #> ℹ change initial estimate of `eta.v` to `10.1` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka   0.60  0.01 #> eta.v    0.01 10.10 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka + eta.v ~ c(0.6, 0.01, 10.1) #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # or equavialtly use the lotri-type syntax for the omega:  f4 <- f |> ini(eta.ka + eta.v ~ c(0.6,                                   0.01, 0.2)) #> ℹ change initial estimate of `eta.ka` to `0.6` #> ℹ add covariance between `eta.v` and `eta.ka` with initial estimate `0.01` #> ℹ change initial estimate of `eta.v` to `0.2` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka   0.60  0.01 #> eta.v    0.01  0.20 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka + eta.v ~ c(0.6, 0.01, 0.2) #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"single-parameter-assignment","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Single parameter assignment","title":"Modifying Models","text":"simplest way change initial parameter estimates simply use ini(parameter=newValue). can also use <- ~ change value: fully worked example showing three types initial value modification : can change bounds like model specification using numeric vector c(low, estimate) c(low, estimate, hi). worked example: Note changing parameters default values might show parameter printout:","code":"f3 <- f |>   ini(tka <- 0.1) #> ℹ change initial estimate of `tka` to `0.1` f4 <- f |>   ini(tka=0.1) #> ℹ change initial estimate of `tka` to `0.1` f5 <- f |>   ini(tka ~ 0.1) #> ℹ change initial estimate of `tka` to `0.1` print(f5) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f |>   ini(tka <- c(0, 0.1, 0.2)) #> ℹ change initial estimate (0.1) and upper/lower bound (0 to 0.2) of `tka` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- c(0, 0.1, 0.2) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f |>   ini(tka <- c(0, 0.1)) #> ℹ change initial estimate (0.1) and lower bound (0) of `tka` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- c(0, 0.1) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f3 <- f |>   ini(tka <- c(0, 0.1, 0.2)) #> ℹ change initial estimate (0.1) and upper/lower bound (0 to 0.2) of `tka` print(f3) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- c(0, 0.1, 0.2) #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # Now reassign f4 <- f3 |>   ini(tka <- c(-Inf, 0.1, Inf)) #> ℹ change initial estimate (0.1) and upper/lower bound (-Inf to Inf) of `tka` print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.10   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.1 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"changing-parameter-types","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Changing parameter types","title":"Modifying Models","text":"can change parameter type two operators either using -par convert parameter covariate ~par toggle population individual parameters. example 3:","code":"# Switch population parameter to between subject variability parameter: f4 <- f |>   ini( ~ tcl) #> ℹ convert 'tcl' from population parameter to between subject variability print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v tcl #> eta.ka    0.6   0.0   0 #> eta.v     0.0   0.1   0 #> tcl       0.0   0.0   1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>         tcl ~ 1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # Switch back to population parameter f5 <- f4 |>   ini( ~ tcl) #> ℹ convert 'tcl' from between subject variability to population parameter print(f5) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd    tcl  #>   0.45   3.45   0.70   1.00  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         tcl <- 1 #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # Change the variable to a covariate parameter (ie it doesn't have an # initial estimate so remove it with the `-` operator):  f6 <- f4 |>   ini(-tcl) #> ℹ changing between subject variability parameter 'tcl' to covariate parameter print(f6) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # You can change the covariate or remove the parameter estimate by # `tcl <- NULL`:  f6 <- f4 |>   ini(tcl <- NULL) #> ℹ changing between subject variability parameter 'tcl' to covariate parameter print(f6) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } # to add it back as a between subject variability or population # parameter you can pipe it as follows:  f7 <- f6 |>   ini(tcl=4) #> ℹ promote `tcl` to population parameter with initial estimate 4 #> ℹ change initial estimate of `tcl` to `4` print(f7) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd    tcl  #>   0.45   3.45   0.70   4.00  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         tcl <- 4 #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f8 <- f6 |>   ini(tcl ~ 0.1) #> ℹ promote `tcl` to between subject variability with initial estimate 0.1 #> ℹ change initial estimate of `tcl` to `0.1` print(f8) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka     tv add.sd  #>   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v tcl #> eta.ka    0.6   0.0 0.0 #> eta.v     0.0   0.1 0.0 #> tcl       0.0   0.0 0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>         tcl ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"changing-parameter-labels","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Changing parameter labels","title":"Modifying Models","text":"want change/add parameter label assign parameter label(\"label add\"). example: can also change order performing operations: want remove labels can remove ini(par=label(NULL)); example:","code":"f4 <- f |>   ini(tka=label(\"Typical Ka (1/hr)\"))  print(f4) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Typical Ka (1/hr)\") #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f5 <- f |>   ini(tka=label(\"Typical Ka (1/hr)\"), append=tcl)  print(f5) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tcl    tka     tv add.sd  #>   1.00   0.45   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tcl <- 1 #>         tka <- 0.45 #>         label(\"Typical Ka (1/hr)\") #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f6 <- f |>   ini(tka=label(NULL))  print(f6) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"changing-parameter-transformations","dir":"Articles","previous_headings":"Modifying initial estimates > Changing the parameter values and possibly bounds","what":"Changing parameter transformations","title":"Modifying Models","text":"Back-transformations -ride back transformations nlmixr2 models. similar modification labels. use |> ini(tka=backTransform(exp)) add exponential back-transformation data: wish remove can also |> ini(tka=backTransform(NULL)):","code":"f7 <- f |>   ini(tka=backTransform(exp))  print(f7) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         backTransform(\"exp\") #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> } f8 <- f |>   ini(tka=backTransform(NULL))  print(f8) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   0.45   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/Modifying-Models.html","id":"more-granular-access-of-initial-conditions","dir":"Articles","previous_headings":"Modifying initial estimates","what":"More granular access of initial conditions","title":"Modifying Models","text":"Just like model() can modify underlying data frame represents ini() block. case simply change initial estimate first parameter (tka):","code":"f <- rxode2(f)  ini <- f$iniDf  print(ini) #>   ntheta neta1 neta2   name lower  est upper   fix label backTransform #> 1      1    NA    NA    tka  -Inf 0.45   Inf FALSE  <NA>          <NA> #> 2      2    NA    NA    tcl  -Inf 1.00   Inf FALSE  <NA>          <NA> #> 3      3    NA    NA     tv  -Inf 3.45   Inf FALSE  <NA>          <NA> #> 4      4    NA    NA add.sd     0 0.70   Inf FALSE  <NA>          <NA> #> 5     NA     1     1 eta.ka  -Inf 0.60   Inf FALSE  <NA>          <NA> #> 6     NA     2     2  eta.v  -Inf 0.10   Inf FALSE  <NA>          <NA> #>   condition  err #> 1      <NA> <NA> #> 2      <NA> <NA> #> 3      <NA> <NA> #> 4        cp  add #> 5        id <NA> #> 6        id <NA> ini$est[1] <- 7  ini(f) <- ini  print(f) #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>    tka    tcl     tv add.sd  #>   7.00   1.00   3.45   0.70  #>  #> Omega ($omega):  #>        eta.ka eta.v #> eta.ka    0.6   0.0 #> eta.v     0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 7 #>         tcl <- 1 #>         tv <- 3.45 #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-cmt.html","id":"how-rxode2-assigns-compartment-numbers","dir":"Articles","previous_headings":"","what":"How rxode2 assigns compartment numbers","title":"Changing rxode2 compartment numbers","text":"rxode2 automatically assigns compartment numbers parsing. example, Mavoglurant PBPK model following model may used: look printout, can see rxode2 assigned compartment number(s) can also see classic rxode2 model. case use summary() function: case, Venous_Blood assigned compartment 15. Figuring can inconvenient also lead re-numbering compartment simulation estimation datasets. easy probably clearer specify compartment name, tools support compartment numbers. Therefore, way number compartment easily can lead less data modification multiple tools.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` pbpk <- function() {   model({     KbBR = exp(lKbBR)     KbMU = exp(lKbMU)     KbAD = exp(lKbAD)     CLint= exp(lCLint + eta.LClint)     KbBO = exp(lKbBO)     KbRB = exp(lKbRB)      ## Regional blood flows     # Cardiac output (L/h) from White et al (1968)     CO  = (187.00*WT^0.81)*60/1000      QHT = 4.0 *CO/100     QBR = 12.0*CO/100     QMU = 17.0*CO/100     QAD = 5.0 *CO/100     QSK = 5.0 *CO/100     QSP = 3.0 *CO/100     QPA = 1.0 *CO/100     QLI = 25.5*CO/100     QST = 1.0 *CO/100     QGU = 14.0*CO/100     # Hepatic artery blood flow     QHA = QLI - (QSP + QPA + QST + QGU)      QBO = 5.0 *CO/100     QKI = 19.0*CO/100     QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI)     QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI + QRB      ## Organs' volumes = organs' weights / organs' density     VLU = (0.76 *WT/100)/1.051     VHT = (0.47 *WT/100)/1.030     VBR = (2.00 *WT/100)/1.036     VMU = (40.00*WT/100)/1.041     VAD = (21.42*WT/100)/0.916     VSK = (3.71 *WT/100)/1.116     VSP = (0.26 *WT/100)/1.054     VPA = (0.14 *WT/100)/1.045     VLI = (2.57 *WT/100)/1.040     VST = (0.21 *WT/100)/1.050     VGU = (1.44 *WT/100)/1.043     VBO = (14.29*WT/100)/1.990     VKI = (0.44 *WT/100)/1.050     VAB = (2.81 *WT/100)/1.040     VVB = (5.62 *WT/100)/1.040     VRB = (3.86 *WT/100)/1.040      ## Fixed parameters     BP = 0.61      # Blood:plasma partition coefficient     fup = 0.028    # Fraction unbound in plasma     fub = fup/BP   # Fraction unbound in blood      KbLU = exp(0.8334)     KbHT = exp(1.1205)     KbSK = exp(-.5238)     KbSP = exp(0.3224)     KbPA = exp(0.3224)     KbLI = exp(1.7604)     KbST = exp(0.3224)     KbGU = exp(1.2026)     KbKI = exp(1.3171)      ##-----------------------------------------     S15 = VVB*BP/1000     C15 = Venous_Blood/S15      ##-----------------------------------------     d/dt(Lungs) = QLU*(Venous_Blood/VVB - Lungs/KbLU/VLU)     d/dt(Heart) = QHT*(Arterial_Blood/VAB - Heart/KbHT/VHT)     d/dt(Brain) = QBR*(Arterial_Blood/VAB - Brain/KbBR/VBR)     d/dt(Muscles) = QMU*(Arterial_Blood/VAB - Muscles/KbMU/VMU)     d/dt(Adipose) = QAD*(Arterial_Blood/VAB - Adipose/KbAD/VAD)     d/dt(Skin) = QSK*(Arterial_Blood/VAB - Skin/KbSK/VSK)     d/dt(Spleen) = QSP*(Arterial_Blood/VAB - Spleen/KbSP/VSP)     d/dt(Pancreas) = QPA*(Arterial_Blood/VAB - Pancreas/KbPA/VPA)     d/dt(Liver) = QHA*Arterial_Blood/VAB + QSP*Spleen/KbSP/VSP +       QPA*Pancreas/KbPA/VPA + QST*Stomach/KbST/VST +       QGU*Gut/KbGU/VGU - CLint*fub*Liver/KbLI/VLI - QLI*Liver/KbLI/VLI     d/dt(Stomach) = QST*(Arterial_Blood/VAB - Stomach/KbST/VST)     d/dt(Gut) = QGU*(Arterial_Blood/VAB - Gut/KbGU/VGU)     d/dt(Bones) = QBO*(Arterial_Blood/VAB - Bones/KbBO/VBO)     d/dt(Kidneys) = QKI*(Arterial_Blood/VAB - Kidneys/KbKI/VKI)     d/dt(Arterial_Blood) = QLU*(Lungs/KbLU/VLU - Arterial_Blood/VAB)     d/dt(Venous_Blood) = QHT*Heart/KbHT/VHT + QBR*Brain/KbBR/VBR +       QMU*Muscles/KbMU/VMU + QAD*Adipose/KbAD/VAD + QSK*Skin/KbSK/VSK +       QLI*Liver/KbLI/VLI + QBO*Bones/KbBO/VBO + QKI*Kidneys/KbKI/VKI +       QRB*Rest_of_Body/KbRB/VRB - QLU*Venous_Blood/VVB     d/dt(Rest_of_Body) = QRB*(Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB)   }) } pbpk <- pbpk() print(pbpk) #>  -- rxode2-based free-form 16-cmt ODE model -------------------------------------  #>  #> States ($state or $stateDf):  #>    Compartment Number Compartment Name #> 1                   1            Lungs #> 2                   2            Heart #> 3                   3            Brain #> 4                   4          Muscles #> 5                   5          Adipose #> 6                   6             Skin #> 7                   7           Spleen #> 8                   8         Pancreas #> 9                   9            Liver #> 10                 10          Stomach #> 11                 11              Gut #> 12                 12            Bones #> 13                 13          Kidneys #> 14                 14   Arterial_Blood #> 15                 15     Venous_Blood #> 16                 16     Rest_of_Body #>  -- Model (Normalized Syntax): --  #> function() { #>     model({ #>         KbBR = exp(lKbBR) #>         KbMU = exp(lKbMU) #>         KbAD = exp(lKbAD) #>         CLint = exp(lCLint + eta.LClint) #>         KbBO = exp(lKbBO) #>         KbRB = exp(lKbRB) #>         CO = (187 * WT^0.81) * 60/1000 #>         QHT = 4 * CO/100 #>         QBR = 12 * CO/100 #>         QMU = 17 * CO/100 #>         QAD = 5 * CO/100 #>         QSK = 5 * CO/100 #>         QSP = 3 * CO/100 #>         QPA = 1 * CO/100 #>         QLI = 25.5 * CO/100 #>         QST = 1 * CO/100 #>         QGU = 14 * CO/100 #>         QHA = QLI - (QSP + QPA + QST + QGU) #>         QBO = 5 * CO/100 #>         QKI = 19 * CO/100 #>         QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO +  #>             QKI) #>         QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI +  #>             QRB #>         VLU = (0.76 * WT/100)/1.051 #>         VHT = (0.47 * WT/100)/1.03 #>         VBR = (2 * WT/100)/1.036 #>         VMU = (40 * WT/100)/1.041 #>         VAD = (21.42 * WT/100)/0.916 #>         VSK = (3.71 * WT/100)/1.116 #>         VSP = (0.26 * WT/100)/1.054 #>         VPA = (0.14 * WT/100)/1.045 #>         VLI = (2.57 * WT/100)/1.04 #>         VST = (0.21 * WT/100)/1.05 #>         VGU = (1.44 * WT/100)/1.043 #>         VBO = (14.29 * WT/100)/1.99 #>         VKI = (0.44 * WT/100)/1.05 #>         VAB = (2.81 * WT/100)/1.04 #>         VVB = (5.62 * WT/100)/1.04 #>         VRB = (3.86 * WT/100)/1.04 #>         BP = 0.61 #>         fup = 0.028 #>         fub = fup/BP #>         KbLU = exp(0.8334) #>         KbHT = exp(1.1205) #>         KbSK = exp(-0.5238) #>         KbSP = exp(0.3224) #>         KbPA = exp(0.3224) #>         KbLI = exp(1.7604) #>         KbST = exp(0.3224) #>         KbGU = exp(1.2026) #>         KbKI = exp(1.3171) #>         S15 = VVB * BP/1000 #>         C15 = Venous_Blood/S15 #>         d/dt(Lungs) = QLU * (Venous_Blood/VVB - Lungs/KbLU/VLU) #>         d/dt(Heart) = QHT * (Arterial_Blood/VAB - Heart/KbHT/VHT) #>         d/dt(Brain) = QBR * (Arterial_Blood/VAB - Brain/KbBR/VBR) #>         d/dt(Muscles) = QMU * (Arterial_Blood/VAB - Muscles/KbMU/VMU) #>         d/dt(Adipose) = QAD * (Arterial_Blood/VAB - Adipose/KbAD/VAD) #>         d/dt(Skin) = QSK * (Arterial_Blood/VAB - Skin/KbSK/VSK) #>         d/dt(Spleen) = QSP * (Arterial_Blood/VAB - Spleen/KbSP/VSP) #>         d/dt(Pancreas) = QPA * (Arterial_Blood/VAB - Pancreas/KbPA/VPA) #>         d/dt(Liver) = QHA * Arterial_Blood/VAB + QSP * Spleen/KbSP/VSP +  #>             QPA * Pancreas/KbPA/VPA + QST * Stomach/KbST/VST +  #>             QGU * Gut/KbGU/VGU - CLint * fub * Liver/KbLI/VLI -  #>             QLI * Liver/KbLI/VLI #>         d/dt(Stomach) = QST * (Arterial_Blood/VAB - Stomach/KbST/VST) #>         d/dt(Gut) = QGU * (Arterial_Blood/VAB - Gut/KbGU/VGU) #>         d/dt(Bones) = QBO * (Arterial_Blood/VAB - Bones/KbBO/VBO) #>         d/dt(Kidneys) = QKI * (Arterial_Blood/VAB - Kidneys/KbKI/VKI) #>         d/dt(Arterial_Blood) = QLU * (Lungs/KbLU/VLU - Arterial_Blood/VAB) #>         d/dt(Venous_Blood) = QHT * Heart/KbHT/VHT + QBR * Brain/KbBR/VBR +  #>             QMU * Muscles/KbMU/VMU + QAD * Adipose/KbAD/VAD +  #>             QSK * Skin/KbSK/VSK + QLI * Liver/KbLI/VLI + QBO *  #>             Bones/KbBO/VBO + QKI * Kidneys/KbKI/VKI + QRB * Rest_of_Body/KbRB/VRB -  #>             QLU * Venous_Blood/VVB #>         d/dt(Rest_of_Body) = QRB * (Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB) #>     }) #> } pbpk <- pbpk$simulationModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ summary(pbpk) #> rxode2 2.1.3 model named rx_0a5b43d0b1e7bff7f47de827bf0daaa5 model (ready).  #> DLL: /tmp/RtmpPXkf2x/rxode2/rx_0a5b43d0b1e7bff7f47de827bf0daaa5__.rxd/rx_0a5b43d0b1e7bff7f47de827bf0daaa5_.so #> NULL #>  #> Calculated Variables: #>  [1] \"KbBR\"  \"KbMU\"  \"KbAD\"  \"CLint\" \"KbBO\"  \"KbRB\"  \"CO\"    \"QHT\"   \"QBR\"   #> [10] \"QMU\"   \"QAD\"   \"QSK\"   \"QSP\"   \"QPA\"   \"QLI\"   \"QST\"   \"QGU\"   \"QHA\"   #> [19] \"QBO\"   \"QKI\"   \"QRB\"   \"QLU\"   \"VLU\"   \"VHT\"   \"VBR\"   \"VMU\"   \"VAD\"   #> [28] \"VSK\"   \"VSP\"   \"VPA\"   \"VLI\"   \"VST\"   \"VGU\"   \"VBO\"   \"VKI\"   \"VAB\"   #> [37] \"VVB\"   \"VRB\"   \"fub\"   \"KbLU\"  \"KbHT\"  \"KbSK\"  \"KbSP\"  \"KbPA\"  \"KbLI\"  #> [46] \"KbST\"  \"KbGU\"  \"KbKI\"  \"S15\"   \"C15\"   #> -- rxode2 Model Syntax -- #> rxode2({ #>     param(lKbBR, lKbMU, lKbAD, lCLint, eta.LClint, lKbBO, lKbRB,  #>         WT) #>     KbBR = exp(lKbBR) #>     KbMU = exp(lKbMU) #>     KbAD = exp(lKbAD) #>     CLint = exp(lCLint + eta.LClint) #>     KbBO = exp(lKbBO) #>     KbRB = exp(lKbRB) #>     CO = (187 * WT^0.81) * 60/1000 #>     QHT = 4 * CO/100 #>     QBR = 12 * CO/100 #>     QMU = 17 * CO/100 #>     QAD = 5 * CO/100 #>     QSK = 5 * CO/100 #>     QSP = 3 * CO/100 #>     QPA = 1 * CO/100 #>     QLI = 25.5 * CO/100 #>     QST = 1 * CO/100 #>     QGU = 14 * CO/100 #>     QHA = QLI - (QSP + QPA + QST + QGU) #>     QBO = 5 * CO/100 #>     QKI = 19 * CO/100 #>     QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI) #>     QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI + QRB #>     VLU = (0.76 * WT/100)/1.051 #>     VHT = (0.47 * WT/100)/1.03 #>     VBR = (2 * WT/100)/1.036 #>     VMU = (40 * WT/100)/1.041 #>     VAD = (21.42 * WT/100)/0.916 #>     VSK = (3.71 * WT/100)/1.116 #>     VSP = (0.26 * WT/100)/1.054 #>     VPA = (0.14 * WT/100)/1.045 #>     VLI = (2.57 * WT/100)/1.04 #>     VST = (0.21 * WT/100)/1.05 #>     VGU = (1.44 * WT/100)/1.043 #>     VBO = (14.29 * WT/100)/1.99 #>     VKI = (0.44 * WT/100)/1.05 #>     VAB = (2.81 * WT/100)/1.04 #>     VVB = (5.62 * WT/100)/1.04 #>     VRB = (3.86 * WT/100)/1.04 #>     BP = 0.61 #>     fup = 0.028 #>     fub = fup/BP #>     KbLU = exp(0.8334) #>     KbHT = exp(1.1205) #>     KbSK = exp(-0.5238) #>     KbSP = exp(0.3224) #>     KbPA = exp(0.3224) #>     KbLI = exp(1.7604) #>     KbST = exp(0.3224) #>     KbGU = exp(1.2026) #>     KbKI = exp(1.3171) #>     S15 = VVB * BP/1000 #>     C15 = Venous_Blood/S15 #>     d/dt(Lungs) = QLU * (Venous_Blood/VVB - Lungs/KbLU/VLU) #>     d/dt(Heart) = QHT * (Arterial_Blood/VAB - Heart/KbHT/VHT) #>     d/dt(Brain) = QBR * (Arterial_Blood/VAB - Brain/KbBR/VBR) #>     d/dt(Muscles) = QMU * (Arterial_Blood/VAB - Muscles/KbMU/VMU) #>     d/dt(Adipose) = QAD * (Arterial_Blood/VAB - Adipose/KbAD/VAD) #>     d/dt(Skin) = QSK * (Arterial_Blood/VAB - Skin/KbSK/VSK) #>     d/dt(Spleen) = QSP * (Arterial_Blood/VAB - Spleen/KbSP/VSP) #>     d/dt(Pancreas) = QPA * (Arterial_Blood/VAB - Pancreas/KbPA/VPA) #>     d/dt(Liver) = QHA * Arterial_Blood/VAB + QSP * Spleen/KbSP/VSP +  #>         QPA * Pancreas/KbPA/VPA + QST * Stomach/KbST/VST + QGU *  #>         Gut/KbGU/VGU - CLint * fub * Liver/KbLI/VLI - QLI * Liver/KbLI/VLI #>     d/dt(Stomach) = QST * (Arterial_Blood/VAB - Stomach/KbST/VST) #>     d/dt(Gut) = QGU * (Arterial_Blood/VAB - Gut/KbGU/VGU) #>     d/dt(Bones) = QBO * (Arterial_Blood/VAB - Bones/KbBO/VBO) #>     d/dt(Kidneys) = QKI * (Arterial_Blood/VAB - Kidneys/KbKI/VKI) #>     d/dt(Arterial_Blood) = QLU * (Lungs/KbLU/VLU - Arterial_Blood/VAB) #>     d/dt(Venous_Blood) = QHT * Heart/KbHT/VHT + QBR * Brain/KbBR/VBR +  #>         QMU * Muscles/KbMU/VMU + QAD * Adipose/KbAD/VAD + QSK *  #>         Skin/KbSK/VSK + QLI * Liver/KbLI/VLI + QBO * Bones/KbBO/VBO +  #>         QKI * Kidneys/KbKI/VKI + QRB * Rest_of_Body/KbRB/VRB -  #>         QLU * Venous_Blood/VVB #>     d/dt(Rest_of_Body) = QRB * (Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB) #> })"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-cmt.html","id":"changing-compartment-numbers-by-pre-declaring-the-compartments","dir":"Articles","previous_headings":"","what":"Changing compartment numbers by pre-declaring the compartments","title":"Changing rxode2 compartment numbers","text":"add compartments rxode2 model order desire simply need pre-declare compartments cmt. example specifying Venous_Blood Skin 1st 2nd compartments, respectively, simple: can see change simple printout first two compartments Venous_Blood followed Skin.","code":"pbpk2 <- function() {   model({     ## Now this is the first compartment, ie cmt=1     cmt(Venous_Blood)     ## Skin may be a compartment you wish to dose to as well,     ##  so it is now cmt=2     cmt(Skin)     KbBR = exp(lKbBR)     KbMU = exp(lKbMU)     KbAD = exp(lKbAD)     CLint= exp(lCLint + eta.LClint)     KbBO = exp(lKbBO)     KbRB = exp(lKbRB)      ## Regional blood flows     # Cardiac output (L/h) from White et al (1968)m     CO  = (187.00*WT^0.81)*60/1000;     QHT = 4.0 *CO/100;     QBR = 12.0*CO/100;     QMU = 17.0*CO/100;     QAD = 5.0 *CO/100;     QSK = 5.0 *CO/100;     QSP = 3.0 *CO/100;     QPA = 1.0 *CO/100;     QLI = 25.5*CO/100;     QST = 1.0 *CO/100;     QGU = 14.0*CO/100;     QHA = QLI - (QSP + QPA + QST + QGU); # Hepatic artery blood flow     QBO = 5.0 *CO/100;     QKI = 19.0*CO/100;     QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI);     QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI + QRB;      ## Organs' volumes = organs' weights / organs' density     VLU = (0.76 *WT/100)/1.051;     VHT = (0.47 *WT/100)/1.030;     VBR = (2.00 *WT/100)/1.036;     VMU = (40.00*WT/100)/1.041;     VAD = (21.42*WT/100)/0.916;     VSK = (3.71 *WT/100)/1.116;     VSP = (0.26 *WT/100)/1.054;     VPA = (0.14 *WT/100)/1.045;     VLI = (2.57 *WT/100)/1.040;     VST = (0.21 *WT/100)/1.050;     VGU = (1.44 *WT/100)/1.043;     VBO = (14.29*WT/100)/1.990;     VKI = (0.44 *WT/100)/1.050;     VAB = (2.81 *WT/100)/1.040;     VVB = (5.62 *WT/100)/1.040;     VRB = (3.86 *WT/100)/1.040;      ## Fixed parameters     BP = 0.61;      # Blood:plasma partition coefficient     fup = 0.028;    # Fraction unbound in plasma     fub = fup/BP;   # Fraction unbound in blood      KbLU = exp(0.8334);     KbHT = exp(1.1205);     KbSK = exp(-.5238);     KbSP = exp(0.3224);     KbPA = exp(0.3224);     KbLI = exp(1.7604);     KbST = exp(0.3224);     KbGU = exp(1.2026);     KbKI = exp(1.3171);      ##-----------------------------------------     S15 = VVB*BP/1000;     C15 = Venous_Blood/S15      ##-----------------------------------------     d/dt(Lungs) = QLU*(Venous_Blood/VVB - Lungs/KbLU/VLU);     d/dt(Heart) = QHT*(Arterial_Blood/VAB - Heart/KbHT/VHT);     d/dt(Brain) = QBR*(Arterial_Blood/VAB - Brain/KbBR/VBR);     d/dt(Muscles) = QMU*(Arterial_Blood/VAB - Muscles/KbMU/VMU);     d/dt(Adipose) = QAD*(Arterial_Blood/VAB - Adipose/KbAD/VAD);     d/dt(Skin) = QSK*(Arterial_Blood/VAB - Skin/KbSK/VSK);     d/dt(Spleen) = QSP*(Arterial_Blood/VAB - Spleen/KbSP/VSP);     d/dt(Pancreas) = QPA*(Arterial_Blood/VAB - Pancreas/KbPA/VPA);     d/dt(Liver) = QHA*Arterial_Blood/VAB + QSP*Spleen/KbSP/VSP +       QPA*Pancreas/KbPA/VPA + QST*Stomach/KbST/VST + QGU*Gut/KbGU/VGU -       CLint*fub*Liver/KbLI/VLI - QLI*Liver/KbLI/VLI;       d/dt(Stomach) = QST*(Arterial_Blood/VAB - Stomach/KbST/VST);       d/dt(Gut) = QGU*(Arterial_Blood/VAB - Gut/KbGU/VGU);       d/dt(Bones) = QBO*(Arterial_Blood/VAB - Bones/KbBO/VBO);       d/dt(Kidneys) = QKI*(Arterial_Blood/VAB - Kidneys/KbKI/VKI);       d/dt(Arterial_Blood) = QLU*(Lungs/KbLU/VLU - Arterial_Blood/VAB);       d/dt(Venous_Blood) = QHT*Heart/KbHT/VHT + QBR*Brain/KbBR/VBR +         QMU*Muscles/KbMU/VMU + QAD*Adipose/KbAD/VAD + QSK*Skin/KbSK/VSK +         QLI*Liver/KbLI/VLI + QBO*Bones/KbBO/VBO + QKI*Kidneys/KbKI/VKI +         QRB*Rest_of_Body/KbRB/VRB - QLU*Venous_Blood/VVB;         d/dt(Rest_of_Body) = QRB*(Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB);   }) } pbpk2 <- pbpk2() pbpk2 #>  -- rxode2-based free-form 16-cmt ODE model -------------------------------------  #>  #> States ($state or $stateDf):  #>    Compartment Number Compartment Name #> 1                   1     Venous_Blood #> 2                   2             Skin #> 3                   3            Lungs #> 4                   4            Heart #> 5                   5            Brain #> 6                   6          Muscles #> 7                   7          Adipose #> 8                   8           Spleen #> 9                   9         Pancreas #> 10                 10            Liver #> 11                 11          Stomach #> 12                 12              Gut #> 13                 13            Bones #> 14                 14          Kidneys #> 15                 15   Arterial_Blood #> 16                 16     Rest_of_Body #>  -- Model (Normalized Syntax): --  #> function() { #>     model({ #>         cmt(Venous_Blood) #>         cmt(Skin) #>         KbBR = exp(lKbBR) #>         KbMU = exp(lKbMU) #>         KbAD = exp(lKbAD) #>         CLint = exp(lCLint + eta.LClint) #>         KbBO = exp(lKbBO) #>         KbRB = exp(lKbRB) #>         CO = (187 * WT^0.81) * 60/1000 #>         QHT = 4 * CO/100 #>         QBR = 12 * CO/100 #>         QMU = 17 * CO/100 #>         QAD = 5 * CO/100 #>         QSK = 5 * CO/100 #>         QSP = 3 * CO/100 #>         QPA = 1 * CO/100 #>         QLI = 25.5 * CO/100 #>         QST = 1 * CO/100 #>         QGU = 14 * CO/100 #>         QHA = QLI - (QSP + QPA + QST + QGU) #>         QBO = 5 * CO/100 #>         QKI = 19 * CO/100 #>         QRB = CO - (QHT + QBR + QMU + QAD + QSK + QLI + QBO +  #>             QKI) #>         QLU = QHT + QBR + QMU + QAD + QSK + QLI + QBO + QKI +  #>             QRB #>         VLU = (0.76 * WT/100)/1.051 #>         VHT = (0.47 * WT/100)/1.03 #>         VBR = (2 * WT/100)/1.036 #>         VMU = (40 * WT/100)/1.041 #>         VAD = (21.42 * WT/100)/0.916 #>         VSK = (3.71 * WT/100)/1.116 #>         VSP = (0.26 * WT/100)/1.054 #>         VPA = (0.14 * WT/100)/1.045 #>         VLI = (2.57 * WT/100)/1.04 #>         VST = (0.21 * WT/100)/1.05 #>         VGU = (1.44 * WT/100)/1.043 #>         VBO = (14.29 * WT/100)/1.99 #>         VKI = (0.44 * WT/100)/1.05 #>         VAB = (2.81 * WT/100)/1.04 #>         VVB = (5.62 * WT/100)/1.04 #>         VRB = (3.86 * WT/100)/1.04 #>         BP = 0.61 #>         fup = 0.028 #>         fub = fup/BP #>         KbLU = exp(0.8334) #>         KbHT = exp(1.1205) #>         KbSK = exp(-0.5238) #>         KbSP = exp(0.3224) #>         KbPA = exp(0.3224) #>         KbLI = exp(1.7604) #>         KbST = exp(0.3224) #>         KbGU = exp(1.2026) #>         KbKI = exp(1.3171) #>         S15 = VVB * BP/1000 #>         C15 = Venous_Blood/S15 #>         d/dt(Lungs) = QLU * (Venous_Blood/VVB - Lungs/KbLU/VLU) #>         d/dt(Heart) = QHT * (Arterial_Blood/VAB - Heart/KbHT/VHT) #>         d/dt(Brain) = QBR * (Arterial_Blood/VAB - Brain/KbBR/VBR) #>         d/dt(Muscles) = QMU * (Arterial_Blood/VAB - Muscles/KbMU/VMU) #>         d/dt(Adipose) = QAD * (Arterial_Blood/VAB - Adipose/KbAD/VAD) #>         d/dt(Skin) = QSK * (Arterial_Blood/VAB - Skin/KbSK/VSK) #>         d/dt(Spleen) = QSP * (Arterial_Blood/VAB - Spleen/KbSP/VSP) #>         d/dt(Pancreas) = QPA * (Arterial_Blood/VAB - Pancreas/KbPA/VPA) #>         d/dt(Liver) = QHA * Arterial_Blood/VAB + QSP * Spleen/KbSP/VSP +  #>             QPA * Pancreas/KbPA/VPA + QST * Stomach/KbST/VST +  #>             QGU * Gut/KbGU/VGU - CLint * fub * Liver/KbLI/VLI -  #>             QLI * Liver/KbLI/VLI #>         d/dt(Stomach) = QST * (Arterial_Blood/VAB - Stomach/KbST/VST) #>         d/dt(Gut) = QGU * (Arterial_Blood/VAB - Gut/KbGU/VGU) #>         d/dt(Bones) = QBO * (Arterial_Blood/VAB - Bones/KbBO/VBO) #>         d/dt(Kidneys) = QKI * (Arterial_Blood/VAB - Kidneys/KbKI/VKI) #>         d/dt(Arterial_Blood) = QLU * (Lungs/KbLU/VLU - Arterial_Blood/VAB) #>         d/dt(Venous_Blood) = QHT * Heart/KbHT/VHT + QBR * Brain/KbBR/VBR +  #>             QMU * Muscles/KbMU/VMU + QAD * Adipose/KbAD/VAD +  #>             QSK * Skin/KbSK/VSK + QLI * Liver/KbLI/VLI + QBO *  #>             Bones/KbBO/VBO + QKI * Kidneys/KbKI/VKI + QRB * Rest_of_Body/KbRB/VRB -  #>             QLU * Venous_Blood/VVB #>         d/dt(Rest_of_Body) = QRB * (Arterial_Blood/VAB - Rest_of_Body/KbRB/VRB) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-cmt.html","id":"appending-compartments-to-the-model","dir":"Articles","previous_headings":"","what":"Appending compartments to the model","title":"Changing rxode2 compartment numbers","text":"can also append “compartments” model. ODE solving internals, add fake compartments model differential equations defined. example legal: can see clearly underlying classic rxode2 model: compartments defined differential equations supported; model : give error:","code":"ode.1c.ka <- function(){   model({     C2 = center/V     d / dt(depot) = -KA * depot     d/dt(center) = KA * depot - CL*C2     cmt(eff)   }) }  ode.1c.ka <- ode.1c.ka() print(ode.1c.ka) #>  -- rxode2-based free-form 2-cmt ODE model --------------------------------------  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  -- Model (Normalized Syntax): --  #> function() { #>     model({ #>         C2 = center/V #>         d/dt(depot) = -KA * depot #>         d/dt(center) = KA * depot - CL * C2 #>         cmt(eff) #>     }) #> } ode.1c.ka$simulationModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_1e03154980d52540647eaa8bacb10b0f model (ready).  #> x$state: depot, center #> x$stateExtra: eff #> x$params: V, KA, CL #> x$lhs: C2 ode.1c.ka <- rxode2({     cmt(eff)     C2 = center/V;     d / dt(depot) = -KA * depot     d/dt(center) = KA * depot - CL*C2 }) Error in rxModelVars_(obj) :    Evaluation error: Compartment 'eff' needs differential equations defined."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-covariates.html","id":"individual-covariates","dir":"Articles","previous_headings":"","what":"Individual Covariates","title":"rxode2 Covariates","text":"individual covariate wish solve may specify iCov dataset:","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` library(units) ## udunits database from /usr/share/xml/udunits/udunits2.xml library(xgxr)  mod3 <- function() {   ini({     TKA   <- 2.94E-01     ## Clearance with individuals     TCL   <- 1.86E+01      TV2   <-4.02E+01     TQ    <-1.05E+01     TV3   <-2.97E+02     TKin  <- 1     TKout <- 1     TEC50 <-200   })   model({     KA            <- TKA     CL            <- TCL * (WT / 70) ^ 0.75     V2            <- TV2     Q             <- TQ     V3            <- TV3     Kin           <- TKin     Kout          <- TKout     EC50          <- TEC50     Tz            <- 8     amp           <- 0.1     C2            <- central/V2     C3            <- peri/V3     d/dt(depot)   <- -KA*depot     d/dt(central) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)    <-                    Q*C2 - Q*C3     d/dt(eff)     <-  Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0)        <- 1  ## This specifies that the effect compartment starts at 1.   }) }  ev <- et(amount.units=\"mg\", time.units=\"hours\") %>%   et(amt=10000, cmt=1) %>%   et(0,48,length.out=100) %>%   et(id=1:4)  set.seed(10) rxSetSeed(10) ## Now use iCov to simulate a 4-id sample r1 <- solve(mod3, ev,             # Create individual covariate data-frame             iCov=data.frame(id=1:4, WT=rnorm(4, 70, 10))) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(r1) ## ── Solved rxode2 object ── ## ── Parameters ($params): ── ##     TKA     TCL     TV2      TQ     TV3    TKin   TKout   TEC50      Tz     amp  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000   8.000   0.100  ## ── Initial Conditions ($inits): ── ##   depot central    peri     eff  ##       0       0       0       1  ## ── First part of data (object): ── ## # A tibble: 400 × 17 ##      id  time    KA    CL    V2     Q    V3   Kin  Kout  EC50    C2    C3  depot ##   <int>   [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> ## 1     1 0     0.294  18.6  40.2  10.5   297     1     1   200   0   0     10000  ## 2     1 0.485 0.294  18.6  40.2  10.5   297     1     1   200  27.8 0.257  8671. ## 3     1 0.970 0.294  18.6  40.2  10.5   297     1     1   200  43.7 0.873  7519. ## 4     1 1.45  0.294  18.6  40.2  10.5   297     1     1   200  51.7 1.68   6520. ## 5     1 1.94  0.294  18.6  40.2  10.5   297     1     1   200  54.7 2.56   5654. ## 6     1 2.42  0.294  18.6  40.2  10.5   297     1     1   200  54.5 3.45   4903. ## # ℹ 394 more rows ## # ℹ 4 more variables: central <dbl>, peri <dbl>, eff <dbl>, WT <dbl> plot(r1, C2, log=\"y\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-covariates.html","id":"time-varying-covariates","dir":"Articles","previous_headings":"","what":"Time Varying Covariates","title":"rxode2 Covariates","text":"Covariates easy specify rxode2, can specify variable. Time-varying covariates, like clock time circadian rhythm model, can also used. Extending indirect response model already discussed, : Now covariate present event dataset, system can solved combining dataset model: solving ODE equations, solver may sample times outside data. happens, ODE solver can use linear interpolation covariate values. equivalent R’s approxfun method=\"linear\".   Note linear approximation case leads kinks solved system 24-hours covariate linear interpolation near 24 near 0. linear seems reasonable, cases like clock time make interpolation methods attractive. rxode2 default covariate interpolation last observation carried forward (locf), constant approximation. equivalent R’s approxfun method=\"constant\". gives following plots:   case, plots seem smoother. can also use NONMEM’s preferred interpolation style next observation carried backward (NOCB): gives following plots:","code":"library(rxode2) library(units)  mod4 <- mod3 %>%   model(d/dt(eff) <-  Kin - Kout*(1-C2/(EC50+C2))*eff) %>%   model(-Kin) %>%   model(Kin <- TKin + amp *cos(2*pi*(ctime-Tz)/24), append=C2, cov=\"ctime\")   ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=10000, cmt=1) %>%     et(0,48,length.out=100)   ## Create data frame of 8 am dosing for the first dose This is done ## with base R but it can be done with dplyr or data.table ev$ctime <- (ev$time+set_units(8,hr)) %% 24 ev$WT <- 70 r1 <- solve(mod4, ev, covsInterpolation=\"linear\") print(r1) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>        TKA        TCL        TV2         TQ        TV3      TKout      TEC50  #>   0.294000  18.600000  40.200000  10.500000 297.000000   1.000000 200.000000  #>       TKin         Tz        amp         pi  #>   1.000000   8.000000   0.100000   3.141593  #> ── Initial Conditions ($inits): ── #>   depot central    peri     eff  #>       0       0       0       1  #> ── First part of data (object): ── #> # A tibble: 100 × 17 #>    time    KA    CL    V2     Q    V3  Kout  EC50    C2   Kin    C3  depot #>     [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> #> 1 0     0.294  18.6  40.2  10.5   297     1   200   0    1.1  0     10000  #> 2 0.485 0.294  18.6  40.2  10.5   297     1   200  27.8  1.10 0.257  8671. #> 3 0.970 0.294  18.6  40.2  10.5   297     1   200  43.7  1.10 0.874  7519. #> 4 1.45  0.294  18.6  40.2  10.5   297     1   200  51.8  1.09 1.68   6520. #> 5 1.94  0.294  18.6  40.2  10.5   297     1   200  54.8  1.09 2.56   5654. #> 6 2.42  0.294  18.6  40.2  10.5   297     1   200  54.6  1.08 3.45   4903. #> # ℹ 94 more rows #> # ℹ 5 more variables: central <dbl>, peri <dbl>, eff <dbl>, ctime [h], WT <dbl> plot(r1,C2, ylab=\"Central Concentration\") plot(r1,eff) + ylab(\"Effect\") + xlab(\"Time\") r1 <- solve(mod4, ev,covsInterpolation=\"locf\") print(r1) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>        TKA        TCL        TV2         TQ        TV3      TKout      TEC50  #>   0.294000  18.600000  40.200000  10.500000 297.000000   1.000000 200.000000  #>       TKin         Tz        amp         pi  #>   1.000000   8.000000   0.100000   3.141593  #> ── Initial Conditions ($inits): ── #>   depot central    peri     eff  #>       0       0       0       1  #> ── First part of data (object): ── #> # A tibble: 100 × 17 #>    time    KA    CL    V2     Q    V3  Kout  EC50    C2   Kin    C3  depot #>     [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> #> 1 0     0.294  18.6  40.2  10.5   297     1   200   0    1.1  0     10000  #> 2 0.485 0.294  18.6  40.2  10.5   297     1   200  27.8  1.10 0.257  8671. #> 3 0.970 0.294  18.6  40.2  10.5   297     1   200  43.7  1.10 0.874  7519. #> 4 1.45  0.294  18.6  40.2  10.5   297     1   200  51.8  1.09 1.68   6520. #> 5 1.94  0.294  18.6  40.2  10.5   297     1   200  54.8  1.09 2.56   5654. #> 6 2.42  0.294  18.6  40.2  10.5   297     1   200  54.6  1.08 3.45   4903. #> # ℹ 94 more rows #> # ℹ 5 more variables: central <dbl>, peri <dbl>, eff <dbl>, ctime [h], WT <dbl> plot(r1,C2, ylab=\"Central Concentration\", xlab=\"Time\") plot(r1,eff, ylab=\"Effect\", xlab=\"Time\") r1 <- solve(mod4, ev,covsInterpolation=\"nocb\") print(r1) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>        TKA        TCL        TV2         TQ        TV3      TKout      TEC50  #>   0.294000  18.600000  40.200000  10.500000 297.000000   1.000000 200.000000  #>       TKin         Tz        amp         pi  #>   1.000000   8.000000   0.100000   3.141593  #> ── Initial Conditions ($inits): ── #>   depot central    peri     eff  #>       0       0       0       1  #> ── First part of data (object): ── #> # A tibble: 100 × 17 #>    time    KA    CL    V2     Q    V3  Kout  EC50    C2   Kin    C3  depot #>     [h] <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> #> 1 0     0.294  18.6  40.2  10.5   297     1   200   0    1.1  0     10000  #> 2 0.485 0.294  18.6  40.2  10.5   297     1   200  27.8  1.10 0.257  8671. #> 3 0.970 0.294  18.6  40.2  10.5   297     1   200  43.7  1.10 0.874  7519. #> 4 1.45  0.294  18.6  40.2  10.5   297     1   200  51.8  1.09 1.68   6520. #> 5 1.94  0.294  18.6  40.2  10.5   297     1   200  54.8  1.09 2.56   5654. #> 6 2.42  0.294  18.6  40.2  10.5   297     1   200  54.6  1.08 3.45   4903. #> # ℹ 94 more rows #> # ℹ 5 more variables: central <dbl>, peri <dbl>, eff <dbl>, ctime [h], WT <dbl> plot(r1,C2, ylab=\"Central Concentration\", xlab=\"Time\") plot(r1,eff, ylab=\"Effect\", xlab=\"Time\")"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"creating-an-interactive-data-frame","dir":"Articles","previous_headings":"Using rxode2 data frames","what":"Creating an interactive data frame","title":"rxode2 Data Frames","text":"rxode2 supports returning solved object modified data-frame. done predict(), solve(), rxSolve() methods.","code":"library(rxode2) library(units)  ## Setup example model mod1 <- function() {   ini({     # central      KA <- 2.94E-01     CL <- 1.86E+01     # peripheral     V2 <- 4.02E+01      Q  <- 1.05E+01     V3 <- 2.97E+02     # effects     Kin  <- 1     Kout <- 1     EC50 <- 200   })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <-  KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                     Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) <- 1   }) }  ## Seup parameters and initial conditions   ## Setup dosing event information ev <- et(amountUnits=\"mg\", timeUnits = \"hours\") %>%   et(amt=10000, addl=9, ii=12) %>%   et(amt=20000, addl=4, time=120, ii=24) %>%   et(0:240)   ## Now solve x <- predict(mod1, ev)  x #> -- Solved rxode2 object -- #> -- Parameters (x$params): -- #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> -- Initial Conditions (x$inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 241 x 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # i 235 more rows"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"using-the-solved-object-as-a-simple-data-frame","dir":"Articles","previous_headings":"Using rxode2 data frames","what":"Using the solved object as a simple data frame","title":"rxode2 Data Frames","text":"solved object acts data.frame tbl can filtered dpylr. example filter easily.","code":"library(dplyr) ## You can  drop units for comparisons and filtering x <- mod1 %>% solve(ev) %>%   drop_units() %>% filter(time <= 3) %>% as_tibble()  ## or keep them and compare with the proper units. x <- mod1 %>% solve(ev) %>%     filter(time <= set_units(3, hr)) %>% as_tibble() x #> # A tibble: 4 x 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"updating-the-data-set-interactively","dir":"Articles","previous_headings":"","what":"Updating the data-set interactively","title":"rxode2 Data Frames","text":"However isn’t just simple data object. can use solved object update parameters fly, even change sampling time. First need recreate original solved system:","code":"x <- mod1 %>% solve(ev) print(x) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 241 x 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # i 235 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"modifying-observation-times-for-rxode2","dir":"Articles","previous_headings":"Updating the data-set interactively","what":"Modifying observation times for rxode2","title":"rxode2 Data Frames","text":"Notice initial effect now 2. can also change sampling times easily method changing t time. example:","code":"x$t <- seq(0,5,length.out=20) print(x) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 20 x 7 #>    time    C2     C3  depot centr  peri   eff #>     [h] <dbl>  <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1 0       0   0      10000     0    0    1    #> 2 0.263  16.8 0.0817  9255.  677.  24.3  1.01 #> 3 0.526  29.5 0.299   8566. 1187.  88.7  1.03 #> 4 0.789  38.9 0.615   7929. 1562. 183.   1.06 #> 5 1.05   45.5 1.00    7338. 1830. 298.   1.09 #> 6 1.32   50.1 1.44    6792. 2013. 427.   1.12 #> # i 14 more rows plot(x)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-data-frame.html","id":"modifying-simulation-parameters","dir":"Articles","previous_headings":"Updating the data-set interactively","what":"Modifying simulation parameters","title":"rxode2 Data Frames","text":"can also access change parameters $ operator. example, accessing KA can done : may change assigning new value.  can access/change parameters, initialization(s) events $params, $inits, $events accessor syntax, similar used . syntax makes easy update explore effect various parameters solved object.","code":"x$KA #> [1] 0.294 x$KA <- 1 print(x) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #>    KA    CL    V2     Q    V3   Kin  Kout  EC50  #>   1.0  18.6  40.2  10.5 297.0   1.0   1.0 200.0  #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #> -- First part of data (object): -- #> # A tibble: 20 x 7 #>    time    C2    C3  depot centr   peri   eff #>     [h] <dbl> <dbl>  <dbl> <dbl>  <dbl> <dbl> #> 1 0       0   0     10000     0     0    1    #> 2 0.263  52.2 0.261  7686. 2098.   77.6  1.03 #> 3 0.526  83.3 0.900  5908. 3348.  267.   1.09 #> 4 0.789  99.8 1.75   4541. 4010.  519.   1.15 #> 5 1.05  106.  2.69   3490. 4273.  800.   1.21 #> 6 1.32  106.  3.66   2683. 4272. 1086.   1.26 #> # i 14 more rows plot(x)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"datasets-for-rxode2-nlmixr","dir":"Articles","previous_headings":"","what":"Datasets for rxode2 & nlmixr","title":"rxode2 Event Types","text":"Data input nlmixr type data input rxode2, similar data NONMEM (NONMEM-ready datasets can used directly nlmixr).","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"subject-identification-columns","dir":"Articles","previous_headings":"Columns Described by Type of Use","what":"Subject Identification Columns","title":"rxode2 Event Types","text":"subject identification column separates subjects identification random effects. ID: subject identifier may integer, character, factor.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"observation-columns","dir":"Articles","previous_headings":"Columns Described by Type of Use","what":"Observation Columns","title":"rxode2 Event Types","text":"Observation columns used indicate dependent variable use measure . DV: numeric column measurement CENS: numeric column indication censoring, limit quantification assay. LIMIT: numeric column helping indicate type censoring, limit quantification assay. MDV: indicator missing DV values CMT: name number compartment DVID: dependent variable identifier EVID event identifier","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dosing-columns","dir":"Articles","previous_headings":"Columns Described by Type of Use","what":"Dosing Columns","title":"rxode2 Event Types","text":"AMT: amount dose CMT: name number compartment EVID: event identifier ADDL: number additional doses RATE DUR: rate duration dose","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"details-for-specific-dataset-columns","dir":"Articles","previous_headings":"","what":"Details for Specific Dataset Columns","title":"rxode2 Event Types","text":"details sorted alphabetically column name. grouping use, see documentation .","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"amt-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"AMT Column","title":"rxode2 Event Types","text":"AMT column defines amount dose. observation rows, 0 NA. dosing rows, amount dose administered CMT. dose zero-order rate (constant infusion), infusion may setup using RATE DUR column.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"censlimit-columns","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"CENS/LIMIT Columns","title":"rxode2 Event Types","text":"CENS column indicator column indicating censoring occurred. pharmacokinetic modeling, censoring typically sample limit quantification. Internally rxode2 saves values nlmixr can use likelihood calculations. CENS = 0 indicates value DV measured without censoring. CENS = 1 indicates value left censored (limit quantitation) value DV censoring/quantitation limit. CENS = -1 indicates value right censored (limit quantitation) value DV censoring/quantitation limit. LIMIT additional information censoring handled nlmixr stored rxode2’s data structure well. value left censored, like limit 1 may also believe value certain threshold, like zero. case, limit 0 indicates censored value 0 1. short : CENS = 0 LIMIT ignored observation censored CENS = 1 value censored (LIMIT, DV) CENS = -1 value censored (DV, LIMIT)","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"cmt-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"CMT Column","title":"rxode2 Event Types","text":"CMT column indicates compartment event occurs. given character string factor (preferred method), matched name model. given integer, matched order compartments appear model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dur-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"DUR Column","title":"rxode2 Event Types","text":"DUR column defines duration infusion. used set duration zero-order rate infusion.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dv-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"DV Column","title":"rxode2 Event Types","text":"DV column indicates current measurement current compartment (see CMT) current measurement identifier (see DVID) may missing (see MDV) censored (see CENS).","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"dvid-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"DVID Column","title":"rxode2 Event Types","text":"TODO","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"evid-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"EVID Column","title":"rxode2 Event Types","text":"EVID column event identifier row data. observation records, 0. normal dosing records, 1. Many EVID values detailed rxode2 Event Types Classic rxode2 Events vignettes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"id-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"ID Column","title":"rxode2 Event Types","text":"ID column subject identifier. column used separate one individual (usually single person animal) another. model, ID column used separate individuals. numerical integrator re-initializes new individual, new values random effects selected.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-datasets.html","id":"rate-column","dir":"Articles","previous_headings":"Details for Specific Dataset Columns","what":"RATE Column","title":"rxode2 Event Types","text":"TODO","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"creating-rxode2s-event-tables","dir":"Articles","previous_headings":"","what":"Creating rxode2’s event tables","title":"Easy rxode2 Event Tables","text":"event table rxode2 specialized data frame acts container rxode2’s events observation times. create rxode2 event table may use code eventTable(), et(), even create data frame right event information contained . closely related types events rxode2 supports. event table can add sampling/observations doses piping direct access. short table two main functions create dosing Sampling times can added add.sampling( sampling times ) et( sampling times ). Dosing intervals sampling windows also supported. documentation using et() syntax, though supported. models, can illustrate using model shared rxode2 tutorial:","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` library(units) #> udunits database from /usr/share/xml/udunits/udunits2.xml (ev <- et()) #> -- EventTable with 0 records -- #> 0 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) (ev <- et()) #> -- EventTable with 0 records -- #> 0 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) ## Model from rxode2 tutorial m1 <-function() {   ini({     KA   <- 2.94E-01     CL   <- 1.86E+01     V2   <- 4.02E+01     Q    <- 1.05E+01     V3   <- 2.97E+02     Kin  <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250   })   model({     C2          <- centr/V2     C3          <- peri/V3     d/dt(depot) <- -KA*depot     f(depot)    <- fdepot     dur(depot)  <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff;     eff(0)      <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"adding-doses-to-the-event-table","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Adding doses to the event table","title":"Easy rxode2 Event Tables","text":"created can add dosing event table add.dosing(), et() functions. Using add.dosing() function : Notice units specified table. specified, units use units package keep track units convert needed. Additionally, ggforce uses label ggplot axes. set_units drop_units useful set drop rxode2 event table units. example, can see time axes labeled:  familiar NONMEM/rxode2 event records, can also specify dosing using et dose elements directly: gives:  shows easy creating event tables can .","code":"ev <- eventTable(amount.units=\"mg\", time.units=\"hr\")  ## The methods ar attached to the event table, so you can use  ## them directly ev$add.dosing(dose=10000, nbr.doses = 3)# loading doses ## Starts at time 0; Default dosing interval is 24  ## You can also pipe the event tables to these methods. ev <- ev %>%   add.dosing(dose=5000, nbr.doses=14,              dosing.interval=12)# maintenance  ev #> -- EventTable with 2 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 2 x 5 #>   time   amt  ii  addl evid         #>    [h]  [mg] [h] <int> <evid>       #> 1    0 10000  24     2 1:Dose (Add) #> 2    0  5000  12    13 1:Dose (Add) rxSolve(m1, ev) %>% plot(C2) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ev <- et(timeUnits=\"hr\") %>%   et(amt=10000, until = set_units(3, days),      ii=12) # loading doses  ev #> -- EventTable with 1 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 1 x 5 #>   time   amt  ii  addl evid         #>    [h] <dbl> [h] <int> <evid>       #> 1    0 10000  12     6 1:Dose (Add) rxSolve(m1, ev) %>% plot(C2) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"adding-sampling-to-an-event-table","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Adding sampling to an event table","title":"Easy rxode2 Event Tables","text":"notice examples, rxode2 generated default sampling times since sampling times. wish control sampling time, add samples rxode2 event table add.sampling et gives:  use et can simply add similar way add.sampling: gives following rxode2 solve:  Note jagged nature plots since sample times.","code":"ev <- eventTable(amount.units=\"mg\", time.units=\"hr\")  ## The methods ar attached to the event table, so you can use them ## directly ev$add.dosing(dose=10000, nbr.doses = 3)# loading doses  ev$add.sampling(seq(0,24,by=4))  ev #> -- EventTable with 8 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 7 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 8 x 5 #>   time   amt  ii  addl evid          #>    [h]  [mg] [h] <int> <evid>        #> 1    0    NA  NA    NA 0:Observation #> 2    0 10000  24     2 1:Dose (Add)  #> 3    4    NA  NA    NA 0:Observation #> 4    8    NA  NA    NA 0:Observation #> 5   12    NA  NA    NA 0:Observation #> 6   16    NA  NA    NA 0:Observation #> 7   20    NA  NA    NA 0:Observation #> 8   24    NA  NA    NA 0:Observation solve(m1, ev) %>% plot(C2) ev <- et(timeUnits=\"hr\") %>%   et(amt=10000, until = set_units(3, days),      ii=12) %>% # loading doses   et(seq(0,24,by=4))  ev #> -- EventTable with 8 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 7 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 8 x 5 #>   time   amt  ii  addl evid          #>    [h] <dbl> [h] <int> <evid>        #> 1    0    NA  NA    NA 0:Observation #> 2    0 10000  12     6 1:Dose (Add)  #> 3    4    NA  NA    NA 0:Observation #> 4    8    NA  NA    NA 0:Observation #> 5   12    NA  NA    NA 0:Observation #> 6   16    NA  NA    NA 0:Observation #> 7   20    NA  NA    NA 0:Observation #> 8   24    NA  NA    NA 0:Observation solve(m1, ev) %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"expand-the-event-table-to-a-multi-subject-event-table-","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Expand the event table to a multi-subject event table.","title":"Easy rxode2 Event Tables","text":"thing needed expand event table list IDs want expand; can see following simulation 4 individuals solved :","code":"ev <- et(timeUnits=\"hr\") %>%   et(amt=10000, until = set_units(3, days),      ii=12) %>% # loading doses   et(seq(0,48,length.out=200)) %>%   et(id=1:4)  ev #> -- EventTable with 804 records -- #> 4 dosing records (see x$get.dosing(); add with add.dosing or et) #> 800 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 804 x 6 #>       id  time   amt  ii  addl evid          #>    <int>   [h] <dbl> [h] <int> <evid>        #>  1     1 0        NA  NA    NA 0:Observation #>  2     1 0     10000  12     6 1:Dose (Add)  #>  3     1 0.241    NA  NA    NA 0:Observation #>  4     1 0.482    NA  NA    NA 0:Observation #>  5     1 0.724    NA  NA    NA 0:Observation #>  6     1 0.965    NA  NA    NA 0:Observation #>  7     1 1.21     NA  NA    NA 0:Observation #>  8     1 1.45     NA  NA    NA 0:Observation #>  9     1 1.69     NA  NA    NA 0:Observation #> 10     1 1.93     NA  NA    NA 0:Observation #> # i 794 more rows set.seed(42) rxSetSeed(42) solve(m1, ev,       params=data.frame(KA=0.294*exp(rnorm(4)),                         CL=18.6*exp(rnorm(4)))) %>%     plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"add-doses-and-samples-within-a-sampling-window","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Add doses and samples within a sampling window","title":"Easy rxode2 Event Tables","text":"addition adding fixed doses fixed sampling times, can windows sample draw doses . dosing windows specify time ordered numerical vector lowest dosing time highest dosing time inside list. example, start dosing time 6 hour dosing window: can clearly see different dosing times following simulation:  course reality dosing interval may 2 hours:  sort thing can specified sampling times. specify sampling times terms sampling window, can create list sampling times. sampling time two element ordered numeric vector.  shows flexibility dosing sampling rxode2 event tables allow.","code":"set.seed(42) rxSetSeed(42) ev <- et(timeUnits=\"hr\") %>%   et(time=list(c(0,6)), amt=10000, until = set_units(2, days),      ii=12) %>% # loading doses   et(id=1:4)  ev #> -- EventTable with 16 records -- #> 16 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 16 x 6 #>       id low   time high   amt evid         #>    <int> [h]    [h]  [h] <dbl> <evid>       #>  1     1   0  5.49     6 10000 1:Dose (Add) #>  2     1  12 17.0     18 10000 1:Dose (Add) #>  3     1  24 25.7     30 10000 1:Dose (Add) #>  4     1  36 41.6     42 10000 1:Dose (Add) #>  5     2   0  4.31     6 10000 1:Dose (Add) #>  6     2  12 14.7     18 10000 1:Dose (Add) #>  7     2  24 28.2     30 10000 1:Dose (Add) #>  8     2  36 39.9     42 10000 1:Dose (Add) #>  9     3   0  0.808    6 10000 1:Dose (Add) #> 10     3  12 16.4     18 10000 1:Dose (Add) #> 11     3  24 27.1     30 10000 1:Dose (Add) #> 12     3  36 39.9     42 10000 1:Dose (Add) #> 13     4   0  4.98     6 10000 1:Dose (Add) #> 14     4  12 13.7     18 10000 1:Dose (Add) #> 15     4  24 29.6     30 10000 1:Dose (Add) #> 16     4  36 41.5     42 10000 1:Dose (Add) ev <- ev %>% et(seq(0,48,length.out=200))  solve(m1, ev,       params=data.frame(KA=0.294*exp(rnorm(4)),                         CL=18.6*exp(rnorm(4)))) %>%   plot(C2) set.seed(42) rxSetSeed(42) ev <- et(timeUnits=\"hr\") %>%   et(time=list(c(0,2)), amt=10000, until = set_units(2, days),      ii=12) %>% # loading doses   et(id=1:4) %>%   et(seq(0,48,length.out=200))  solve(m1, ev,       params=data.frame(KA=0.294*exp(rnorm(4)),                         CL=18.6*exp(rnorm(4)))) %>%   plot(C2) rxSetSeed(42) set.seed(42) ev <- et(timeUnits=\"hr\") %>%   et(time=list(c(0,2)), amt=10000, until = set_units(2, days),      ii=12) %>% # loading doses   et(id=1:4)  ## Create 20 samples in the first 24 hours and 20 samples in the ## second 24 hours samples <- c(lapply(1:20, function(...){c(0,24)}),              lapply(1:20, function(...){c(20,48)}))  ## Add the random collection to the event table ev <- ev %>% et(samples)  library(ggplot2) solve(m1, ev, params=data.frame(KA=0.294*exp(rnorm(4)),                                 CL=18.6*exp(rnorm(4)))) %>%   plot(C2) + geom_point()"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"combining-event-tables","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Combining event tables","title":"Easy rxode2 Event Tables","text":"Since can create dosing records sampling records, can create complex dosing regimen wish. addition, rxode2 allows combine event tables c, seq, rep, rbind.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"sequencing-event-tables","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Sequencing event tables","title":"Easy rxode2 Event Tables","text":"One way combine event table sequence c, seq etSeq. takes two dosing groups adds least one inter-dose interval :  sequencing events, can also separate sequence period time; example wanted separate week, easily following sequence event tables:  Note example time bid qd event tables exactly one week, 1 week plus 24 hours inter-dose interval. want behavior, can sequence using wait=\"+ii\".  Also note, rxode2 assumes dosing want space event tables , clears sampling records combine event tables. true, can also use option samples=\"use\"","code":"## bid for 5 days bid <- et(timeUnits=\"hr\") %>%        et(amt=10000,ii=12,until=set_units(5, \"days\"))  ## qd for 5 days qd <- et(timeUnits=\"hr\") %>%       et(amt=20000,ii=24,until=set_units(5, \"days\"))  ## bid for 5 days followed by qd for 5 days et <- seq(bid,qd) %>% et(seq(0,11*24,length.out=100));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days followed by qd for 5 days et <- seq(bid,set_units(1, \"week\"), qd) %>%     et(seq(0,18*24,length.out=100));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days followed by qd for 5 days et <- seq(bid,set_units(1, \"week\"), qd,wait=\"+ii\") %>%     et(seq(0,18*24,length.out=100));  rxSolve(m1, et) %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"repeating-event-tables","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Repeating event tables","title":"Easy rxode2 Event Tables","text":"can event table can repeat etRep rep. example 4 rounds 2 weeks QD therapy 1 week therapy can simply specified:  simplified way use sequence event tables. Therefore, many options still apply; samples cleared unless use samples=\"use\", time event tables least inter-dose interval. can adjust timing wait option.","code":"qd <-et(timeUnits = \"hr\") %>%   et(amt=10000, ii=24, until=set_units(2, \"weeks\"), cmt=\"depot\")  et <- rep(qd, times=4, wait=set_units(1,\"weeks\")) %>%       add.sampling(set_units(seq(0, 12.5,by=0.005),weeks))  rxSolve(m1, et)  %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"combining-event-tables-with-rbind","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Combining event tables with rbind","title":"Easy rxode2 Event Tables","text":"may combine event tables rbind. consider event times combining event tables, keeps times. space event tables waiting period, also consider inter-dose interval. Using previous seq can clearly see difference. sequence:  bind together rbind  Still waiting period applies (consider inter-dose interval)  can also bind tables together make ID event table unique; can good combine cohorts different expected dosing sampling times. requires id=\"unique\" option; Using first example shows different case:","code":"## bid for 5 days bid <- et(timeUnits=\"hr\") %>%        et(amt=10000,ii=12,until=set_units(5, \"days\"))  ## qd for 5 days qd <- et(timeUnits=\"hr\") %>%       et(amt=20000,ii=24,until=set_units(5, \"days\"))  et <- seq(bid,qd) %>%     et(seq(0,18*24,length.out=500));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days et <- rbind(bid,qd) %>%     et(seq(0,18*24,length.out=500));  rxSolve(m1, et) %>% plot(C2) et <- rbind(bid,wait=set_units(10,days),qd) %>%     et(seq(0,18*24,length.out=500));  rxSolve(m1, et) %>% plot(C2) ## bid for 5 days et <- etRbind(bid,qd, id=\"unique\") %>%     et(seq(0,150,length.out=500));  library(ggplot2) rxSolve(m1, et) %>% plot(C2) + facet_wrap( ~ id)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-table.html","id":"event-tables-in-rstudio-notebooks","dir":"Articles","previous_headings":"Creating rxode2’s event tables","what":"Event tables in Rstudio Notebooks","title":"Easy rxode2 Event Tables","text":"addition output console shown examples, Rstudio notebook output different can seen following screenshots; first screenshot shows event table looks evaluating Rstduio notebook  simple dataframe allows page contents. click first box Rstudio notebook output, notes event table: ## Expanding events Event tables can expanded contain addl data item, like following example: can expand events addl items $expand() etExpand(ev): first, etExpand(ev) expands event table without modifying original data frame: can see addl events expanded, however original data frame remained intact: use ev$expand() modify ev object. similar object-oriented method:","code":"ev <- et() %>%   et(dose=50, ii=8, until=48)  ev #> -- EventTable with 1 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 1 x 5 #>    time   amt    ii  addl evid         #>   <dbl> <dbl> <dbl> <int> <evid>       #> 1     0    50     8     6 1:Dose (Add) etExpand(ev) #> -- EventTable with 7 records -- #> 7 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 7 x 4 #>    time   amt    ii evid         #>   <dbl> <dbl> <dbl> <evid>       #> 1     0    50     0 1:Dose (Add) #> 2     8    50     0 1:Dose (Add) #> 3    16    50     0 1:Dose (Add) #> 4    24    50     0 1:Dose (Add) #> 5    32    50     0 1:Dose (Add) #> 6    40    50     0 1:Dose (Add) #> 7    48    50     0 1:Dose (Add) print(ev) #> -- EventTable with 1 records -- #> 1 dosing records (see $get.dosing(); add with add.dosing or et) #> 0 observation times (see $get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with $expand(); or etExpand() #> -- First part of : -- #> # A tibble: 1 x 5 #>    time   amt    ii  addl evid         #>   <dbl> <dbl> <dbl> <int> <evid>       #> 1     0    50     8     6 1:Dose (Add) ev$expand() ev #> -- EventTable with 7 records -- #> 7 dosing records (see x$get.dosing(); add with add.dosing or et) #> 0 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 7 x 4 #>    time   amt    ii evid         #>   <dbl> <dbl> <dbl> <evid>       #> 1     0    50     0 1:Dose (Add) #> 2     8    50     0 1:Dose (Add) #> 3    16    50     0 1:Dose (Add) #> 4    24    50     0 1:Dose (Add) #> 5    32    50     0 1:Dose (Add) #> 6    40    50     0 1:Dose (Add) #> 7    48    50     0 1:Dose (Add)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"rxode2-event-tables","dir":"Articles","previous_headings":"","what":"rxode2 event tables","title":"rxode2 Event Types","text":"general, rxode2 event tables follow NONMEM dataset convention exceptions: may turn compartment negative compartment number “-cmt” cmt compartment name. compartment data item (cmt) can still number, number compartment defined appearance compartment name model. can tedious count, can specify compartment numbers easier using cmt(cmtName) beginning model. Bioavailability changes change rate infusion since dur/amt fixed input data. Similarly, specifying rate/amt infusion, bioavailability change infusion duration since rate/amt fixed input data. infrequent NONMEM columns supported: pcmt, call. evid=5 replace event; replaces value compartment value specified amt column. equivalent deSolve=replace. evid=6 multiply event; multiplies value compartment value specified amt column. equivalent deSolve=multiply. evid=7 transit compartment model/phantom event. puts dose dose() function calculates time since last dose tad() doesn’t actually put dose compartment. allows transit() function easily apply compartment. legal entries data table: notes: evid can classic RxODE (described ) NONMEM-style evid described . NONMEM’s DV required; rxode2 ODE solving framework. NONMEM’s MDV required, since captured EVID. Instead NONMEM-compatible data, can accept deSolve compatible data-frames. returning rxode2 solved data-set additional event ids (EVID) may see depending solving options: EVID = -1 modeled rate ends (corresponds rate = -1) EVID = -2 modeled duration ends (corresponds rate=-2) EVID = -10 rate specified zero-order infusion ends (corresponds rate > 0) EVID = -20 duration specified zero-order infusion ends (corresponds dur > 0) EVID = 101, 102, 103,... correspond 1, 2, 3, … modeled time (mtime). can accessed solving option combination addDosing=TRUE subsetNonmem=FALSE. want see classic EVID equivalents can use addDosing=NA. illustrate event types use model original rxode2 tutorial.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` ## Model from rxode2 tutorial  m1 <- function() {   ini({     KA   <- 2.94E-01     CL   <- 1.86E+01     V2   <- 4.02E+01     Q    <- 1.05E+01     V3   <- 2.97E+02     Kin  <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250   })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     f(depot) <- fdepot     dur(depot) <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)  <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"bolusadditive-doses","dir":"Articles","previous_headings":"","what":"Bolus/Additive Doses","title":"rxode2 Event Types","text":"bolus dose default type dose rxode2 requires amt/dose. Note uses convenience function et() described rxode2 event tables","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 5 #>     time   amt  ii  addl evid          #>      [h] <dbl> [h] <int> <evid>        #>  1 0        NA  NA    NA 0:Observation #>  2 0     10000  12     2 1:Dose (Add)  #>  3 0.242    NA  NA    NA 0:Observation #>  4 0.485    NA  NA    NA 0:Observation #>  5 0.727    NA  NA    NA 0:Observation #>  6 0.970    NA  NA    NA 0:Observation #>  7 1.21     NA  NA    NA 0:Observation #>  8 1.45     NA  NA    NA 0:Observation #>  9 1.70     NA  NA    NA 0:Observation #> 10 1.94     NA  NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev) %>% plot(C2) +     xlab(\"Time\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"infusion-doses","dir":"Articles","previous_headings":"","what":"Infusion Doses","title":"rxode2 Event Types","text":"different type infusions rxode2 supports: Constant Rate Infusion (rate) Constant Duration Infusion (dur) Estimated Rate Infusion Estimated Duration Infusion","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"constant-infusion-in-terms-of-duration-and-rate","dir":"Articles","previous_headings":"Infusion Doses","what":"Constant Infusion (in terms of duration and rate)","title":"rxode2 Event Types","text":"next type event infusion; two ways specify infusion; first dur keyword. example :  can also specified rate component:  exception bioavailability changes infusion. case modeling rate, bioavailability decrease, decreases infusion duration, NONMEM. example:  Similarly increasing bioavailability increases infusion duration.  rationale behavior rate amt specified event table, thing can change bioavailability increase duration infusion. specify amt dur components event table, bioavailability changes affect rate infusion. can see side--side comparison bioavailability changes affecting rate instead duration records following plots:","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, dur=8) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt  ii  addl evid          dur #>      [h] <dbl> [h] <int> <evid>        [h] #>  1 0        NA  NA    NA 0:Observation  NA #>  2 0     10000  12     2 1:Dose (Add)    8 #>  3 0.242    NA  NA    NA 0:Observation  NA #>  4 0.485    NA  NA    NA 0:Observation  NA #>  5 0.727    NA  NA    NA 0:Observation  NA #>  6 0.970    NA  NA    NA 0:Observation  NA #>  7 1.21     NA  NA    NA 0:Observation  NA #>  8 1.45     NA  NA    NA 0:Observation  NA #>  9 1.70     NA  NA    NA 0:Observation  NA #> 10 1.94     NA  NA    NA 0:Observation  NA #> # i 91 more rows rxSolve(m1, ev) %>% plot(depot, C2) +     xlab(\"Time\") ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, rate=10000/8) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt rate        ii  addl evid          #>      [h] <dbl> <rate/dur> [h] <int> <evid>        #>  1 0        NA NA          NA    NA 0:Observation #>  2 0     10000  1250       12     2 1:Dose (Add)  #>  3 0.242    NA NA          NA    NA 0:Observation #>  4 0.485    NA NA          NA    NA 0:Observation #>  5 0.727    NA NA          NA    NA 0:Observation #>  6 0.970    NA NA          NA    NA 0:Observation #>  7 1.21     NA NA          NA    NA 0:Observation #>  8 1.45     NA NA          NA    NA 0:Observation #>  9 1.70     NA NA          NA    NA 0:Observation #> 10 1.94     NA NA          NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev) %>% plot(depot, C2) +     xlab(\"Time\") rxSolve(m1, ev, c(fdepot=0.25)) %>% plot(depot, C2) +     xlab(\"Time\") rxSolve(m1, ev, c(fdepot=1.25)) %>% plot(depot, C2) +     xlab(\"Time\") ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, dur=8) %>%     et(seq(0, 24, length.out=100)) library(ggplot2) library(patchwork)  p1 <- rxSolve(m1, ev, c(fdepot=1.25)) %>% plot(depot) +     xlab(\"Time\") + ylim(0,5000)  p2 <- rxSolve(m1, ev, c(fdepot=0.25)) %>% plot(depot) +     xlab(\"Time\")+ ylim(0,5000)  ## Use patchwork syntax to combine plots p1 * p2"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"modeled-rate-and-duration-of-infusion","dir":"Articles","previous_headings":"Infusion Doses","what":"Modeled Rate and Duration of Infusion","title":"rxode2 Event Types","text":"can model duration, equivalent NONMEM’s rate=-2.  Similarly, may also model rate. equivalent NONMEM’s rate=-1 rxode2’s event table specifies data item well.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, rate=-2) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt rate        ii  addl evid          #>      [h] <dbl> <rate/dur> [h] <int> <evid>        #>  1 0        NA NA          NA    NA 0:Observation #>  2 0     10000 -2:dur      12     2 1:Dose (Add)  #>  3 0.242    NA NA          NA    NA 0:Observation #>  4 0.485    NA NA          NA    NA 0:Observation #>  5 0.727    NA NA          NA    NA 0:Observation #>  6 0.970    NA NA          NA    NA 0:Observation #>  7 1.21     NA NA          NA    NA 0:Observation #>  8 1.45     NA NA          NA    NA 0:Observation #>  9 1.70     NA NA          NA    NA 0:Observation #> 10 1.94     NA NA          NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev, c(durDepot=7)) %>% plot(depot, C2) +     xlab(\"Time\") ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12,until=24, rate=-1) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 101 x 6 #>     time   amt rate        ii  addl evid          #>      [h] <dbl> <rate/dur> [h] <int> <evid>        #>  1 0        NA NA          NA    NA 0:Observation #>  2 0     10000 -1:rate     12     2 1:Dose (Add)  #>  3 0.242    NA NA          NA    NA 0:Observation #>  4 0.485    NA NA          NA    NA 0:Observation #>  5 0.727    NA NA          NA    NA 0:Observation #>  6 0.970    NA NA          NA    NA 0:Observation #>  7 1.21     NA NA          NA    NA 0:Observation #>  8 1.45     NA NA          NA    NA 0:Observation #>  9 1.70     NA NA          NA    NA 0:Observation #> 10 1.94     NA NA          NA    NA 0:Observation #> # i 91 more rows rxSolve(m1, ev, c(rateDepot=10000/3)) %>% plot(depot, C2) +     xlab(\"Time\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"steady-state","dir":"Articles","previous_headings":"","what":"Steady State","title":"rxode2 Event Types","text":"doses solved steady state reached constant inter-dose interval.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, ss=1) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 101 records -- #> 1 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> -- First part of x: -- #> # A tibble: 101 x 5 #>     time   amt  ii evid             ss #>      [h] <dbl> [h] <evid>        <int> #>  1 0        NA  NA 0:Observation    NA #>  2 0     10000  12 1:Dose (Add)      1 #>  3 0.242    NA  NA 0:Observation    NA #>  4 0.485    NA  NA 0:Observation    NA #>  5 0.727    NA  NA 0:Observation    NA #>  6 0.970    NA  NA 0:Observation    NA #>  7 1.21     NA  NA 0:Observation    NA #>  8 1.45     NA  NA 0:Observation    NA #>  9 1.70     NA  NA 0:Observation    NA #> 10 1.94     NA  NA 0:Observation    NA #> # i 91 more rows rxSolve(m1, ev) %>% plot(C2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"steady-state-for-complex-dosing","dir":"Articles","previous_headings":"Steady State","what":"Steady state for complex dosing","title":"rxode2 Event Types","text":"using ss=2 flag, can use super-positioning principle linear kinetics get steady state nonstandard dosing (.e. morning 100 mg vs evening 150 mg). done : Saving state values Resetting states solving system steady state Adding back prior state values  can see takes full dose cycle reach true complex steady state dosing.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=24, ss=1) %>%     et(time=12, amt=15000, ii=24, ss=2) %>%     et(time=24, amt=10000, ii=24, addl=3) %>%     et(time=36, amt=15000, ii=24, addl=3) %>%     et(seq(0, 64, length.out=500))  library(ggplot2)  rxSolve(m1, ev,maxsteps=10000) %>% plot(C2) +   annotate(\"rect\", xmin=0, xmax=24, ymin=-Inf, ymax=Inf,            alpha=0.2) +   annotate(\"text\", x=12.5, y=7,            label=\"Initial Steady State Period\") +   annotate(\"text\", x=44,   y=7,            label=\"Steady State AM/PM dosing\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"steady-state-for-constant-infusion-or-zero-order-processes","dir":"Articles","previous_headings":"Steady State","what":"Steady state for constant infusion or zero order processes","title":"rxode2 Event Types","text":"last type steady state rxode2 supports steady-state constant infusion rate. can specified way NONMEM, : inter-dose interval ii=0 steady state dose, ie ss=1 Either positive rate (rate>0) estimated rate rate=-1. zero dose, ie amt=0 steady-state constant infusion achieved, infusion turned using record, just like NONMEM. Note rate=-2 model duration infusion doesn’t make much sense since solving infusion steady state. duration specified steady state solution. Also note bioavailability changes steady state infusion also make sense neither change rate duration steady state infusion. Hence modeled bioavailability type dosing event ignored. example:  can used PK, can used steady-state disease processes.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=0, ss=1,rate=10000/8)  p1 <- rxSolve(m1, ev) %>% plot(C2, eff)   ev <- et(timeUnits=\"hr\") %>%     et(amt=200000, rate=10000/8) %>%     et(0, 250, length.out=1000)  p2 <- rxSolve(m1, ev) %>% plot(C2, eff)  library(patchwork)  p1 / p2"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"reset-events","dir":"Articles","previous_headings":"","what":"Reset Events","title":"rxode2 Event Types","text":"Reset events implemented evid=3 evid=reset, reset evid=4 reset dose. solving show happens system system reset 6 hours post-dose.  can see compartments reset initial values. next dose start dosing cycle . case, whole system reset dose given","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, evid=reset) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 102 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 102 x 5 #>     time   amt  ii  addl evid          #>      [h] <dbl> [h] <int> <evid>        #>  1 0        NA  NA    NA 0:Observation #>  2 0     10000  12     3 1:Dose (Add)  #>  3 0.242    NA  NA    NA 0:Observation #>  4 0.485    NA  NA    NA 0:Observation #>  5 0.727    NA  NA    NA 0:Observation #>  6 0.970    NA  NA    NA 0:Observation #>  7 1.21     NA  NA    NA 0:Observation #>  8 1.45     NA  NA    NA 0:Observation #>  9 1.70     NA  NA    NA 0:Observation #> 10 1.94     NA  NA    NA 0:Observation #> # i 92 more rows rxSolve(m1, ev) %>% plot(depot,C2, eff) ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, amt=10000, evid=4) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 102 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 102 x 5 #>     time   amt  ii  addl evid          #>      [h] <dbl> [h] <int> <evid>        #>  1 0        NA  NA    NA 0:Observation #>  2 0     10000  12     3 1:Dose (Add)  #>  3 0.242    NA  NA    NA 0:Observation #>  4 0.485    NA  NA    NA 0:Observation #>  5 0.727    NA  NA    NA 0:Observation #>  6 0.970    NA  NA    NA 0:Observation #>  7 1.21     NA  NA    NA 0:Observation #>  8 1.45     NA  NA    NA 0:Observation #>  9 1.70     NA  NA    NA 0:Observation #> 10 1.94     NA  NA    NA 0:Observation #> # i 92 more rows rxSolve(m1, ev) %>% plot(depot,C2, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-event-types.html","id":"turning-off-compartments","dir":"Articles","previous_headings":"","what":"Turning off compartments","title":"rxode2 Event Types","text":"may also turn compartment, similar reset event. Solving shows system:  case, depot turned , depot compartment concentrations set initial values compartment concentrations/levels reset. another dose depot administered depot compartment turned back . Note dose compartment turns back compartment dosed. Hence turn effect compartment, continues another dose depot.  turn back compartment, zero-dose compartment evid=2 compartment needed.","code":"ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, cmt=\"-depot\", evid=2) %>%     et(seq(0, 24, length.out=100))  ev #> -- EventTable with 102 records -- #> 2 dosing records (see x$get.dosing(); add with add.dosing or et) #> 100 observation times (see x$get.sampling(); add with add.sampling or et) #> multiple doses in `addl` columns, expand with x$expand(); or etExpand(x) #> -- First part of x: -- #> # A tibble: 102 x 6 #>     time cmt         amt  ii  addl evid          #>      [h] <chr>     <dbl> [h] <int> <evid>        #>  1 0     (obs)        NA  NA    NA 0:Observation #>  2 0     (default) 10000  12     3 1:Dose (Add)  #>  3 0.242 (obs)        NA  NA    NA 0:Observation #>  4 0.485 (obs)        NA  NA    NA 0:Observation #>  5 0.727 (obs)        NA  NA    NA 0:Observation #>  6 0.970 (obs)        NA  NA    NA 0:Observation #>  7 1.21  (obs)        NA  NA    NA 0:Observation #>  8 1.45  (obs)        NA  NA    NA 0:Observation #>  9 1.70  (obs)        NA  NA    NA 0:Observation #> 10 1.94  (obs)        NA  NA    NA 0:Observation #> # i 92 more rows rxSolve(m1, ev) %>% plot(depot,C2, eff) ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, cmt=\"-eff\", evid=2) %>%     et(seq(0, 24, length.out=100))  rxSolve(m1, ev) %>% plot(depot,C2, eff) ev <- et(timeUnits=\"hr\") %>%     et(amt=10000, ii=12, addl=3) %>%     et(time=6, cmt=\"-eff\", evid=2) %>%     et(time=12,cmt=\"eff\",evid=2) %>%     et(seq(0, 24, length.out=100))  rxSolve(m1, ev) %>% plot(depot,C2, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"classic-rxode-evid-values","dir":"Articles","previous_headings":"","what":"Classic RxODE evid values","title":"Classic rxode2 Events","text":"Originally RxODE supported compound event IDs; rxode2 still supports parameters, often useful use normal NONMEM dataset standard used many modeling tools like NONMEM, Monolix nlmixr, described rxode2 types article. Classically, RxODE supported event coding single event id evid described following table. classic EVID concatenate numbers table, infusion compartment 1 10101 infusion compartment 199 119901. EVID = 0 (observations), EVID=2 (type event) EVID=3 supported. Internally EVID=9 non-observation event makes sure system initialized zero; EVID=9 manually set. EVID 10-99 represents modeled time interventions, similar NONMEM’s MTIME. along amount (amt) time columns specify events ODE system. infusions specified EVIDs > 100 amt column represents rate value. Infusion flags 1 2 +amt turn infusion specific compartment -amt turn infusion specific compartment. specify dose/duration place dosing records time duration starts stops. modeled rate/duration infusion flags infusion flag must followed infusion record. number concatenated together form full RxODE event ID, shown following examples:","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"bolus-dose-examples","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Bolus Dose Examples","title":"Classic rxode2 Events","text":"100 bolus dose compartment #1 time 0 100 bolus dose compartment #99 time 0 100 bolus dose compartment #199 time 0","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"infusion-event-examples","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Infusion Event Examples","title":"Classic rxode2 Events","text":"Bolus infusion rate 50 compartment 1 1.5 hr, (modeled bioavailability changes duration infusion) Bolus infusion rate 50 compartment 1 1.5 hr (modeled bioavailability changes rate infusion) Modeled rate amount 50 Modeled duration amount 50","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"steady-state-for-classic-rxode-evid-example","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Steady State for classic RxODE EVID example","title":"Classic rxode2 Events","text":"Steady state dose cmt 1 Steady State super-positioning principle 50 pm 100 dose","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"steady-state-with-lagged-dose-for-classic-rxode-evid","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Steady state with lagged dose for classic RxODE evid","title":"Classic rxode2 Events","text":"Steady state lagged dose bolus dose: time | evid | amt | ii | 0 | 109 | 100 | 24 | 0 | 101 | 100 | 0 | event 109 calculates trough amount steady state passed (ii-lag_time) event 101 applies next steady state dose. Steady state (=2) lagged bolus dose Steady state infusion lag time time | evid | amt | ii | 0 | 10109 | 5 | 24 | 0 | 10108 | -5 | 24 | 0 | 10101 | 5 | | 20 | 10101 | -5 | | case bolus, 10109 event calculates trough concentration ii=24, may (may ) still infusion running. infusion running 10108 event turn infusion appropriate time. infusion completed, 10108 event ignored. next 2 events 10101 represent event times (assuming bioavailability applied).","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-events-classic.html","id":"turning-off-a-compartment-with-classic-rxode-evid","dir":"Articles","previous_headings":"Classic RxODE evid values","what":"Turning off a compartment with classic RxODE EVID","title":"Classic rxode2 Events","text":"Turn first compartment time 12 Event coding rxode2 encoded single event number evid. compartments 100, coded : event 0 observation events. 100*(Compartment Number) + 1 dose captured amt 10000 + 100*(Compartment Number) + 1 infusion rate captured amt column infusion turned subtracting amt evid stop infusion. compartments greater equal 100, 100s place digits transferred 100,000th place digit. doses 99th compartment evid bolus dose 9901 evid infusion 19901. bolus dose 199th compartment evid bolus dose 109901. infusion dosing record 199th compartment 119901.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"introduction","dir":"Articles","previous_headings":"","what":"Introduction","title":"Introduction to rxode2","text":"rxode2 R package facilitates simulation ODE models R. designed pharmacometrics models mind, can applied generally ODE model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"description-of-rxode2-illustrated-through-an-example","dir":"Articles","previous_headings":"","what":"Description of rxode2 illustrated through an example","title":"Introduction to rxode2","text":"model equations can specified text string, model file R expression. differential algebraic equations permitted. Differential equations specified d/dt(var_name) =. equation can separated semicolon. load rxode2 package compile model: Model parameters may specified ini({}) model block, initial conditions can specified within model cmt(0)= X, like model eff(0) <- 1. may also specify subject variability initial conditions residual error components just like nlmixr2. allows single interface nlmixr2/rxode2 models. Also note, classic rxode2 interface still works just like past (don’t worry breaking code time). fact, can get classic rxode2 model $simulationModel ui object:","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod1 <- function() {   ini({     # central      KA=2.94E-01     CL=1.86E+01     V2=4.02E+01     # peripheral     Q=1.05E+01     V3=2.97E+02     # effects     Kin=1     Kout=1     EC50=200    })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <- Q*C2 - Q*C3     eff(0) <- 1     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff   }) } mod1 <- mod1() # create the ui object (can also use `rxode2(mod1)`) mod1  summary(mod1$simulationModel)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"specify-dosing-and-sampling-in-rxode2","dir":"Articles","previous_headings":"Description of rxode2 illustrated through an example","what":"Specify Dosing and sampling in rxode2","title":"Introduction to rxode2","text":"rxode2 provides simple flexible way specify dosing sampling functions generate event table. First, empty event table generated “et()” function. interface similar NONMEM event tables: can see code, can dose compartment named rxode2 model. slight deviation NONMEM can reduce need compartment renumbering. events can also combined expanded (multi-subject events complex regimens) rbind, c, seq, rep. information creating complex dosing regimens using rxode2 see rxode2 events vignette.","code":"ev  <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(0:240) # Add sampling"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-intro.html","id":"solving-odes","dir":"Articles","previous_headings":"Description of rxode2 illustrated through an example","what":"Solving ODEs","title":"Introduction to rxode2","text":"ODE can now solved using rxSolve: returns modified data frame. can see compartment values plot :  ,  Note labels automatically labeled units initial event table. rxode2 extracts units label plot (present).","code":"x <- mod1 %>% rxSolve(ev) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ x #> ── Solved rxode2 object ── #> ── Parameters (x$params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions (x$inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows library(ggplot2) plot(x,C2) + ylab(\"Central Concentration\") plot(x,eff)  + ylab(\"Effect\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-model-types.html","id":"prediction-only-models","dir":"Articles","previous_headings":"","what":"Prediction only models","title":"rxode2 additional model types","text":"Prediction models simple create. use rxode2 syntax without ODE systems . simple example one-compartment model. Solving rxode2 models saving simple ODE system, faster course.","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` mod <- function(){   model({     ipre <- 10 * exp(-ke * t)   }) } et  <- et(seq(0,24,length.out=50)) cmt1 <- rxSolve(mod,et,params=c(ke=0.5)) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ cmt1 ## -- Solved rxode2 object -- ## -- Parameters (x$params): -- ##  ke  ## 0.5  ## -- Initial Conditions (x$inits): -- ## named numeric(0) ## -- First part of data (object): -- ## # A tibble: 50 x 2 ##    time  ipre ##   <dbl> <dbl> ## 1 0     10    ## 2 0.490  7.83 ## 3 0.980  6.13 ## 4 1.47   4.80 ## 5 1.96   3.75 ## 6 2.45   2.94 ## # i 44 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-model-types.html","id":"solved-compartment-models","dir":"Articles","previous_headings":"","what":"Solved compartment models","title":"rxode2 additional model types","text":"Solved models also simple create. simply place linCmt() psuedo-function code. linCmt() function figures type model use based parameter names specified. often, pharmacometric models parameterized terms volume clearances. Clearances specified NONMEM-style names CL, Q, Q1, Q2, etc. distributional clearances CLD, CLD2. Volumes specified Central (VC V), Peripheral/Tissue (VP, VT). translations available, example translations : Another popular parameterization terms micro-constants. rxode2 assumes compartment 1 central compartment. elimination constant specified K, Ke Kel. example translations : last parameterization possible using alpha V //B/C. example translations : linCmt() sleuthing complete, 1, 2 3 compartment model solution used value linCmt(). compartments can dose linear solved system depot central linear absorption constant model ka. Without additional ODEs, compartments numbered depot=1 central=2. absorption constant ka missing, may dose central compartment. Without additional ODEs compartment number central=1. compartments take sort events ODE model can take, discussed rxode2 events vignette. acts ODE model; specify dose depot compartment solve system:","code":"mod <- function() {   ini({     kel <- 0.5     V <- 1   })   model({     ipre <- linCmt(V, kel)   }) } et  <- et(amt=10,time=0,cmt=depot) %>%   et(seq(0,24,length.out=50)) cmt1 <- rxSolve(mod,et) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ cmt1 ## -- Solved rxode2 object -- ## -- Parameters (x$params): -- ## kel   V  ## 0.5 1.0  ## -- Initial Conditions (x$inits): -- ## named numeric(0) ## -- First part of data (object): -- ## # A tibble: 50 x 2 ##    time  ipre ##   <dbl> <dbl> ## 1 0     10    ## 2 0.490  7.83 ## 3 0.980  6.13 ## 4 1.47   4.80 ## 5 1.96   3.75 ## 6 2.45   2.94 ## # i 44 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-model-types.html","id":"mixing-solved-systems-and-odes","dir":"Articles","previous_headings":"","what":"Mixing Solved Systems and ODEs","title":"rxode2 additional model types","text":"addition pure ODEs, may mix solved systems ODEs. prior 2-compartment indirect response model can simplified linCmt() function: allows indirect response model assign 2-compartment model C2 variable used indirect response model. mixing solved systems ODEs, solved system’s compartment always last compartment. solved system technically isn’t compartment solved. Adding dosing compartment end interfere actual ODE solved. Therefore,two-compartment indirect response model, effect compartment compartment #1 PK dosing compartment depot compartment #2. compartment model requires new event table since compartment number changed: can solved following command: Note solving require specifying effect compartment initial condition 1. Rather, already pre-specified eff(0)=1. can solved different initial conditions easily: rxode2 detective also require specify variables linCmt() function already defined block. Therefore, following function also work solve system. Note specify parameters solving system since built model, can override parameters:","code":"library(rxode2) ## Setup example model mod1 <-function() {   model({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) =-KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  =                    Q*C2 - Q*C3     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff   }) }  ## Seup parameters and initial conditions  theta <-     c(KA=2.94E-01, CL=1.86E+01, V2=4.02E+01, # central       Q=1.05E+01,  V3=2.97E+02,              # peripheral       Kin=1, Kout=1, EC50=200)               # effects  inits <- c(eff=1)  ## Setup dosing event information ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=10000, addl=9, ii=12) %>%     et(amt=20000, addl=4, time=120, ii=24) %>%     add.sampling(0:240)  ## Setup a mixed solved/ode system: mod2 <- function() {   model({     ## the order of variables do not matter, the type of compartmental     ## model is determined by the parameters specified.     C2   = linCmt(KA, CL, V2, Q, V3);     eff(0) = 1  ## This specifies that the effect compartment starts at 1.     d/dt(eff) =  Kin - Kout*(1-C2/(EC50+C2))*eff;   }) } ev <- et(amountUnits='mg', timeUnits='hours') %>%   et(amt=10000, addl=9, ii=12, cmt=2) %>%   et(amt=20000, addl=4, time=120, ii=24, cmt=2) %>%   et(0:240) x <- mod2 %>%  solve(theta, ev) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##      KA      CL      V2       Q      V3     Kin    Kout    EC50  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  ## -- Initial Conditions ($inits): -- ## eff  ##   1  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   1    ## 2    1 121.   1.35 ## 3    2  60.3  1.38 ## 4    3  31.0  1.28 ## 5    4  17.0  1.18 ## 6    5  10.2  1.11 ## # i 235 more rows x <- mod2 %>%  solve(theta, ev,c(eff=2)) print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##      KA      CL      V2       Q      V3     Kin    Kout    EC50  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  ## -- Initial Conditions ($inits): -- ## eff  ##   2  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   2    ## 2    1 121.   1.93 ## 3    2  60.3  1.67 ## 4    3  31.0  1.41 ## 5    4  17.0  1.23 ## 6    5  10.2  1.13 ## # i 235 more rows mod3 <- function() {   ini({     KA <- 2.94E-01     CL <- 1.86E+01     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200   })   model({     # Since the parameters are in the ini block, put them in linCmt so     # that the model is detected correctly     C2   <- linCmt(KA, CL, V2, Q, V3)     eff(0) <- 1  ## This specifies that the effect compartment starts at 1.     d/dt(eff) <-  Kin - Kout*(1-C2/(EC50+C2))*eff;   }) }  x <- mod3 %>%  solve(ev)  print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##      KA      CL      V2       Q      V3     Kin    Kout    EC50  ##   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  ## -- Initial Conditions ($inits): -- ## eff  ##   1  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   1    ## 2    1 121.   1.35 ## 3    2  60.3  1.38 ## 4    3  31.0  1.28 ## 5    4  17.0  1.18 ## 6    5  10.2  1.11 ## # i 235 more rows x <- mod3 %>%  solve(c(KA=10),ev) print(x) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##    KA    CL    V2     Q    V3   Kin  Kout  EC50  ##  10.0  18.6  40.2  10.5 297.0   1.0   1.0 200.0  ## -- Initial Conditions ($inits): -- ## eff  ##   1  ## -- First part of data (object): -- ## # A tibble: 241 x 3 ##   time    C2   eff ##    [h] <dbl> <dbl> ## 1    0 249.   1    ## 2    1 121.   1.35 ## 3    2  60.3  1.38 ## 4    3  31.0  1.28 ## 5    4  17.0  1.18 ## 6    5  10.2  1.11 ## # i 235 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"nesting-in-rxode2","dir":"Articles","previous_headings":"","what":"Nesting in rxode2","title":"Nesting levels in rxode2","text":"one level nesting possible rxode2; example using following uncertainties sources variability:","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"event-table","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Event table","title":"Nesting levels in rxode2","text":"event table contains nesting variables: inv: investigator id id: subject id eye: eye id (left right) occ: occasion","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(seq(0, 240, by=4)) %>% # Assumes sampling when there is no dosing information   et(seq(0, 240, by=4) + 0.1) %>% ## adds 0.1 for separate eye   et(id=1:20) %>%   ## Add an occasion per dose   mutate(occ=cumsum(!is.na(amt))) %>%   mutate(occ=ifelse(occ == 0, 1, occ)) %>%   mutate(occ=2- occ %% 2) %>%   mutate(eye=ifelse(round(time) == time, 1, 2)) %>%   mutate(inv=ifelse(id < 10, 1, 2)) %>% as_tibble ->   ev"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"rxode2-model","dir":"Articles","previous_headings":"Nesting in rxode2","what":"rxode2 model","title":"Nesting levels in rxode2","text":"creates rxode2 model multi-level nesting. Note variables inv.Cl, inv.Ka, eta.Cl etc; need one variable level nesting.","code":"mod <- rxode2({   ## Clearance with individuals   eff(0) = 1   C2 = centr/V2*(1+prop.sd)   C3 = peri/V3   CL =  TCl*exp(eta.Cl + eye.Cl + iov.Cl + inv.Cl)   KA = TKA * exp(eta.Ka + eye.Ka + iov.Cl + inv.Ka)   d/dt(depot) =-KA*depot   d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3   d/dt(peri)  =                    Q*C2 - Q*C3   d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff   ef0 = eff + add.sd }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"uncertainty-in-model-parameters","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Uncertainty in Model parameters","title":"Nesting levels in rxode2","text":"","code":"theta <- c(\"TKA\"=0.294, \"TCl\"=18.6, \"V2\"=40.2,            \"Q\"=10.5, \"V3\"=297, \"Kin\"=1, \"Kout\"=1, \"EC50\"=200)  ## Creating covariance matrix tmp <- matrix(rnorm(8^2), 8, 8) tMat <- tcrossprod(tmp, tmp) / (8 ^ 2) dimnames(tMat) <- list(names(theta), names(theta))  tMat #>               TKA           TCl           V2            Q           V3 #> TKA   0.173571236 -0.1003204607  0.038185010 -0.004108928 -0.095032973 #> TCl  -0.100320461  0.2195710868 -0.043849095  0.013295549 -0.007477895 #> V2    0.038185010 -0.0438490948  0.129784612 -0.017270432 -0.038004762 #> Q    -0.004108928  0.0132955493 -0.017270432  0.022145634  0.020376451 #> V3   -0.095032973 -0.0074778948 -0.038004762  0.020376451  0.165568340 #> Kin  -0.040867119 -0.0492597458 -0.003056722 -0.033468634 -0.003021883 #> Kout  0.035469225  0.0275087955  0.033725901  0.027668205  0.005497301 #> EC50  0.026158042  0.0009434711  0.039426946 -0.036283167 -0.093134292 #>               Kin         Kout          EC50 #> TKA  -0.040867119  0.035469225  0.0261580416 #> TCl  -0.049259746  0.027508796  0.0009434711 #> V2   -0.003056722  0.033725901  0.0394269465 #> Q    -0.033468634  0.027668205 -0.0362831667 #> V3   -0.003021883  0.005497301 -0.0931342922 #> Kin   0.226735493 -0.083447793  0.0659884544 #> Kout -0.083447793  0.117195570 -0.0291684598 #> EC50  0.065988454 -0.029168460  0.0928611407"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"nesting-variability","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Nesting Variability","title":"Nesting levels in rxode2","text":"specify multiple levels nesting, can specify nested lotri matrix; using approach use condition operator | specify variable nesting occurs ; Bayesian simulation need specify much information parameter; rxode2 nu parameter. case: - id, nu=100 model came 100 subjects - eye, nu=200 model came 200 eyes - occ, nu=200 model came 200 occasions - inv, nu=10 model came 10 investigators specify lotri can use | var(nu=X), :","code":"omega <- lotri(lotri(eta.Cl ~ 0.1,                      eta.Ka ~ 0.1) | id(nu=100),                lotri(eye.Cl ~ 0.05,                      eye.Ka ~ 0.05) | eye(nu=200),                lotri(iov.Cl ~ 0.01,                      iov.Ka ~ 0.01) | occ(nu=200),                lotri(inv.Cl ~ 0.02,                      inv.Ka ~ 0.02) | inv(nu=10)) omega #> $id #>        eta.Cl eta.Ka #> eta.Cl    0.1    0.0 #> eta.Ka    0.0    0.1 #>  #> $eye #>        eye.Cl eye.Ka #> eye.Cl   0.05   0.00 #> eye.Ka   0.00   0.05 #>  #> $occ #>        iov.Cl iov.Ka #> iov.Cl   0.01   0.00 #> iov.Ka   0.00   0.01 #>  #> $inv #>        inv.Cl inv.Ka #> inv.Cl   0.02   0.00 #> inv.Ka   0.00   0.02 #>  #> Properties: nu"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"unexplained-variability","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Unexplained variability","title":"Nesting levels in rxode2","text":"last piece variability specify unexplained variability","code":"sigma <- lotri(prop.sd ~ .25,                add.sd~ 0.125)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-nesting.html","id":"solving-the-problem","dir":"Articles","previous_headings":"Nesting in rxode2","what":"Solving the problem","title":"Nesting levels in rxode2","text":"multiple investigators study; investigator number individuals enrolled site. rxode2 automatically determines number investigators simulate effect investigator. output, inv.Cl(inv==1) inv.Cl investigator 1, inv.Cl(inv==2) inv.Cl investigator 2, etc. inv.Cl(inv==1), inv.Cl(inv==2), etc simulated study combined form investigator variability. equation form represent following: look simulated parameters can see inv.Cl(inv==1) inv.Cl(inv==2) s$params; study: eye variability occasion variability individual simulates number variables become eye occasion variability; case eye: look simulation variables (ie eye.Cl(eye==1), eye.Cl(eye==2), etc) change individual combined make eye variability occasion variability can seen pharamcometric models.","code":"s <- rxSolve(mod, theta, ev,              thetaMat=tMat, omega=omega,              sigma=sigma, sigmaDf=400,              nStud=400) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> unhandled error message: EE:[lsoda] 70000 steps taken before reaching tout #>  @(lsoda.c:754 #> Warning: some ID(s) could not solve the ODEs correctly; These values are #> replaced with 'NA' print(s) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #> # A tibble: 8,000 x 24 #>    sim.id id    `inv.Cl(inv==1)` `inv.Cl(inv==2)` `inv.Ka(inv==1)` #>     <int> <fct>            <dbl>            <dbl>            <dbl> #>  1      1 1              0.00900           -0.117            0.312 #>  2      1 2              0.00900           -0.117            0.312 #>  3      1 3              0.00900           -0.117            0.312 #>  4      1 4              0.00900           -0.117            0.312 #>  5      1 5              0.00900           -0.117            0.312 #>  6      1 6              0.00900           -0.117            0.312 #>  7      1 7              0.00900           -0.117            0.312 #>  8      1 8              0.00900           -0.117            0.312 #>  9      1 9              0.00900           -0.117            0.312 #> 10      1 10             0.00900           -0.117            0.312 #> # i 7,990 more rows #> # i 19 more variables: `inv.Ka(inv==2)` <dbl>, `eye.Cl(eye==1)` <dbl>, #> #   `eye.Cl(eye==2)` <dbl>, `eye.Ka(eye==1)` <dbl>, `eye.Ka(eye==2)` <dbl>, #> #   `iov.Cl(occ==1)` <dbl>, `iov.Cl(occ==2)` <dbl>, `iov.Ka(occ==1)` <dbl>, #> #   `iov.Ka(occ==2)` <dbl>, V2 <dbl>, V3 <dbl>, TCl <dbl>, eta.Cl <dbl>, #> #   TKA <dbl>, eta.Ka <dbl>, Q <dbl>, Kin <dbl>, Kout <dbl>, EC50 <dbl> #> -- Initial Conditions ($inits): -- #> depot centr  peri   eff  #>     0     0     0     1  #>  #> Simulation with uncertainty in: #> * parameters ($thetaMat for changes) #> * omega matrix ($omegaList) #>  #> -- First part of data (object): -- #> # A tibble: 976,000 x 21 #>   sim.id    id time  inv.Cl inv.Ka  eye.Cl  eye.Ka iov.Cl iov.Ka    C2     C3 #>    <int> <int>  [h]   <dbl>  <dbl>   <dbl>   <dbl>  <dbl>  <dbl> <dbl>  <dbl> #> 1      1     1  0   0.00900  0.312 -0.330  -0.0547 0.0370  0.139  0    0      #> 2      1     1  0.1 0.00900  0.312  0.0792  0.185  0.0370  0.139  8.04 0.0108 #> 3      1     1  4   0.00900  0.312 -0.330  -0.0547 0.0370  0.139 55.4  5.58   #> 4      1     1  4.1 0.00900  0.312  0.0792  0.185  0.0370  0.139 23.2  5.69   #> 5      1     1  8   0.00900  0.312 -0.330  -0.0547 0.0370  0.139 19.6  7.95   #> 6      1     1  8.1 0.00900  0.312  0.0792  0.185  0.0370  0.139 20.0  7.97   #> # i 975,994 more rows #> # i 10 more variables: CL <dbl>, KA <dbl>, ef0 <dbl>, depot <dbl>, centr <dbl>, #> #   peri <dbl>, eff <dbl>, occ <fct>, eye <fct>, inv <fct> inv.Cl = (inv == 1) * `inv.Cl(inv==1)` + (inv == 2) * `inv.Cl(inv==2)` print(head(s$params)) #>   sim.id id inv.Cl(inv==1) inv.Cl(inv==2) inv.Ka(inv==1) inv.Ka(inv==2) #> 1      1  1    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 2      1  2    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 3      1  3    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 4      1  4    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 5      1  5    0.008995575     -0.1165083      0.3118205   -0.001926179 #> 6      1  6    0.008995575     -0.1165083      0.3118205   -0.001926179 #>   eye.Cl(eye==1) eye.Cl(eye==2) eye.Ka(eye==1) eye.Ka(eye==2) iov.Cl(occ==1) #> 1    -0.33023666    0.079190835    -0.05471051     0.18487157    0.037021761 #> 2    -0.42211135   -0.006773841    -0.15399939     0.22463048   -0.007762485 #> 3    -0.05128596    0.123504458     0.18243012     0.13361259   -0.168921159 #> 4    -0.06360621    0.340443016     0.65606292     0.04807945    0.157027939 #> 5     0.08009126   -0.027627640     0.06662396    -0.05838949   -0.226341946 #> 6    -0.11937190    0.090540884    -0.17320340     0.02039840   -0.211648032 #>   iov.Cl(occ==2) iov.Ka(occ==1) iov.Ka(occ==2)       V2       V3      TCl #> 1    -0.19726803     0.13932112    0.088763868 40.26476 296.4748 19.25043 #> 2     0.04555116     0.07716277    0.049148937 40.26476 296.4748 19.25043 #> 3     0.17916513     0.08347085   -0.002990168 40.26476 296.4748 19.25043 #> 4    -0.13444417     0.24736740   -0.077647794 40.26476 296.4748 19.25043 #> 5     0.02049190    -0.09568733    0.054658457 40.26476 296.4748 19.25043 #> 6     0.12619350    -0.01547841    0.040206608 40.26476 296.4748 19.25043 #>        eta.Cl       TKA      eta.Ka        Q       Kin     Kout     EC50 #> 1  0.09788109 0.2354877 -0.21526790 10.43157 0.9256189 1.044937 200.3679 #> 2  0.33302598 0.2354877 -0.01028112 10.43157 0.9256189 1.044937 200.3679 #> 3  0.08504325 0.2354877  0.37253705 10.43157 0.9256189 1.044937 200.3679 #> 4  0.16521678 0.2354877 -0.01859974 10.43157 0.9256189 1.044937 200.3679 #> 5  0.53131380 0.2354877 -0.37109644 10.43157 0.9256189 1.044937 200.3679 #> 6 -0.30786713 0.2354877  0.14897588 10.43157 0.9256189 1.044937 200.3679 print(head(s$params %>% filter(sim.id == 2))) #>   sim.id id inv.Cl(inv==1) inv.Cl(inv==2) inv.Ka(inv==1) inv.Ka(inv==2) #> 1      2  1    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 2      2  2    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 3      2  3    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 4      2  4    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 5      2  5    -0.07227613    -0.01884171     0.07620864    -0.01962499 #> 6      2  6    -0.07227613    -0.01884171     0.07620864    -0.01962499 #>   eye.Cl(eye==1) eye.Cl(eye==2) eye.Ka(eye==1) eye.Ka(eye==2) iov.Cl(occ==1) #> 1    -0.25102634      0.1354374   -0.135230741    -0.01605664    0.081212198 #> 2    -0.24832622      0.2233348   -0.380770273    -0.37101991    0.207985292 #> 3     0.24081384     -0.3826770    0.133361609    -0.32496203   -0.099983807 #> 4    -0.11021791      0.3655308    0.009423897     0.08736599   -0.002580804 #> 5     0.21155058      0.3626918    0.194977611    -0.19452347    0.058392202 #> 6     0.03663028     -0.1003754    0.068047203     0.32368023    0.167658084 #>   iov.Cl(occ==2) iov.Ka(occ==1) iov.Ka(occ==2)       V2       V3      TCl #> 1    -0.06556217   -0.032678168   -0.061277262 40.21091 296.6424 18.86433 #> 2    -0.05016896    0.130258486   -0.103721118 40.21091 296.6424 18.86433 #> 3     0.06144127   -0.262726137   -0.104186144 40.21091 296.6424 18.86433 #> 4     0.03954265   -0.037289852   -0.021344468 40.21091 296.6424 18.86433 #> 5     0.12713682   -0.007530921    0.081238197 40.21091 296.6424 18.86433 #> 6    -0.02811353    0.026801132   -0.002702762 40.21091 296.6424 18.86433 #>        eta.Cl       TKA      eta.Ka        Q       Kin     Kout   EC50 #> 1  0.19368101 0.5686675  0.26184170 10.54673 0.5434136 1.323825 200.09 #> 2 -0.16262369 0.5686675  0.12969487 10.54673 0.5434136 1.323825 200.09 #> 3  0.22164455 0.5686675  0.24501105 10.54673 0.5434136 1.323825 200.09 #> 4  0.05997662 0.5686675  0.03059399 10.54673 0.5434136 1.323825 200.09 #> 5  0.45691421 0.5686675 -0.20098716 10.54673 0.5434136 1.323825 200.09 #> 6 -0.19943300 0.5686675 -0.36846082 10.54673 0.5434136 1.323825 200.09 eye.Cl = (eye == 1) * `eye.Cl(eye==1)` + (eye == 2) * `eye.Cl(eye==2)`"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"setting-up-the-rxode2-model-for-the-pipeline","dir":"Articles","previous_headings":"","what":"Setting up the rxode2 model for the pipeline","title":"rxode2 in a pipeline","text":"example show use rxode2 simple pipeline. can start model can used different simulation workflows rxode2 can handle: tumor growth model described Ribba 2012. case, compiled model R object Ribba2012, though rxode2 simulation pipeline, assign compiled model object, though think makes sense.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` Ribba2012 <- function() {   ini({     k = 100     tkde = 0.24     eta.tkde = 0     tkpq = 0.0295     eta.kpq = 0     tkqpp = 0.0031     eta.kqpp = 0     tlambdap = 0.121     eta.lambdap = 0     tgamma = 0.729     eta.gamma = 0     tdeltaqp = 0.00867     eta.deltaqp = 0     prop.sd <- 0     tpt0 = 7.13     eta.pt0 = 0     tq0 = 41.2     eta.q0 = 0   })   model({     kde ~ tkde*exp(eta.tkde)     kpq ~ tkpq * exp(eta.kpq)     kqpp ~ tkqpp * exp(eta.kqpp)     lambdap ~ tlambdap*exp(eta.lambdap)     gamma ~ tgamma*exp(eta.gamma)     deltaqp ~ tdeltaqp*exp(eta.deltaqp)     d/dt(c) = -kde * c     d/dt(pt) = lambdap * pt *(1-pstar/k) + kqpp*qp -       kpq*pt - gamma*c*kde*pt     d/dt(q) = kpq*pt -gamma*c*kde*q     d/dt(qp) = gamma*c*kde*q - kqpp*qp - deltaqp*qp     ## initial conditions     pt0 ~ tpt0*exp(eta.pt0)     q0 ~ tq0*exp(eta.q0)     pt(0) = pt0     q(0) = q0     pstar <- (pt+q+qp)     pstar ~  prop(prop.sd)   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulating-one-event-table","dir":"Articles","previous_headings":"","what":"Simulating one event table","title":"rxode2 in a pipeline","text":"Simulating single event table quite simple: pipe rxode2 simulation object event table object et(). events completely specified, simply solve ODE system rxSolve(). case can pipe output plot() conveniently view results. pt (Proliferative Tissue), q (quiescent tissue) qp (DNA-Damaged quiescent tissue) pstar (total tumor tissue)","code":"Ribba2012 %>% # Use rxode2     et(time.units=\"months\") %>% # Pipe to a new event table     et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months     et(0, 250, by=0.5) %>% # Add some sampling times (not required)     rxSolve() %>% # Solve the simulation     plot(pt, q, qp, pstar) # Plot it, plotting the variables of interest #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulating-with-between-subject-variability","dir":"Articles","previous_headings":"Simulating multiple subjects from a single event table","what":"Simulating with between subject variability","title":"rxode2 in a pipeline","text":"next sort simulation may useful simulating multiple patients treatments. case, use omega matrix specified paper: information, easy simulate 3 subjects model-based parameters:  Note two different things added simulation: - nSub specify many subjects model - omega specify subject variability.","code":"## Add CVs from paper for individual simulation ## Uses exact formula:  lognCv = function(x){log((x/100)^2+1)}  library(lotri) ## Now create omega matrix ## I'm using lotri to quickly specify names/diagonals omega <- lotri(eta.pt0 ~ lognCv(94),                eta.q0 ~ lognCv(54),                eta.lambdap ~ lognCv(72),                eta.kqp ~ lognCv(76),                eta.kqpp ~ lognCv(97),                eta.deltaqp ~ lognCv(115),                eta.tkde ~ lognCv(70))  omega #>               eta.pt0    eta.q0 eta.lambdap   eta.kqp  eta.kqpp eta.deltaqp #> eta.pt0     0.6331848 0.0000000   0.0000000 0.0000000 0.0000000   0.0000000 #> eta.q0      0.0000000 0.2558818   0.0000000 0.0000000 0.0000000   0.0000000 #> eta.lambdap 0.0000000 0.0000000   0.4176571 0.0000000 0.0000000   0.0000000 #> eta.kqp     0.0000000 0.0000000   0.0000000 0.4559047 0.0000000   0.0000000 #> eta.kqpp    0.0000000 0.0000000   0.0000000 0.0000000 0.6631518   0.0000000 #> eta.deltaqp 0.0000000 0.0000000   0.0000000 0.0000000 0.0000000   0.8426442 #> eta.tkde    0.0000000 0.0000000   0.0000000 0.0000000 0.0000000   0.0000000 #>              eta.tkde #> eta.pt0     0.0000000 #> eta.q0      0.0000000 #> eta.lambdap 0.0000000 #> eta.kqp     0.0000000 #> eta.kqpp    0.0000000 #> eta.deltaqp 0.0000000 #> eta.tkde    0.3987761 set.seed(1089) rxSetSeed(1089) Ribba2012 %>% # Use rxode2     et(time.units=\"months\") %>% # Pipe to a new event table     et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months     et(0, 250, by=0.5) %>% # Add some sampling times (not required)     rxSolve(nSub=3, omega=omega) %>% # Solve the simulation     plot(pt, q, qp, pstar) # Plot it, plotting the variables of interest #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> Warning: multi-subject simulation without without 'omega'"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulation-with-unexplained-variability","dir":"Articles","previous_headings":"Simulating multiple subjects from a single event table","what":"Simulation with unexplained variability","title":"rxode2 in a pipeline","text":"can even add unexplained variability quite easily:  case added sigma matrix unexplained variability pstar total tumor tissue. can even simulate uncertainty theta omega sigma values wish.","code":"Ribba2012 %>% # Use rxode2   ini(prop.sd=0.05) %>% # change variability   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=3, omega=omega) %>%   plot(pt, q, qp, sim) # Plot it, plotting the variables of interest #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega' # note that sim is the simulated pstar since this is simulated from the # model with a nlmixr2 endpoint"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulation-with-uncertainty-in-all-the-parameters-by-matrices","dir":"Articles","previous_headings":"Simulating multiple subjects from a single event table","what":"Simulation with uncertainty in all the parameters (by matrices)","title":"rxode2 in a pipeline","text":"assume parameters came 95 subjects 8 observations apiece, degrees freedom omega matrix 95, degrees freedom sigma matrix 95*8=760 95 items informed omega matrix, 760 items informed sigma matrix.  Often simulations full covariance matrix fixed effect parameters. case, matrix, specified thetaMat. full covariance matrix, can information diagonal elements covariance matrix model paper. can converted follows: Now thetaMat represent uncertainty theta matrix, well pieces simulation. Typically can put information simulation thetaMat matrix. large variability theta easy sample negative rate constant, make sense. example: correct problems simply need use truncated multivariate normal specify reasonable ranges parameters. theta specified thetaLower thetaUpper. Similar parameters matrices: omegaLower, omegaUpper, sigmaLower sigmaUpper. may named vectors, one numeric value, numeric vector matching number parameters specified thetaMat matrix. case simulation simply modified thetaLower=0 make sure rates positive:","code":"Ribba2012 %>% # Use rxode2   ini(prop.sd = 0.05) %>%   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=3, nStud=3, omega=omega,           dfSub=760, dfObs=95) %>% # Solve the simulation   plot(pt, q, qp, sim) # Plot it, plotting the variables of interest #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega' rseVar <- function(est, rse){   return(est*rse/100)^2 }  thetaMat <- lotri(tpt0 ~ rseVar(7.13,25),                   tq0 ~ rseVar(41.2,7),                   tlambdap ~ rseVar(0.121, 16),                   tkqpp ~ rseVar(0.0031, 35),                   tdeltaqp ~ rseVar(0.00867, 21),                   tgamma ~ rseVar(0.729, 37),                   tkde ~ rseVar(0.24, 33)                   )  thetaMat #>            tpt0   tq0 tlambdap    tkqpp  tdeltaqp  tgamma   tkde #> tpt0     1.7825 0.000  0.00000 0.000000 0.0000000 0.00000 0.0000 #> tq0      0.0000 2.884  0.00000 0.000000 0.0000000 0.00000 0.0000 #> tlambdap 0.0000 0.000  0.01936 0.000000 0.0000000 0.00000 0.0000 #> tkqpp    0.0000 0.000  0.00000 0.001085 0.0000000 0.00000 0.0000 #> tdeltaqp 0.0000 0.000  0.00000 0.000000 0.0018207 0.00000 0.0000 #> tgamma   0.0000 0.000  0.00000 0.000000 0.0000000 0.26973 0.0000 #> tkde     0.0000 0.000  0.00000 0.000000 0.0000000 0.00000 0.0792 Ribba2012 %>% # Use rxode2 ini(prop.sd = 0.05) %>% et(time.units=\"months\") %>% # Pipe to a new event table et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months et(0, 250, by=0.5) %>% # Add some sampling times (not required) rxSolve(nSub=2, nStud=2, omega=omega, thetaMat=thetaMat, dfSub=760, dfObs=95) %>% # Solve the simulation plot(pt, q, qp, pstar) # Plot it, plotting the variables of interest  #> ℹ change initial estimate of `prop.sd` to `0.05` #> unhandled error message: EE:[lsoda] 70000 steps taken before reaching tout #> @(lsoda.c:750 #> Warning message: #> In rxSolve_(object, .ctl, .nms, .xtra, params, events, inits, setupOnly = .setupOnly) : #>  Some ID(s) could not solve the ODEs correctly; These values are replaced with NA. Ribba2012 %>% # Use rxode2   ini(prop.sd = 0.05) %>%   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=2, nStud=2, omega=omega,           thetaMat=thetaMat,           thetaLower=0, # Make sure the rates are reasonable           dfSub=760, dfObs=95) %>% # Solve the simulation   plot(pt, q, qp, sim) # Plot it, plotting the variables of interest #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega'"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"summarizing-the-simulation-output","dir":"Articles","previous_headings":"","what":"Summarizing the simulation output","title":"rxode2 in a pipeline","text":"easy use dplyr data.table perform summary simulations, rxode2 also provides ability confint function.","code":"## This takes a little more time; Most of the time is the summary ## time.  sim0 <- Ribba2012 %>% # Use rxode2   ini(prop.sd=0.05) %>%   et(time.units=\"months\") %>% # Pipe to a new event table   et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months   et(0, 250, by=0.5) %>% # Add some sampling times (not required)   rxSolve(nSub=10, nStud=10, omega=omega,           thetaMat=thetaMat,           thetaLower=0, # Make sure the rates are reasonable           dfSub=760, dfObs=95) %>% # Solve the simulation   confint(c(\"pt\",\"q\",\"qp\",\"sim\"),level=0.90); # Create Simulation intervals #> ℹ change initial estimate of `prop.sd` to `0.05` #> Warning: multi-subject simulation without without 'omega' #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done sim0 %>% plot() # Plot the simulation intervals"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-pipeline.html","id":"simulating-from-a-data-frame-of-parameters","dir":"Articles","previous_headings":"Summarizing the simulation output","what":"Simulating from a data-frame of parameters","title":"rxode2 in a pipeline","text":"simulation matrices can useful fast way simulate information, sometimes may want simulate complex scenarios. instance, may reason believe tkde needs tlambdap, therefore need simulated carefully. can generate data frame whatever way want. internal method simulating new parameters exported .","code":"library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union Ribba2012 <- Ribba2012() # Convert to classic rxode2 model with ini attached  r <- Ribba2012$simulationIniModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ pars <- rxInits(r) pars <- pars[regexpr(\"(prop|eta)\",names(pars)) == -1] print(pars) #>           k        tkde        tkpq       tkqpp    tlambdap      tgamma  #>    1.00e+02    2.40e-01    2.95e-02    3.10e-03    1.21e-01    7.29e-01  #>    tdeltaqp        tpt0         tq0 rxerr.pstar  #>    8.67e-03    7.13e+00    4.12e+01    1.00e+00 ## This is the exported method for simulation of Theta/Omega internally in rxode2 df <- rxSimThetaOmega(params=pars, omega=omega,dfSub=760,                       thetaMat=thetaMat, thetaLower=0, nSub=60,nStud=60) %>%     filter(tkde > tlambdap) %>% as_tibble() ## You could also simulate more and bind them together to a data frame. print(df) #> # A tibble: 2,040 × 17 #>        k  tkde   tkpq tkqpp tlambdap tgamma tdeltaqp  tpt0   tq0 rxerr.pstar #>    <dbl> <dbl>  <dbl> <dbl>    <dbl>  <dbl>    <dbl> <dbl> <dbl>       <dbl> #>  1   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  2   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  3   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  4   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  5   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  6   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  7   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  8   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #>  9   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #> 10   100 0.716 0.0295 0.529    0.161  0.887    0.412  7.82  41.9           1 #> # ℹ 2,030 more rows #> # ℹ 7 more variables: eta.pt0 <dbl>, eta.q0 <dbl>, eta.lambdap <dbl>, #> #   eta.kqp <dbl>, eta.kqpp <dbl>, eta.deltaqp <dbl>, eta.tkde <dbl> ## Quick check to make sure that all the parameters are OK. all(df$tkde>df$tlambdap) #> [1] TRUE sim1 <- r %>% # Use rxode2     et(time.units=\"months\") %>% # Pipe to a new event table     et(amt=1, time=50, until=58, ii=1.5) %>% # Add dosing every 1.5 months     et(0, 250, by=0.5) %>% # Add some sampling times (not required)     rxSolve(df) ## Note this information looses information about which ID is in a ## \"study\", so it summarizes the confidence intervals by dividing the ## subjects into sqrt(#subjects) subjects and then summarizes the ## confidence intervals sim2 <- sim1 %>% confint(c(\"pt\",\"q\",\"qp\",\"sim\"),level=0.90); # Create Simulation intervals #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done save(sim2, file = file.path(system.file(package = \"rxode2\"), \"pipeline-sim2.rds\"), version = 2) sim2 %>% plot()"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"setting-up-model-for-plotting-explanation","dir":"Articles","previous_headings":"Plotting in rxode2","what":"Setting up model for plotting explanation","title":"Plotting rxode2 models","text":"first step explain rxode2 model plots setup: rxode2 model event table solved object","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` ## Model from rxode2 tutorial m1 <- function() {   ini({     KA <- 2.94E-01     CL <- 1.86E+01     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250   })   model({     C2 <- centr / V2     C3 <- peri / V3     d/dt(depot) <- -KA * depot     f(depot) <- fdepot     dur(depot) <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3     d/dt(peri) <- Q * C2 - Q * C3     d/dt(eff) <- Kin - Kout * (1 - C2 / (EC50 + C2)) * eff     eff(0) <- 1   }) }  ev <- et(timeUnits = \"hr\") %>%   et(amt = 10000, ii = 12, until = 24) %>%   et(seq(0, 24, length.out = 100))  s <- rxSolve(m1, ev)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"default-rxode2-plot","dir":"Articles","previous_headings":"Plotting in rxode2","what":"Default rxode2 plot","title":"Plotting rxode2 models","text":"default plot rxode2 plot time versus state calculated lhs values:  Things note: First, plots actually ggplot2 plots can modified standard ggplot2 grammar graphics Also note plot method rxSolve objects can currently use arguments default plot method: log character string contains \"x\" x axis logarithmic, \"y\" y axis logarithmic \"xy\" \"yx\" axes logarithmic. available, use xgxr xgx_scale_x_log10() xgx_scale_y_log10() functions. xlab bit different simply supplying x-label ggplot2, retain unit information originally supplied event table. Also note, available, unit information used used meaningful ticks xgx_scale_x_time_units() ylab supplies y label following currently unsupported: . type b. xlim c. ylim d. main e. sub f. ann g. axes","code":"plot(s)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"subsetting-plots","dir":"Articles","previous_headings":"","what":"Subsetting plots","title":"Plotting rxode2 models","text":"interesting, often useful subset plots based values interested ; example, interesting plot eff C2. , simply specify interesting items plot function. instance:","code":"plot(s, C2, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"semi-log-plots","dir":"Articles","previous_headings":"","what":"Semi-log plots","title":"Plotting rxode2 models","text":"Semi-log plots PK concentrations common; simply need use log=\"y\"  uses xgx_scale_y_log10() available clearly show semi-log nature plot.","code":"plot(s, C2, log=\"y\", ylab=\"Concentration\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"plots-with-multi-subject-plots","dir":"Articles","previous_headings":"","what":"Plots with multi-subject plots","title":"Plotting rxode2 models","text":"multi-subject plots can easily plot applying principles . illustrate plot, lets expand event table include plot 4 subjects lognormal random variability subjects; complete, may plot method:  Notice colored individual labeled legend. interested concentration, produces similar plot:  Notice plot individual labeled attached id let know individual comes . done ggrepel available, otherwise legend retained.","code":"# Setup the new problem m2 <- function() {   ini({     KA <- 2.94E-01     TCL <- 1.86E+01     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200     ## Added modeled bioavaiblity, duration and rate     fdepot <- 1     durDepot <- 8     rateDepot <- 1250     eta.Cl ~ 0.4^2       })   model({     CL <- TCL * exp(eta.Cl)     C2 <- centr / V2     C3 <- peri / V3     d/dt(depot) <- -KA * depot     f(depot) <- fdepot     dur(depot) <- durDepot     rate(depot) <- rateDepot     d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3     d/dt(peri) <- Q * C2 - Q * C3     d/dt(eff) <- Kin - Kout * (1 - C2 / (EC50 + C2)) * eff     eff(0) <- 1   }) }  # Create the event table ev <- et(timeUnits = \"hr\") %>%   et(amt = 10000, until = units::set_units(3, days), ii = 12) %>% # loading doses   et(seq(0, 48, length.out = 200)) %>%   et(id = 1:4)  s <- rxSolve(m2, ev) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ plot(s, C2, eff) plot(s, C2, log=\"y\", ylab=\"Concentration\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-plot.html","id":"multi-subject-plots-with-large-number-of-subjects","dir":"Articles","previous_headings":"Plots with multi-subject plots","what":"Multi-subject plots with large number of subjects","title":"Plotting rxode2 models","text":"can get bit much many subjects solved (example lets simulate 100)  case, individuals put plot transparent grayscale plot pane. allows places subjects present darker. number individuals plots switch legend grayscale controlled changing options rxode2.spaghetti. ie. options(rxode2.spaghetti=7). 7 individuals default value plotting changes individual grayscale spaghetti plots. can also create confidence interval simulations confint: plot plot","code":"ev <- et(timeUnits = \"hr\") %>%   et(amt = 10000, until = units::set_units(3, days), ii = 12) %>% # loading doses   et(seq(0, 48, length.out = 200)) %>%   et(id = 1:100) # 100 subjects  s <- rxSolve(m2, ev)  plot(s, C2, log=\"y\", ylab=\"Concentration\") ## Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = ## minor_breaks, : log-10 transformation introduced infinite ## values. s2 <- confint(s, parm=\"C2\") ## ! in order to put confidence bands around the intervals, you need at least 2500 simulations ## summarizing data...done plot(s2)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-prior-data.html","id":"using-prior-data-for-solving","dir":"Articles","previous_headings":"","what":"Using prior data for solving","title":"Using Prior Data for ODE solving","text":"rxode2 can use single subject multiple subjects single event table solve ODEs. Additionally, rxode2 can use arbitrary data frame individualized events. example using nlmixr, use theo_sd data frame","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` library(nlmixr2data)  ## Load data from nlmixr d <- theo_sd  ## Create rxode2 model theo <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v   })  }  ## Create parameter dataset library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union parsDf <- tribble(   ~ eta.ka, ~ eta.cl, ~ eta.v,    0.105, -0.487, -0.080,   0.221, 0.144, 0.021,   0.368, 0.031, 0.058,  -0.277, -0.015, -0.007,  -0.046, -0.155, -0.142,  -0.382, 0.367, 0.203,  -0.791, 0.160, 0.047,  -0.181, 0.168, 0.096,   1.420, 0.042, 0.012,  -0.738, -0.391, -0.170,   0.790, 0.281, 0.146,  -0.527, -0.126, -0.198) %>%     mutate(tka = 0.451, tcl = 1.017, tv = 3.449)  ## Now solve the dataset solveData <- rxSolve(theo, parsDf, d) #> ℹ parameter labels from comments will be replaced by 'label()' #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ plot(solveData, cp) print(solveData) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 12 × 7 #>    id      tka   tcl    tv eta.ka eta.cl  eta.v #>    <fct> <dbl> <dbl> <dbl>  <dbl>  <dbl>  <dbl> #>  1 1     0.451  1.02  3.45  0.105 -0.487 -0.08  #>  2 2     0.451  1.02  3.45  0.221  0.144  0.021 #>  3 3     0.451  1.02  3.45  0.368  0.031  0.058 #>  4 4     0.451  1.02  3.45 -0.277 -0.015 -0.007 #>  5 5     0.451  1.02  3.45 -0.046 -0.155 -0.142 #>  6 6     0.451  1.02  3.45 -0.382  0.367  0.203 #>  7 7     0.451  1.02  3.45 -0.791  0.16   0.047 #>  8 8     0.451  1.02  3.45 -0.181  0.168  0.096 #>  9 9     0.451  1.02  3.45  1.42   0.042  0.012 #> 10 10    0.451  1.02  3.45 -0.738 -0.391 -0.17  #> 11 11    0.451  1.02  3.45  0.79   0.281  0.146 #> 12 12    0.451  1.02  3.45 -0.527 -0.126 -0.198 #> ── Initial Conditions ($inits): ── #>  depot center  #>      0      0  #> ── First part of data (object): ── #> # A tibble: 132 × 8 #>      id  time    ka    cl     v    cp   depot center #>   <int> <dbl> <dbl> <dbl> <dbl> <dbl>   <dbl>  <dbl> #> 1     1  0     1.74  1.70  29.0  0    320.        0  #> 2     1  0.25  1.74  1.70  29.0  3.86 207.      112. #> 3     1  0.57  1.74  1.70  29.0  6.81 118.      198. #> 4     1  1.12  1.74  1.70  29.0  9.06  45.4     263. #> 5     1  2.02  1.74  1.70  29.0  9.79   9.45    284. #> 6     1  3.82  1.74  1.70  29.0  9.10   0.410   264. #> # ℹ 126 more rows ## Of course the fasest way to solve if you don't care about the rxode2 extra parameters is  solveData <- rxSolve(theo, parsDf, d, returnType=\"data.frame\") #> ℹ parameter labels from comments will be replaced by 'label()' #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## solved data dplyr::as_tibble(solveData) #> # A tibble: 132 × 8 #>       id  time    ka    cl     v    cp         depot center #>    <int> <dbl> <dbl> <dbl> <dbl> <dbl>         <dbl>  <dbl> #>  1     1  0     1.74  1.70  29.0  0    320.              0  #>  2     1  0.25  1.74  1.70  29.0  3.86 207.            112. #>  3     1  0.57  1.74  1.70  29.0  6.81 118.            198. #>  4     1  1.12  1.74  1.70  29.0  9.06  45.4           263. #>  5     1  2.02  1.74  1.70  29.0  9.79   9.45          284. #>  6     1  3.82  1.74  1.70  29.0  9.10   0.410         264. #>  7     1  5.1   1.74  1.70  29.0  8.46   0.0440        246. #>  8     1  7.03  1.74  1.70  29.0  7.56   0.00152       219. #>  9     1  9.05  1.74  1.70  29.0  6.71   0.0000449     195. #> 10     1 12.1   1.74  1.70  29.0  5.61   0.000000212   163. #> # ℹ 122 more rows data.table::data.table(solveData) #>         id  time        ka       cl        v       cp        depot    center #>      <int> <num>     <num>    <num>    <num>    <num>        <num>     <num> #>   1:     1  0.00 1.7436838 1.698932 29.04946 0.000000 3.199920e+02   0.00000 #>   2:     1  0.25 1.7436838 1.698932 29.04946 3.861730 2.069289e+02 112.18117 #>   3:     1  0.57 1.7436838 1.698932 29.04946 6.805372 1.184389e+02 197.69240 #>   4:     1  1.12 1.7436838 1.698932 29.04946 9.058196 4.539354e+01 263.13572 #>   5:     1  2.02 1.7436838 1.698932 29.04946 9.791088 9.450361e+00 284.42585 #>  ---                                                                         #> 128:    12  5.07 0.9268162 2.437566 25.81614 8.442535 2.919432e+00 217.95370 #> 129:    12  7.07 0.9268162 2.437566 25.81614 7.074251 4.573778e-01 182.62989 #> 130:    12  9.03 0.9268162 2.437566 25.81614 5.892253 7.436222e-02 152.11524 #> 131:    12 12.05 0.9268162 2.437566 25.81614 4.432614 4.526550e-03 114.43300 #> 132:    12 24.15 0.9268162 2.437566 25.81614 1.414179 6.154773e-08  36.50865"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-rxUse.html","id":"using-pre-compiled-models-in-your-packages","dir":"Articles","previous_headings":"","what":"Using Pre-compiled models in your packages","title":"Compiled rxode2 models in Packages","text":"package like include pre-compiled rxode2 models package easy create package. simple make package rxPkg() command. : Add model package; can use package data idr package loads Add right package requirements DESCRIPTION file. want update describe package modify authors, license etc. Create skeleton model documentation files can add package documentation. case file idr-doc.R R directory Create configure configure.win script removes regenerates src directory based whatever version rxode2 compiled . modified plan compiled code, though suggested. can write R code package interacts rxode2 object can distribute shiny apps similar things package context. present can add models package rxUse(). Simply compile rxode2 model package add model rxUse() Now model idr model library. also create model-doc.R R directory can document model. can use devtools methods install/test model","code":"library(rxode2); ## Now Create a model idr <- rxode2({     C2 = centr/V2;     C3 = peri/V3;     d/dt(depot) =-KA*depot;     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  =                    Q*C2 - Q*C3;     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff; })  ## You can specify as many models as you want to add  rxPkg(idr, package=\"myPackage\"); ## Add the idr model to your package rxUse(model) devtools::load_all() # Load all the functions in the package devtools::document() # Create package documentation devtools::install() # Install package devtools::check() # Check the package devtools::build() # build the package so you can submit it to places like CRAN"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-rxUse.html","id":"using-models-in-a-already-present-package","dir":"Articles","previous_headings":"","what":"Using Models in a already present package","title":"Compiled rxode2 models in Packages","text":"illustrate, lets start blank package rxUse() : - Create rxode2 sources move package’s src/ directory. R source package, also finish directory library-init.c registers rxode2 models package use R. - Create stub R documentation models including package. able see R documentation loading package standard ? interface. still need : - Export least one function. function wish export, can add re-export rxode2 using roxygen follows: want use Suggests instead Depends package, way want export rxode2’s normal routines also need instruct R load model library models included model’s dll. done : R package rxode2 models intend add compiled sources (recommended), can add following configure scripts Depending check may need dummy autoconf script, want integrate sources Rcpp C/Fortan based packages, need include rxModels-compiled.h : - Add define macro compiledModelCall list registered .Call functions. - Register C interface allow model solving R_init0_rxModels_rxode2_models() (rxModels replaced package name). complete, can compile/document standard methods: load package new version rxode2, models recompiled used. However, want models recompiled recent version rxode2, simply need call rxUse() project directory followed standard methods install/create package. Note include rxode2 code required generate model regenerate rxode2 c-code src directory. rxode2 objects, summary show one way recreate model. example compiled models package can found rxModels repository.","code":"library(rxode2) library(usethis) pkgPath  <- file.path(rxTempDir(),\"MyRxModel\") create_package(pkgPath); use_gpl3_license(\"Matt\") use_package(\"rxode2\", \"LinkingTo\") use_package(\"rxode2\", \"Depends\") ##  library(rxode2) on load; Can use imports instead. use_roxygen_md() ##use_readme_md() library(rxode2); ## Now Create a model idr <- rxode2({     C2 = centr/V2;     C3 = peri/V3;     d/dt(depot) =-KA*depot;     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  =                    Q*C2 - Q*C3;     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff; });  rxUse(idr); ## Add the idr model to your package rxUse(); # Update the compiled rxode2 sources for all of your packages ##' @importFrom rxode2 rxode2 ##' @export rxode2::rxode2 ##' @importFrom rxode2 rxode2 ##' @export rxode2::rxode2  ##' @importFrom rxode2 et ##' @export rxode2::et  ##' @importFrom rxode2 etRep ##' @export rxode2::etRep  ##' @importFrom rxode2 etSeq ##' @export rxode2::etSeq  ##' @importFrom rxode2 as.et ##' @export rxode2::as.et  ##' @importFrom rxode2 eventTable ##' @export rxode2::eventTable  ##' @importFrom rxode2 add.dosing ##' @export rxode2::add.dosing  ##' @importFrom rxode2 add.sampling ##' @export rxode2::add.sampling  ##' @importFrom rxode2 rxSolve ##' @export rxode2::rxSolve  ##' @importFrom rxode2 rxControl ##' @export rxode2::rxControl  ##' @importFrom rxode2 rxClean ##' @export rxode2::rxClean  ##' @importFrom rxode2 rxUse ##' @export rxode2::rxUse  ##' @importFrom rxode2 rxShiny ##' @export rxode2::rxShiny  ##' @importFrom rxode2 genShinyApp.template ##' @export rxode2::genShinyApp.template  ##' @importFrom rxode2 cvPost ##' @export rxode2::cvPost  # This is actually from `magrittr` but allows less imports ##' @importFrom rxode2 %>% ##' @export rxode2::`%>%` # In this case `rxModels` is the package name ##' @useDynLib rxModels, .registration=TRUE #!/bin/sh # This should be used for both configure and configure.win echo \"unlink('src', recursive=TRUE);rxode2::rxUse()\" > build.R ${R_HOME}/bin/Rscript build.R rm build.R ## dummy autoconf script ## It is saved to configure.ac devtools::load_all() devtools::document() devtools::install() devtools::load_all() devtools::document() devtools::install()"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-shiny.html","id":"facilities-for-generating-r-shiny-applications","dir":"Articles","previous_headings":"","what":"Facilities for generating R shiny applications","title":"rxode2 and Shiny","text":"example creating R shiny application interactively explore responses various complex dosing regimens available http://qsp.engr.uga.edu:3838/rxode2/RegimenSimulator. Shiny applications like one may programmatically created experimental function genShinyApp.template(). application includes widgets varying dose, dosing regimen, dose cycle, number cycles. Click go Shiny App","code":"genShinyApp.template(appDir = \"shinyExample\", verbose=TRUE)  library(shiny) runApp(\"shinyExample\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-shiny.html","id":"exploring-parameter-fits-graphically-using-shiny","dir":"Articles","previous_headings":"","what":"Exploring parameter fits graphically using shiny","title":"rxode2 and Shiny","text":"rxode2 object can explored rxShiny(obj). rxShiny() also allow try new models see behave.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-variability-with-rxode2","dir":"Articles","previous_headings":"Population Simulations with rxode2","what":"Simulation of Variability with rxode2","title":"rxode2 Simulation","text":"pharmacometrics nonlinear-mixed effect modeling software (like nlmixr) characterizes -subject variability. subject variability can simulate new subjects. Assuming 2-compartment, indirect response model, can set create rxode2 model describing system :","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"setting-up-the-rxode2-model","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Setting up the rxode2 model","title":"rxode2 Simulation","text":"","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` set.seed(32) rxSetSeed(32)  mod <- function() {   ini({     KA <- 2.94E-01     TCl <- 1.86E+01     # between subject variability     eta.Cl ~ 0.4^2     V2 <- 4.02E+01     Q <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200   })   model({     C2 <- centr/V2     C3 <- peri/V3     CL <-  TCl*exp(eta.Cl) ## This is coded as a variable in the model     d/dt(depot) <- -KA*depot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulating","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Simulating","title":"rxode2 Simulation","text":"next step simulate create dosing regimen overall simulation: wish, can also add sampling times (though rxode2 can fill ): Note et takes similar arguments seq adding sampling times. methods adding sampling times events make complex dosing regimens (See event vignette). includes ways add variability sampling dosing times). complete can simulate using rxSolve routine: quickly look customize simulation use default plot routine. Since rxode2 object, create ggplot2 object can modify wish. extra parameter plot tells rxode2/R piece information interested plotting. case, interested looking derived parameter C2:","code":"ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, cmt=\"centr\") ev <- ev %>% et(0,48, length.out=100) sim  <- rxSolve(mod, ev, nSub=100) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"checking-the-simulation-with-plot","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Checking the simulation with plot","title":"rxode2 Simulation","text":"course additional parameter also state value, like eff:  even look two side--side:  stack patchwork","code":"library(ggplot2) ## The plots from rxode2 are ggplots so they can be modified with ## standard ggplot commands. plot(sim, C2, log=\"y\") +     ylab(\"Central Compartment\") ## They also takes many of the standard plot arguments; See ?plot plot(sim, eff, ylab=\"Effect\") plot(sim, C2, eff) library(patchwork) plot(sim, C2, log=\"y\") / plot(sim, eff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"processing-the-data-to-create-summary-plots","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Processing the data to create summary plots","title":"rxode2 Simulation","text":"Usually pharmacometric simulations enough simply simulate system. something easier digest, like look central extreme tendencies simulation. Since rxode2 solve object type data frame now straightforward perform calculations generate plots simulated data. can , 5th, 50th, 95th percentiles simulated data plotted.   Note can see parameters simulated example","code":"confint(sim, \"C2\", level=0.95) %>%     plot(ylab=\"Central Concentration\", log=\"y\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done confint(sim, \"eff\", level=0.95) %>%     plot(ylab=\"Effect\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done head(sim$param) #>   sim.id    KA  TCl   V2    Q  V3 Kin Kout EC50     eta.Cl #> 1      1 0.294 18.6 40.2 10.5 297   1    1  200  0.8067570 #> 2      2 0.294 18.6 40.2 10.5 297   1    1  200  0.0236961 #> 3      3 0.294 18.6 40.2 10.5 297   1    1  200 -0.3340697 #> 4      4 0.294 18.6 40.2 10.5 297   1    1  200 -0.1098197 #> 5      5 0.294 18.6 40.2 10.5 297   1    1  200 -0.2803752 #> 6      6 0.294 18.6 40.2 10.5 297   1    1  200 -0.2114090"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-unexplained-variability-sigma","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Simulation of unexplained variability (sigma)","title":"rxode2 Simulation","text":"addition conveniently simulating subject variability, can also easily simulate unexplained variability. One way create rxode2 model endpoints defined. Model piping can keeping model intact: can see dataset defined $multipleEndpoint: see endpoints defined eff C2: allows solve system: Since simulated model residual specification included multiple endpoint model, can summarize endpoint simply taking confidence interval \"sim\":  want subset specific endpoint, like eff can create confidence interval endpoint using specification sim.eff, endpoint name separated sim dot:","code":"mod2 <- mod %>%   model(eff ~ add(eff.sd), append=TRUE) %>%   model(C2 ~ prop(prop.sd), append=TRUE) %>%   ini(eff.sd=sqrt(0.1), prop.sd=sqrt(0.1)) #> ℹ add residual parameter `eff.sd` and set estimate to 1 #> ℹ add residual parameter `prop.sd` and set estimate to 1 #> ℹ change initial estimate of `eff.sd` to `0.316227766016838` #> ℹ change initial estimate of `prop.sd` to `0.316227766016838` mod2$multipleEndpoint #>   variable                cmt                dvid* #> 1  eff ~ … cmt='eff' or cmt=4 dvid='eff' or dvid=1 #> 2   C2 ~ …  cmt='C2' or cmt=5  dvid='C2' or dvid=2 ev <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, cmt=\"centr\") %>%   et(seq(0,48, length.out=100), cmt=\"eff\") %>%   et(seq(0,48, length.out=100), cmt=\"C2\") sim  <- rxSolve(mod2, ev, nSub=100) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ s <- confint(sim, \"sim\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(s) seff <- confint(sim, \"sim.eff\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(seff)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-individuals","dir":"Articles","previous_headings":"Population Simulations with rxode2 > Simulation of Variability with rxode2","what":"Simulation of Individuals","title":"rxode2 Simulation","text":"Sometimes may want match dosing observations individuals clinical trial. create data.frame using rxode2 event specification well ID column indicate individual. rxode2 event vignette talks datasets created.","code":"library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union ev1 <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=10000, cmt=2) %>%     et(0,48,length.out=10)  ev2 <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%     et(amt=5000, cmt=2) %>%     et(0,48,length.out=8)  dat <- rbind(data.frame(ID=1, ev1$get.EventTable()),              data.frame(ID=2, ev2$get.EventTable()))   ## Note the number of subject is not needed since it is determined by the data sim  <- rxSolve(mod, dat) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ sim %>% select(id, time, eff, C2) #>    id          time      eff          C2 #> 1   1  0.000000 [h] 1.000000 248.7562189 #> 2   1  5.333333 [h] 1.168348  19.1658271 #> 3   1 10.666667 [h] 1.040652   7.1764936 #> 4   1 16.000000 [h] 1.030849   6.0167596 #> 5   1 21.333333 [h] 1.027742   5.4446478 #> 6   1 26.666667 [h] 1.025204   4.9484199 #> 7   1 32.000000 [h] 1.022911   4.4984868 #> 8   1 37.333333 [h] 1.020828   4.0895169 #> 9   1 42.666667 [h] 1.018933   3.7177314 #> 10  1 48.000000 [h] 1.017211   3.3797461 #> 11  2  0.000000 [h] 1.000000 124.3781095 #> 12  2  6.857143 [h] 1.016301   1.9290018 #> 13  2 13.714286 [h] 1.006648   1.2918078 #> 14  2 20.571429 [h] 1.005617   1.0964981 #> 15  2 27.428571 [h] 1.004772   0.9317005 #> 16  2 34.285714 [h] 1.004055   0.7916739 #> 17  2 41.142857 [h] 1.003446   0.6726918 #> 18  2 48.000000 [h] 1.002928   0.5715923"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-of-clinical-trials","dir":"Articles","previous_headings":"","what":"Simulation of Clinical Trials","title":"rxode2 Simulation","text":"either using simple single event table, data clinical trial described , complete clinical trial simulation can performed. Typically clinical trial simulations want account uncertainty fixed parameter estimates, even uncertainty subject variability well unexplained variability. rxode2 allows account uncertainties simulating multiple virtual “studies,” specified parameter nStud. studies samples realization fixed effect parameters covariance matrices subject variability(omega) unexplained variabilities (sigma). Depending information models, strategies simulating realization omega sigma matrices. first strategy occurs either standard errors standard deviations (related parameters), modeled correlation model simulating . case suggested strategy use inverse Wishart (parameterized scale conjugate prior)/scaled inverse chi distribution. approach uses single parameter inform variability covariance matrix sampled (degrees freedom). second strategy occurs standard errors variance/standard deviation modeled correlations covariance matrix. approach perform separate simulations standard deviations correlation matrix. First simulate variance/standard deviation components thetaMat multivariate normal simulation. simulation transformation standard deviations, correlation matrix simulated using degrees freedom covariance matrix. Combining simulated standard deviation simulated correlation matrix give simulated covariance matrix. smaller dimension covariance matrices (dimension < 10x10) recommended use lkj distribution simulate correlation matrix. higher dimension covariance matrices suggested use inverse wishart distribution (transformed correlation matrix) simulations. covariance/variance prior simulated rxode2s cvPost() function.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulation-from-inverse-wishart-correlations","dir":"Articles","previous_headings":"Simulation of Clinical Trials","what":"Simulation from inverse Wishart correlations","title":"rxode2 Simulation","text":"example simulation :  wish can see omega sigma used virtual study accessing solved data object $omega.list $sigma.list: can also see parameter realizations $params data frame.","code":"## Creating covariance matrix tmp <- matrix(rnorm(8^2), 8, 8) tMat <- tcrossprod(tmp, tmp) / (8 ^ 2) dimnames(tMat) <- list(NULL, names(mod2$theta)[1:8])  sim  <- rxSolve(mod2, ev, nSub=100, thetaMat=tMat, nStud=10,                 dfSub=10, dfObs=100)  s <-sim %>% confint(\"sim\") #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(s) head(sim$omegaList) #> [[1]] #>           eta.Cl #> eta.Cl 0.1676778 #>  #> [[2]] #>           eta.Cl #> eta.Cl 0.2917085 #>  #> [[3]] #>           eta.Cl #> eta.Cl 0.1776813 #>  #> [[4]] #>           eta.Cl #> eta.Cl 0.1578682 #>  #> [[5]] #>           eta.Cl #> eta.Cl 0.1845614 #>  #> [[6]] #>           eta.Cl #> eta.Cl 0.3282268 head(sim$sigmaList) #> [[1]] #>            rxerr.eff   rxerr.C2 #> rxerr.eff 1.12416983 0.04197039 #> rxerr.C2  0.04197039 0.97293971 #>  #> [[2]] #>             rxerr.eff    rxerr.C2 #> rxerr.eff  0.84311199 -0.06277998 #> rxerr.C2  -0.06277998  1.22140938 #>  #> [[3]] #>           rxerr.eff  rxerr.C2 #> rxerr.eff 0.9834771 0.1060251 #> rxerr.C2  0.1060251 1.0024751 #>  #> [[4]] #>            rxerr.eff   rxerr.C2 #> rxerr.eff 1.25556975 0.07690868 #> rxerr.C2  0.07690868 0.90991261 #>  #> [[5]] #>           rxerr.eff  rxerr.C2 #> rxerr.eff  1.116261 -0.184748 #> rxerr.C2  -0.184748  1.320288 #>  #> [[6]] #>            rxerr.eff   rxerr.C2 #> rxerr.eff 0.93539238 0.07270049 #> rxerr.C2  0.07270049 0.98648424"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulate-using-variancestandard-deviation-standard-errors","dir":"Articles","previous_headings":"Simulation of Clinical Trials","what":"Simulate using variance/standard deviation standard errors","title":"rxode2 Simulation","text":"Lets assume wish simulate nonmem run included xpose First setup model; Since taking nonmem like use free-form style classic rxode2 model start classic model: Next input estimated parameters: also covariances; , easiest way create named covariance matrix use lotri():","code":"rx1 <- rxode2({   cl <- tcl*(1+crcl.cl*(CLCR-65)) * exp(eta.cl)   v <- tv * WT * exp(eta.v)   ka <- tka * exp(eta.ka)   ipred <- linCmt()   obs <- ipred * (1 + prop.sd) + add.sd  }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ theta <- c(tcl=2.63E+01, tv=1.35E+00, tka=4.20E+00, tlag=2.08E-01,            prop.sd=2.05E-01, add.sd=1.06E-02, crcl.cl=7.17E-03,            ## Note that since we are using the separation strategy the ETA variances are here too            eta.cl=7.30E-02,  eta.v=3.80E-02, eta.ka=1.91E+00) thetaMat <- lotri(     tcl + tv + tka + tlag + prop.sd + add.sd + crcl.cl + eta.cl + eta.v + eta.ka ~         c(7.95E-01,           2.05E-02, 1.92E-03,           7.22E-02, -8.30E-03, 6.55E-01,           -3.45E-03, -6.42E-05, 3.22E-03, 2.47E-04,           8.71E-04, 2.53E-04, -4.71E-03, -5.79E-05, 5.04E-04,           6.30E-04, -3.17E-06, -6.52E-04, -1.53E-05, -3.14E-05, 1.34E-05,           -3.30E-04, 5.46E-06, -3.15E-04, 2.46E-06, 3.15E-06, -1.58E-06, 2.88E-06,           -1.29E-03, -7.97E-05, 1.68E-03, -2.75E-05, -8.26E-05, 1.13E-05, -1.66E-06, 1.58E-04,           -1.23E-03, -1.27E-05, -1.33E-03, -1.47E-05, -1.03E-04, 1.02E-05, 1.67E-06, 6.68E-05, 1.56E-04,           7.69E-02, -7.23E-03, 3.74E-01, 1.79E-03, -2.85E-03, 1.18E-05, -2.54E-04, 1.61E-03, -9.03E-04, 3.12E-01))  evw <- et(amount.units=\"mg\", time.units=\"hours\") %>%     et(amt=100) %>%     ## For this problem we will simulate with sampling windows     et(list(c(0, 0.5),        c(0.5, 1),        c(1, 3),        c(3, 6),        c(6, 12))) %>%     et(id=1:1000)  ## From the run we know that: ##   total number of observations is: 476 ##    Total number of individuals:     74 sim  <- rxSolve(rx1, theta, evw,  nSub=100, nStud=10,                 thetaMat=thetaMat,                 ## Match boundaries of problem                 thetaLower=0,                  sigma=c(\"prop.sd\", \"add.sd\"), ## Sigmas are standard deviations                 sigmaXform=\"identity\", # default sigma xform=\"identity\"                 omega=c(\"eta.cl\", \"eta.v\", \"eta.ka\"), ## etas are variances                 omegaXform=\"variance\", # default omega xform=\"variance\"                 iCov=data.frame(WT=rnorm(1000, 70, 15), CLCR=rnorm(1000, 65, 25)),                 dfSub=74, dfObs=476); #> ℹ thetaMat has too many items, ignored: 'tlag' print(sim) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 10,000 × 9 #>    sim.id id      tcl crcl.cl  eta.cl    tv   eta.v   tka  eta.ka #>     <int> <fct> <dbl>   <dbl>   <dbl> <dbl>   <dbl> <dbl>   <dbl> #>  1      1 1      27.0   0.828 -0.884   1.73  0.531   5.38  0.114  #>  2      1 2      27.0   0.828  1.11    1.73 -0.906   5.38 -0.738  #>  3      1 3      27.0   0.828 -0.690   1.73  0.280   5.38 -0.365  #>  4      1 4      27.0   0.828  0.0751  1.73 -1.17    5.38 -0.575  #>  5      1 5      27.0   0.828 -0.299   1.73 -0.121   5.38 -0.601  #>  6      1 6      27.0   0.828 -0.638   1.73 -0.281   5.38 -0.309  #>  7      1 7      27.0   0.828  1.05    1.73 -0.0495  5.38  0.567  #>  8      1 8      27.0   0.828 -0.371   1.73 -0.372   5.38 -0.903  #>  9      1 9      27.0   0.828 -0.206   1.73  1.97    5.38  0.0203 #> 10      1 10     27.0   0.828  0.754   1.73  0.166   5.38 -1.73   #> # ℹ 9,990 more rows #> ── Initial Conditions ($inits): ── #> named numeric(0) #>  #> Simulation with uncertainty in: #> • parameters ($thetaMat for changes) #> • omega matrix ($omegaList) #> • sigma matrix ($sigmaList) #>  #> ── First part of data (object): ── #> # A tibble: 50,000 × 10 #>   sim.id    id   time     cl     v    ka  ipred    obs    WT  CLCR #>    <int> <int>    [h]  <dbl> <dbl> <dbl>  <dbl>  <dbl> <dbl> <dbl> #> 1      1     1 0.0155   50.4 183.   6.03 0.0487  3.56   62.2  69.3 #> 2      1     1 0.749    50.4 183.   6.03 0.460  -1.59   62.2  69.3 #> 3      1     1 1.02     50.4 183.   6.03 0.431   1.15   62.2  69.3 #> 4      1     1 3.41     50.4 183.   6.03 0.224   1.92   62.2  69.3 #> 5      1     1 7.81     50.4 183.   6.03 0.0666  0.731  62.2  69.3 #> 6      1     2 0.0833 2816.   24.9  2.57 0.0754 -1.20   35.7 105.  #> # ℹ 49,994 more rows ## Notice that the simulation time-points change for the individual  ## If you want the same sampling time-points you can do that as well: evw <- et(amount.units=\"mg\", time.units=\"hours\") %>%     et(amt=100) %>%     et(0, 24, length.out=50) %>%     et(id=1:100)  sim  <- rxSolve(rx1, theta, evw,  nSub=100, nStud=10,                 thetaMat=thetaMat,                 ## Match boundaries of problem                 thetaLower=0,                  sigma=c(\"prop.sd\", \"add.sd\"), ## Sigmas are standard deviations                 sigmaXform=\"identity\", # default sigma xform=\"identity\"                 omega=c(\"eta.cl\", \"eta.v\", \"eta.ka\"), ## etas are variances                 omegaXform=\"variance\", # default omega xform=\"variance\"                 iCov=data.frame(WT=rnorm(100, 70, 15), CLCR=rnorm(100, 65, 25)),                 dfSub=74, dfObs=476,                 resample=TRUE) #> ℹ thetaMat has too many items, ignored: 'tlag' s <-sim %>% confint(c(\"ipred\")) #> summarizing data... #> done plot(s)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-sim-var.html","id":"simulate-without-uncertainty-in-omega-or-sigma-parameters","dir":"Articles","previous_headings":"Simulation of Clinical Trials","what":"Simulate without uncertainty in omega or sigma parameters","title":"rxode2 Simulation","text":"wish sample prior distributions either omega sigma matrices, can turn feature specifying simVariability = FALSE option solving:  Note since realizations omega sigma simulated, $omegaList $sigmaList return NULL.","code":"sim  <- rxSolve(mod2, ev, nSub=100, thetaMat=tMat, nStud=10,                 simVariability=FALSE)  s <-sim %>% confint(c(\"centr\", \"eff\")) #> ! in order to put confidence bands around the intervals, you need at least 2500 simulations #> summarizing data...done plot(s)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-single-subject.html","id":"single-subject-solving","dir":"Articles","previous_headings":"","what":"Single Subject solving","title":"Single Subject ODE solving -- differences from multiple subject","text":"Originally, rxode2 created solve ODEs one individual. single system without changes individual parameters. course still supported, classic examples found rxode2 intro. article discusses differences multiple subject single subject solving. three differences: Single solving solve ID parallel Single solving lacks id column parameters($params) well actual dataset. Single solving allows parameter exploration easier parameter can modified. multiple subject solves, make sure update individual parameter. first obvious difference speed; multiple subjects can run subject ID parallel. information examples speed gains multiple subject solving see Speeding rxode2 vignette. next difference amount information output final data. Taking 2 compartment indirect response model originally tutorial: Now simple solve  better see differences single solve, can solve 2 individuals  observing two solves, can see: multiple subject solve contains id column data frame data frame parameters subject. last feature obvious, modifying individual parameters. single subject data, can modify rxode2 data frame changing initial conditions parameter values part data frame, described rxode2 Data Frames. multiple subject solving, feature still works, requires care supplying individual’s parameter value, otherwise may change solve drop parameter key individuals.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod1 <- function() {   ini({     KA=2.94E-01     CL=1.86E+01     V2=4.02E+01     Q=1.05E+01     V3=2.97E+02     Kin=1     Kout=1     EC50=200   })   model({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) =-KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  =                    Q*C2 - Q*C3     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1   }) }    et <- et(amount.units='mg', time.units='hours') %>%     et(dose=10000, addl=9, ii=12) %>%     et(amt=20000, nbr.doses=5, start.time=120, dosing.interval=24) %>%     et(0:240) # sampling x <- rxSolve(mod1, et) x #> ── Solved rxode2 object ── #> ── Parameters (x$params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions (x$inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows print(x) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions ($inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows plot(x, C2, eff) x2 <- rxSolve(mod1, et %>% et(id=1:2), params=data.frame(CL=c(18.6, 7.6))) print(x2) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 2 × 9 #>   id       KA    CL    V2     Q    V3   Kin  Kout  EC50 #>   <fct> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 1     0.294  18.6  40.2  10.5   297     1     1   200 #> 2 2     0.294   7.6  40.2  10.5   297     1     1   200 #> ── Initial Conditions ($inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 482 × 8 #>      id time    C2    C3  depot centr  peri   eff #>   <int>  [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1     1    0   0   0     10000     0     0   1    #> 2     1    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3     1    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4     1    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5     1    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6     1    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 476 more rows plot(x2, C2, eff)"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"a-note-about-the-speed-of-the-functional-form-for-rxode2","dir":"Articles","previous_headings":"","what":"A note about the speed of the functional form for rxode2","title":"Speeding up rxode2","text":"functional form benefit supported nlmixr2 therefore one interface solving estimating, takes computation time get underlying “classic” simulation code. models form : can also specify end-points simulation/estimation just like nlmixr2: every solve, compile (cached compile) underlying model. wish speed process can use two underlying rxode2 classic models. takes two steps: Parsing/evaluating model Creating simulation model first step can done rxode2(mod1) mod1() (second model ). second step create underlying “classic” rxode2 model, can done two different methods:$simulationModel $simulationIniModel. $simulationModel provide simulation code without initial conditions pre-pended, $simulationIniModel pre-pend values. endpoints specified, simulation code endpoint also output. can see differences : wish speed multiple simualtions rxode2 functions, need pre-calculate care steps : functions can act like normal ui model solved. can convert back UI .rxUi() function .function() needed. increase speed multiple simulations model use lower level simulation model (ie $simulationModel $simulationIniModel depending need)","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod1 <- function() {   ini({     KA   <- 0.3     CL   <- 7     V2   <- 40     Q    <- 10     V3   <- 300     Kin  <- 0.2     Kout <- 0.2     EC50 <- 8   })   model({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     d/dt(eff) = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1   }) } mod2f <- function() {   ini({     TKA   <- 0.3     TCL   <- 7     TV2   <- 40     TQ    <- 10     TV3   <- 300     TKin  <- 0.2     TKout <- 0.2     TEC50 <- 8     eta.cl + eta.v ~ c(0.09,                        0.08, 0.25)     c2.prop.sd <- 0.1     eff.add.sd <- 0.1   })   model({     KA <- TKA     CL <- TCL*exp(eta.cl)     V2  <- TV2*exp(eta.v)     Q   <- TQ     V3  <- TV3     Kin  <- TKin     Kout <- TKout     EC50 <- TEC50     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     d/dt(eff) = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1     C2 ~ prop(c2.prop.sd)     eff ~ add(eff.add.sd)   }) } mod1 <- mod1() mod2f <- rxode2(mod2f) summary(mod1$simulationModel) #> rxode2 2.1.3 model named rx_cb1fa822f5309328f76b6ae6134387bc model (✔ ready).  #> DLL: /tmp/RtmpPXkf2x/rxode2/rx_cb1fa822f5309328f76b6ae6134387bc__.rxd/rx_cb1fa822f5309328f76b6ae6134387bc_.so #> NULL #>  #> Calculated Variables: #> [1] \"C2\" \"C3\" #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(KA, CL, V2, Q, V3, Kin, Kout, EC50) #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #> }) summary(mod1$simulationIniModel) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_f8478a8ad7138c076241f845fcf72d6c model (✔ ready).  #> DLL: /tmp/RtmpPXkf2x/rxode2/rx_f8478a8ad7138c076241f845fcf72d6c__.rxd/rx_f8478a8ad7138c076241f845fcf72d6c_.so #> NULL #>  #> Calculated Variables: #> [1] \"C2\" \"C3\" #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(KA, CL, V2, Q, V3, Kin, Kout, EC50) #>     KA = 0.3 #>     CL = 7 #>     V2 = 40 #>     Q = 10 #>     V3 = 300 #>     Kin = 0.2 #>     Kout = 0.2 #>     EC50 = 8 #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #> }) summary(mod2f$simulationModel) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_881bb97ef0dcf691ed7a5d08770c2985 model (✔ ready).  #> DLL: /tmp/RtmpPXkf2x/rxode2/rx_881bb97ef0dcf691ed7a5d08770c2985__.rxd/rx_881bb97ef0dcf691ed7a5d08770c2985_.so #> NULL #>  #> Calculated Variables: #>  [1] \"KA\"       \"CL\"       \"V2\"       \"Q\"        \"V3\"       \"Kin\"      #>  [7] \"Kout\"     \"EC50\"     \"C2\"       \"C3\"       \"ipredSim\" \"sim\"      #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(TKA, TCL, TV2, TQ, TV3, TKin, TKout, TEC50, c2.prop.sd,  #>         eff.add.sd, eta.cl, eta.v) #>     KA = TKA #>     CL = TCL * exp(eta.cl) #>     V2 = TV2 * exp(eta.v) #>     Q = TQ #>     V3 = TV3 #>     Kin = TKin #>     Kout = TKout #>     EC50 = TEC50 #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #>     if (CMT == 5) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ C2 #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (rx_pred_f_ * c2.prop.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.C2, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     if (CMT == 4) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ eff #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (eff.add.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.eff, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     cmt(C2) #>     dvid(5, 4) #> }) summary(mod2f$simulationIniModel) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_a19506ad281e7f985f0efe169504bf02 model (✔ ready).  #> DLL: /tmp/RtmpPXkf2x/rxode2/rx_a19506ad281e7f985f0efe169504bf02__.rxd/rx_a19506ad281e7f985f0efe169504bf02_.so #> NULL #>  #> Calculated Variables: #>  [1] \"KA\"       \"CL\"       \"V2\"       \"Q\"        \"V3\"       \"Kin\"      #>  [7] \"Kout\"     \"EC50\"     \"C2\"       \"C3\"       \"ipredSim\" \"sim\"      #> ── rxode2 Model Syntax ── #> rxode2({ #>     param(TKA, TCL, TV2, TQ, TV3, TKin, TKout, TEC50, c2.prop.sd,  #>         eff.add.sd, eta.cl, eta.v) #>     rxerr.C2 = 1 #>     rxerr.eff = 1 #>     TKA = 0.3 #>     TCL = 7 #>     TV2 = 40 #>     TQ = 10 #>     TV3 = 300 #>     TKin = 0.2 #>     TKout = 0.2 #>     TEC50 = 8 #>     c2.prop.sd = 0.1 #>     eff.add.sd = 0.1 #>     eta.cl = 0 #>     eta.v = 0 #>     KA = TKA #>     CL = TCL * exp(eta.cl) #>     V2 = TV2 * exp(eta.v) #>     Q = TQ #>     V3 = TV3 #>     Kin = TKin #>     Kout = TKout #>     EC50 = TEC50 #>     C2 = centr/V2 #>     C3 = peri/V3 #>     d/dt(depot) = -KA * depot #>     d/dt(centr) = KA * depot - CL * C2 - Q * C2 + Q * C3 #>     d/dt(peri) = Q * C2 - Q * C3 #>     d/dt(eff) = Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     eff(0) = 1 #>     if (CMT == 5) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ C2 #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (rx_pred_f_ * c2.prop.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.C2, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     if (CMT == 4) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ eff #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (eff.add.sd)^2 #>         ipredSim = rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim = rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.eff, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     cmt(C2) #>     dvid(5, 4) #> }) mod1 <- mod1$simulationModel  mod2 <- mod2f$simulationModel"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"increasing-rxode2-speed-by-multi-subject-parallel-solving","dir":"Articles","previous_headings":"","what":"Increasing rxode2 speed by multi-subject parallel solving","title":"Speeding up rxode2","text":"Using classic rxode2 model specification (can convert functional/ui model style) continue discussion rxode2 speed enhancements. rxode2 originally developed ODE solver allowed ODE solve single subject. flexibility still supported. original code rxode2 tutorial :","code":"library(rxode2)  library(microbenchmark) library(ggplot2)  mod1 <- rxode2({     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     d/dt(eff) = Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0) = 1 }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## Create an event table  ev <- et() %>%     et(amt=10000, addl=9,ii=12) %>%     et(time=120, amt=20000, addl=4, ii=24) %>%     et(0:240) ## Add Sampling  nsub <- 100 # 100 sub-problems sigma <- matrix(c(0.09,0.08,0.08,0.25),2,2) # IIV covariance matrix mv <- rxRmvn(n=nsub, rep(0,2), sigma) # Sample from covariance matrix CL <- 7*exp(mv[,1]) V2 <- 40*exp(mv[,2]) params.all <- cbind(KA=0.3, CL=CL, V2=V2, Q=10, V3=300,                     Kin=0.2, Kout=0.2, EC50=8)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"for-loop","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"For Loop","title":"Speeding up rxode2","text":"slowest way code use loop. example enclose function compare timing.","code":"runFor <- function(){     res <- NULL     for (i in 1:nsub) {         params <- params.all[i,]         x <- mod1$solve(params, ev)         ##Store results for effect compartment         res <- cbind(res, x[, \"eff\"])     }     return(res) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"running-with-apply","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Running with apply","title":"Speeding up rxode2","text":"general R, apply types functions perform better loop, tutorial also suggests speed enhancement","code":"runSapply <- function(){     res <- apply(params.all, 1, function(theta)         mod1$run(theta, ev)[, \"eff\"]) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"run-using-a-single-threaded-solve","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Run using a single-threaded solve","title":"Speeding up rxode2","text":"can also rxode2 solve subject simultaneously without collecting results R, using single threaded solve. data output slightly different , still gives information:","code":"runSingleThread <- function(){   solve(mod1, params.all, ev, cores=1)[,c(\"sim.id\", \"time\", \"eff\")] }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"run-a-2-threaded-solve","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Run a 2 threaded solve","title":"Speeding up rxode2","text":"rxode2 supports multi-threaded solves, another option 2 threads (called cores solve options, can see options rxControl() rxSolve()).","code":"run2Thread <- function(){   solve(mod1, params.all, ev, cores=2)[,c(\"sim.id\", \"time\", \"eff\")] }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"compare-the-times-between-all-the-methods","dir":"Articles","previous_headings":"Increasing rxode2 speed by multi-subject parallel solving","what":"Compare the times between all the methods","title":"Speeding up rxode2","text":"Now moment truth, timings:  clear largest jump performance using solve method providing parameters rxode2 solve without looping subject either sapply. number cores/threads applied solve also plays role solving. can explore number threads following code:  can suite spot speed vs number cores. system type (mac, linux, windows /processor), complexity ODE solving number subjects may affect arbitrary number threads. 4 threads good number use without prior knowledge systems days least 4 threads (2 processors 4 threads).","code":"bench <- microbenchmark(runFor(), runSapply(), runSingleThread(),run2Thread()) print(bench) #> Unit: milliseconds #>               expr       min        lq      mean    median        uq       max #>           runFor() 263.75899 267.88252 276.54611 269.91252 273.66833 391.77390 #>        runSapply() 262.40141 268.24234 276.38984 270.40109 272.75447 434.24999 #>  runSingleThread()  27.45217  27.61226  27.94130  27.69391  27.96168  37.54928 #>       run2Thread()  16.17052  16.32538  17.99204  16.45466  16.85388  27.76178 #>  neval #>    100 #>    100 #>    100 #>    100 autoplot(bench) runThread <- function(n){     solve(mod1, params.all, ev, cores=n)[,c(\"sim.id\", \"time\", \"eff\")] }  bench <- eval(parse(text=sprintf(\"microbenchmark(%s)\",                                      paste(paste0(\"runThread(\", seq(1, 2 * rxCores()),\")\"),                                            collapse=\",\")))) print(bench) #> Unit: milliseconds #>          expr      min       lq     mean   median       uq      max neval #>  runThread(1) 27.45592 27.64329 29.31929 28.28103 29.36378 48.39516   100 #>  runThread(2) 16.17788 16.39189 18.17140 17.07855 17.39019 28.93695   100 #>  runThread(3) 17.42292 19.35593 20.04593 19.99469 20.21361 27.37969   100 #>  runThread(4) 15.80004 16.03155 18.46875 16.62245 16.95347 40.09879   100 autoplot(bench)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"increasing-speed-with-compiler-options","dir":"Articles","previous_headings":"","what":"Increasing speed with compiler options","title":"Speeding up rxode2","text":"One way allows faster ODE solving make approximations make math operators like exp() faster technically accurate enough follow IEEE standard math functions values (implications cover ). optimizations opt-Julia since compile everything session, CRAN conservative approach since individuals compile R function running . Still, rxode2 models can compiled option without disturbing CRAN policies. key set option. example:  Note compiler settings can tricky setup system wide Makevars may interact setting. example use ccache compile may produced options since cached options. example, github runner (generates page), advantage \"fast\" compile. However, development laptop minimal speed increase. probably check using . disabled default since minimum increase speed.","code":"# Using the first example subset to PK mod2f <- function() {   ini({     TKA   <- 0.3     TCL   <- 7     TV2   <- 40     TQ    <- 10     TV3   <- 300     TKin  <- 0.2     TKout <- 0.2     TEC50 <- 8     eta.cl + eta.v ~ c(0.09,                        0.08, 0.25)     c2.prop.sd <- 0.1   })   model({     KA <- TKA     CL <- TCL*exp(eta.cl)     V2  <- TV2*exp(eta.v)     Q   <- TQ     V3  <- TV3     Kin  <- TKin     Kout <- TKout     EC50 <- TEC50     C2 = centr/V2     C3 = peri/V3     d/dt(depot) = -KA*depot     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri) = Q*C2 - Q*C3     C2 ~ prop(c2.prop.sd)   }) }  mod2f <- mod2f()  mod2s <- mod2f$simulationIniModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ev  <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(0:240) # Add sampling  bench1 <- microbenchmark(standardCompile=rxSolve(mod2s, ev, nSub=1000))  # Now clear the cache of models so we can change the compile options for the same model rxClean()  # Use withr to preserve the options withr::with_options(list(rxode2.compile.O=\"fast\"), {   mod2s <- mod2f$simulationIniModel }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ bench2 <- microbenchmark(fastCompile=rxSolve(mod2s, ev, nSub=1000))  bench <- rbind(bench1, bench2)  print(bench) #> Unit: milliseconds #>             expr      min       lq     mean   median       uq      max neval #>  standardCompile 173.3969 181.5005 218.6953 188.4703 239.3224 375.4217   100 #>      fastCompile 174.1782 179.6236 214.7402 182.9570 230.5892 352.4374   100 autoplot(bench)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"a-real-life-example","dir":"Articles","previous_headings":"","what":"A real life example","title":"Speeding up rxode2","text":"cBefore parallel solving implemented, fastest way run rxode2 lapply. Rik Schoemaker created data-set nlmixr comparisons, reduced run faster automatic building pkgdown website. applying new parallel solving concepts can simply run simulation less code faster: can see striking time difference two methods; things keep mind: rxode2 use thread-safe sitmo threefry routines simulation eta values. Therefore results expected different (also random samples taken different order different) prior simulation run R 3.5, different random number generator results simulation different actual nlmixr comparison using slower simulation. speed comparison used data.table. rxode2 uses data.table internally (available) try speed sorting, different installations data.table installed. can force rxode2 use order() sorting using forderForceBase(TRUE). case little difference two, though examples data.table’s presence leads speed increase (less likely lead slowdown).","code":"library(rxode2) library(data.table) #Define the rxode2 model   ode1 <- \"   d/dt(abs)    = -KA*abs;   d/dt(centr)  =  KA*abs-(CL/V)*centr;   C2=centr/V;   \"  #Create the rxode2 simulation object mod1 <- rxode2(model = ode1) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #Population parameter values on log-scale   paramsl <- c(CL = log(4),                V = log(70),                KA = log(1)) #make 10,000 subjects to sample from:   nsubg <- 300 # subjects per dose   doses <- c(10, 30, 60, 120)   nsub <- nsubg * length(doses) #IIV of 30% for each parameter   omega <- diag(c(0.09, 0.09, 0.09))# IIV covariance matrix   sigma <- 0.2 #Sample from the multivariate normal   set.seed(98176247)   rxSetSeed(98176247)   library(MASS)   mv <-     mvrnorm(nsub, rep(0, dim(omega)[1]), omega) # Sample from covariance matrix #Combine population parameters with IIV   params.all <-     data.table(       \"ID\" = seq(1:nsub),       \"CL\" = exp(paramsl['CL'] + mv[, 1]),       \"V\" = exp(paramsl['V'] + mv[, 2]),       \"KA\" = exp(paramsl['KA'] + mv[, 3])     ) #set the doses (looping through the 4 doses) params.all[, AMT := rep(100 * doses,nsubg)]  Startlapply <- Sys.time()  #Run the simulations using lapply for speed   s = lapply(1:nsub, function(i) { #selects the parameters associated with the subject to be simulated     params <- params.all[i] #creates an eventTable with 7 doses every 24 hours     ev <- eventTable()     ev$add.dosing(       dose = params$AMT,       nbr.doses = 1,       dosing.to = 1,       rate = NULL,       start.time = 0     ) #generates 4 random samples in a 24 hour period     ev$add.sampling(c(0, sort(round(sample(runif(600, 0, 1440), 4) / 60, 2)))) #runs the rxode2 simulation     x <- as.data.table(mod1$run(params, ev)) #merges the parameters and ID number to the simulation output     x[, names(params) := params]   })  #runs the entire sequence of 100 subjects and binds the results to the object res   res = as.data.table(do.call(\"rbind\", s))  Stoplapply <- Sys.time()  print(Stoplapply - Startlapply) #> Time difference of 12.08713 secs rx <- rxode2({     CL =  log(4)     V = log(70)     KA = log(1)     CL = exp(CL + eta.CL)     V = exp(V + eta.V)     KA = exp(KA + eta.KA)     d/dt(abs)    = -KA*abs;     d/dt(centr)  =  KA*abs-(CL/V)*centr;     C2=centr/V; }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ omega <- lotri(eta.CL ~ 0.09,                eta.V ~ 0.09,                eta.KA ~ 0.09)  doses <- c(10, 30, 60, 120)   startParallel <- Sys.time() ev <- do.call(\"rbind\",         lapply(seq_along(doses), function(i){             et() %>%                 et(amt=doses[i]) %>% # Add single dose                 et(0) %>% # Add 0 observation                 ## Generate 4 samples in 24 hour period                 et(lapply(1:4, function(...){c(0, 24)})) %>%                 et(id=seq(1, nsubg) + (i - 1) * nsubg) %>%                 ## Convert to data frame to skip sorting the data                 ## When binding the data together                 as.data.frame         })) ## To better compare, use the same output, that is data.table res <- rxSolve(rx, ev, omega=omega, returnType=\"data.table\") endParallel <- Sys.time() print(endParallel - startParallel) #> Time difference of 0.1223662 secs"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"want-more-ways-to-run-multi-subject-simulations","dir":"Articles","previous_headings":"A real life example","what":"Want more ways to run multi-subject simulations","title":"Speeding up rxode2","text":"version since tutorial even ways run multi-subject simulations, including adding variability sampling dosing times et() (see rxode2 events information), ability supply omega sigma matrix well adding thetaMat R simulate uncertainty omega, sigma theta matrices; see rxode2 simulation vignette.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-speed.html","id":"session-information","dir":"Articles","previous_headings":"A real life example","what":"Session Information","title":"Speeding up rxode2","text":"session information:","code":"sessionInfo() #> R version 4.4.0 (2024-04-24) #> Platform: x86_64-pc-linux-gnu #> Running under: Ubuntu 22.04.4 LTS #>  #> Matrix products: default #> BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3  #> LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.20.so;  LAPACK version 3.10.0 #>  #> locale: #>  [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8        #>  [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8    #>  [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C           #> [10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C    #>  #> time zone: UTC #> tzcode source: system (glibc) #>  #> attached base packages: #> [1] stats     graphics  grDevices utils     datasets  methods   base      #>  #> other attached packages: #> [1] MASS_7.3-60.2         data.table_1.15.4     ggplot2_3.5.1         #> [4] microbenchmark_1.4.10 rxode2_2.1.3          #>  #> loaded via a namespace (and not attached): #>  [1] gtable_0.3.5         xfun_0.44            bslib_0.7.0          #>  [4] htmlwidgets_1.6.4    RApiSerialize_0.1.3  lattice_0.22-6       #>  [7] vctrs_0.6.5          tools_4.4.0          generics_0.1.3       #> [10] tibble_3.2.1         symengine_0.2.6      fansi_1.0.6          #> [13] highr_0.11           pkgconfig_2.0.3      checkmate_2.3.1      #> [16] desc_1.4.3           RcppParallel_5.1.7   lifecycle_1.0.4      #> [19] compiler_4.4.0       farver_2.1.2         textshaping_0.4.0    #> [22] munsell_0.5.1        qs_0.26.3            htmltools_0.5.8.1    #> [25] sys_3.4.2            sass_0.4.9           yaml_2.3.8           #> [28] lazyeval_0.2.2       pillar_1.9.0         pkgdown_2.0.9        #> [31] crayon_1.5.2         jquerylib_0.1.4      cachem_1.1.0         #> [34] nlme_3.1-164         tidyselect_1.2.1     digest_0.6.35        #> [37] lotri_0.4.4          dplyr_1.1.4          purrr_1.0.2          #> [40] labeling_0.4.3       rxode2ll_2.0.11.9000 fastmap_1.2.0        #> [43] grid_4.4.0           colorspace_2.1-0     rxode2parse_2.0.19   #> [46] cli_3.6.2            dparser_1.3.1-11     magrittr_2.0.3       #> [49] utf8_1.2.4           withr_3.0.0          scales_1.3.0         #> [52] backports_1.5.0      rmarkdown_2.27       ragg_1.3.2           #> [55] stringfish_0.16.0    memoise_2.0.1        evaluate_0.23        #> [58] knitr_1.46           rex_1.2.1            rxode2et_2.0.13      #> [61] rxode2random_2.1.1   PreciseSums_0.6      rlang_1.1.3          #> [64] Rcpp_1.0.12          glue_1.7.0           jsonlite_1.8.8       #> [67] R6_2.5.1             systemfonts_1.1.0    fs_1.6.4             #> [70] units_0.8-5"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-stiff.html","id":"stiff-odes-with-jacobian-specification","dir":"Articles","previous_headings":"","what":"Stiff ODEs with Jacobian Specification","title":"rxode2 Jacobian specification and Stiff Systems","text":"Occasionally, may come across stiff differential equation, differential equation numerically unstable small variations parameters cause different solutions ODEs. One way tackle choose stiff-solver, hybrid stiff solver (like default LSODA). Typically enough. However exact Jacobian solutions may increase stability ODE. (Note Jacobian derivative ODE specification respect variable). rxode2 can specify Jacobian df(state)/dy(variable)= statement. classic ODE stiff properties various conditions Van der Pol differential equations. rxode2 can specified following: stiff mu=1, mu=1000 stiff system easy enough , bit tedious. rxode2 setup appropriately, can use computer algebra system sympy calculate Jacobian automatically. done rxode2 option calcJac option:","code":"library(rxode2) ## rxode2 2.1.3 using 2 threads (see ?getRxThreads) ##   no cache: create with `rxCreateCache()` Vtpol2 <- function() {   ini({     mu <- 1 ## nonstiff; 10 moderately stiff; 1000 stiff   })   model({     d/dt(y)       <- dy     d/dt(dy)      <- mu*(1-y^2)*dy - y     ## Jacobian     df(y)/dy(dy)  <- 1     df(dy)/dy(y)  <- -2*dy*mu*y - 1     df(dy)/dy(dy) <- mu*(1-y^2)     ## Initial conditions     y(0)          <- 2     dy(0)         <- 0   }) }  et <- et(0, 10, length.out=200) %>%   et(amt=0)  s1 <- Vtpol2 %>%  solve(et, method=\"lsoda\") ## i parameter labels from comments will be replaced by 'label()' ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(s1) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ## mu  ##  1  ## -- Initial Conditions ($inits): -- ##  y dy  ##  2  0  ## -- First part of data (object): -- ## # A tibble: 200 x 3 ##     time     y      dy ##    <dbl> <dbl>   <dbl> ## 1 0       2     0      ## 2 0.0503  2.00 -0.0933 ## 3 0.101   1.99 -0.173  ## 4 0.151   1.98 -0.242  ## 5 0.201   1.97 -0.302  ## 6 0.251   1.95 -0.353  ## # i 194 more rows s2 <- Vtpol2 %>%  solve(c(mu=1000), et) ## i parameter labels from comments will be replaced by 'label()' ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(s2) ## -- Solved rxode2 object -- ## -- Parameters ($params): -- ##   mu  ## 1000  ## -- Initial Conditions ($inits): -- ##  y dy  ##  2  0  ## -- First part of data (object): -- ## # A tibble: 200 x 3 ##     time     y        dy ##    <dbl> <dbl>     <dbl> ## 1 0       2     0        ## 2 0.0503  2.00 -0.000667 ## 3 0.101   2.00 -0.000667 ## 4 0.151   2.00 -0.000667 ## 5 0.201   2.00 -0.000667 ## 6 0.251   2.00 -0.000667 ## # i 194 more rows Vtpol <- function() {   ini({     mu <- 1 ## nonstiff; 10 moderately stiff; 1000 stiff   })   model({     d/dt(y)       <- dy     d/dt(dy)      <- mu*(1-y^2)*dy - y     y(0)          <- 2     dy(0)         <- 0   }) }  Vtpol <- Vtpol()  # you can also use $symengineModelPrune if there is if/else blocks # that need to be converted: Vtpol <- rxode2(Vtpol$symengineModelNoPrune,  calcJac=TRUE) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## > pruning branches (`if`/`else`)... ## v done ## > loading into symengine environment... ## v done ## > calculate jacobian ## [====|====|====|====|====|====|====|====|====|====] 0:00:00 ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ summary(Vtpol) ## rxode2 2.1.3 model named rx_ebf53954066c14e2215d9e98d501de9f model (ready).  ## DLL: /tmp/RtmpPXkf2x/rxode2/rx_ebf53954066c14e2215d9e98d501de9f__.rxd/rx_ebf53954066c14e2215d9e98d501de9f_.so ## NULL ## -- rxode2 Model Syntax -- ## rxode2({ ##     cmt(y) ##     cmt(dy) ##     d/dt(y) = dy ##     d/dt(dy) = -y + mu * dy * (1 - Rx_pow_di(y, 2)) ##     y(0) = 2 ##     dy(0) = 0 ##     df(y)/dy(y) = 0 ##     df(dy)/dy(y) = -1 - 2 * y * mu * dy ##     df(y)/dy(dy) = 1 ##     df(dy)/dy(dy) = mu * (1 - Rx_pow_di(y, 2)) ##     df(y)/dy(mu) = 0 ##     df(dy)/dy(mu) = dy * (1 - Rx_pow_di(y, 2)) ## })"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"introduction","dir":"Articles","previous_headings":"","what":"Introduction","title":"rxode2 ODE solving syntax","text":"briefly describes syntax used define models rxode2 translate R-callable compiled code. also describes communication variables R rxode2 modeling specification.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"creating-rxode2-models","dir":"Articles","previous_headings":"","what":"Creating rxode2 models","title":"rxode2 ODE solving syntax","text":"ODE-based model specification may coded inside four places: Inside rxode2({}) block statements: Inside rxode2(\"\") string statement: file name loaded rxode2: model function can parsed rxode2: block statement, character string text file internal rxode2 compilation manager translates ODE system C, compiles loads R session. call rxode2 produces object class rxode2 consists list-like structure (environment) various member functions. last type model (model function), call rxode2 creates parsed rxode2 ui can translated rxode2 compilation model. type function required nlmixr2 estimation can extended modified model piping. reason focused documentation.","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` mod <- rxode2({   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; }) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ mod <- rxode2(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \") #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ writeLines(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \", \"modelFile.rxode2\") mod <- rxode2(filename='modelFile.rxode2') unlink(\"modelFile.rxode2\") mod <- function() {   model({     # simple assignment     C2 <- centr/V2      # time-derivative assignment     d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;   }) }  mod <- rxode2(mod) # or simply mod() if the model is at the end of the function  # These model functions often have residual components and initial # (`ini({})`) conditions attached as well. For example the # theophylline model can be written as:  one.compartment <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  # after parsing the model mod <- one.compartment() mod$simulationModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_0577787cc435460f1d0396f10467e345 model (✔ ready).  #> x$state: depot, center #> x$stateExtra: cp #> x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp #> x$lhs: ka, cl, v, cp, ipredSim, sim # or  mod$simulationIniModel #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_b3707fe11efa354614e6788ba8c23ed4 model (✔ ready).  #> x$state: depot, center #> x$stateExtra: cp #> x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp #> x$lhs: ka, cl, v, cp, ipredSim, sim"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"syntax","dir":"Articles","previous_headings":"","what":"Syntax","title":"rxode2 ODE solving syntax","text":"basic model specification consists one statements optionally terminated semi-colons ; optional comments (comments delimited # end--line). block statements set statements delimited curly braces, { ... }. Statements can either assignments, conditional /else /else, loops (can exited break), special statements, printing statements (debugging/testing). Assignment statements can : simple assignments, left hand identifier (.e., variable) special time-derivative assignments, left hand specifies change amount corresponding state variable (compartment) respect time e.g., d/dt(depot): special initial-condition assignments left hand specifies compartment initial condition specified, e.g. depot(0) = 0 special model event changes including bioavailability (f(depot)=1), lag time (alag(depot)=0), modeled rate (rate(depot)=2) modeled duration (dur(depot)=2). example model features event specification modeled infusions rxode2 data specification found rxode2 events vignette. special change point syntax, model times. model times specified mtime(var)=time special Jacobian-derivative assignments, left hand specifies change compartment ode respect variable. example, d/dt(y) = dy, Jacobian compartment can specified df(y)/dy(dy) = 1. may advantage obtaining solution specifying Jacobian stiff ODE systems. However, stiff systems tried LSODA, actually slightly slowed solving. Note assignment can done =, <- ~. assigning ~ operator, simple assignments time-derivative assignments output. Note rxode2 model functions assignment ~ can also overloaded residual distribution specification. Special statements can : Compartment declaration statements, can change default dosing compartment assumed compartment number(s) well add extra compartment names end (useful multiple-endpoint nlmixr models); specified cmt(compartmentName) Parameter declaration statements, can make sure input parameters certain order instead ordering parameters order parsed. useful keeping parameter order using 2 different ODE models. specified param(par1, par2,...) example model shown : Expressions assignment statements can numeric logical. Numeric expressions can include following numeric operators +, -, *, /, ^ mathematical functions defined C R math libraries (e.g., fabs, exp, log, sin, abs). may also access R’s functions R math libraries, like lgammafn log gamma function. rxode2 syntax case-sensitive, .e., ABC different abc, Abc, ABc, etc.","code":"# simple assignment    C2 <- centr/V2     # time-derivative assignment    d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"identifiers","dir":"Articles","previous_headings":"Syntax","what":"Identifiers","title":"rxode2 ODE solving syntax","text":"Like R, Identifiers (variable names) may consist one alphanumeric, underscore _ period . characters, first character digit underscore _. Identifiers model specification can refer : State variables dynamic system (e.g., compartments pharmacokinetics model). Implied input variable, t (time), tlast (last time point), podo (oral dose, undocumented case absorption transit models). Special constants like pi R’s predefined constants. Model parameters (e.g., ka rate absorption, CL clearance, etc.) Others, created assignments part model specification; referred LHS (left-hand side) variable. Currently, rxode2 modeling language recognizes system state variables “parameters”, thus, values need passed R ODE model (e.g., age) either passed params argument integrator function rxSolve() supplied event data-set. certain variable names rxode2 event tables. avoid confusion, following event table-related items assigned, used state can accessed rxode2 code: cmt dvid addl ss rate id However following variables used model specification: evid ii Sometimes rxode2 generates variables fed back rxode2. Similarly, nlmixr2 generates variables used nlmixr estimation simulation. variables start either rx nlmixr prefixes. avoid problems, suggested use variables starting either rx nlmixr prefixes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"logical-operators","dir":"Articles","previous_headings":"","what":"Logical Operators","title":"rxode2 ODE solving syntax","text":"Logical operators support standard R operators ==, != >= <= > <. Like R can () () statements, ifelse() expressions. Additionally can standard assignment. instance, following valid: Notice can also use character expressions comparisons. convenience comes cost since character comparisons slower numeric expressions. Unlike R, .numeric .integer logical statements needed, cause syntax error try use function.","code":"cov1 = covm*(sexf == \"female\") + covm*(sexf != \"female\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"supported-functions","dir":"Articles","previous_headings":"","what":"Supported functions","title":"rxode2 ODE solving syntax","text":"supported functions rxode2 can seen rxSupportedFuns(). brief description built-functions following table: Note lag(cmt) = equivalent alag(cmt) = = lag(wt)","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"reserved-keywords","dir":"Articles","previous_headings":"","what":"Reserved keywords","title":"rxode2 ODE solving syntax","text":"reserved keywords rxode2 model. following table: Note rxFlag always output 11 calc_lhs since final variables calculated, though can tweak test certain parts rxode2 using flag.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"residual-functions-when-using-rxode2-functions","dir":"Articles","previous_headings":"","what":"Residual functions when using rxode2 functions","title":"rxode2 ODE solving syntax","text":"addition ~ hiding output certain types output, also used specify residual output endpoint input rxode2 model function (includes residual model({}) block). specifications form: Indicating variable var variable represents individual central tendencies model also represents compartment specification data-set. can also change compartment name using | syntax, : case var represents central tendency cmt represents compartment dvid specification.","code":"var ~ add(add.sd) var ~ add(add.sd) | cmt"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"transformations","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Transformations","title":"rxode2 ODE solving syntax","text":"normal related distributions, can apply transformation sides using keywords/functions apply transformations. default likelihood transformations calculated untransformed scale. bounded variables like logit-normal probit-normal low high values defaulted 0 1 missing. models wish proportional model one transformation can replace standard deviation NA allow transformations, lnorm(), probitNorm() logitNorm() can combined variance stabilizing yeoJohnson() transformation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"normal-and-t-related-distributions","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Normal and t-related distributions","title":"rxode2 ODE solving syntax","text":"normal t-related distributions, wanted keep ability use skewed distributions additive proportional t/cauchy-space, distributions specified differently comparison supported distributions within nlmixr2: Note normal t-related distributions nlmixr2 calculate cwres npde normal assumption help assess goodness fit model. Also note +dnorm() mostly testing purposes slow estimation procedure nlmixr2. suggest adding (except explicit testing). multiple endpoint models mix non-normal normal distributions, whole problem shifted log-likelihood method estimation nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"notes-on-additive-proportional-models","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Notes on additive + proportional models","title":"rxode2 ODE solving syntax","text":"two different ways specify additive proportional models, call combined1 combined2, way Monolix calls two distributions (avoid software differences naming). first, combined1, assumes additive proportional differences standard deviation scale, : y=f+(+b* f^c)*err second, combined2, assumes additive proportional differences combined variance scale: y=f+\\[sqrt(^2+b^2 *f^(2c))\\]*err default nlmixr2/rxode2 otherwise specified combined2 since mirrors adding 2 normal distributions statistics add variances (standard deviations). However, combined1 can describe data possibly even better combined2 possible options rxode2/nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"distributions-of-known-likelihoods","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Distributions of known likelihoods","title":"rxode2 ODE solving syntax","text":"residuals related normal, t-distribution cauchy, often residual specification form: compartment specification left handed side specification. generalized likelihood can specify:","code":"cmt ~ dbeta(alpha, beta) ll(cmt) ~ llik specification"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"ordinal-likelihoods","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"Ordinal likelihoods","title":"rxode2 ODE solving syntax","text":"Finally, ordinal likelihoods/simulations can specified 2 ways. first : err represents compartment p0 probability specific category: model ensure sum p values less 1. Additionally can write arbitrary number categories ordinal model described . seems little p0 probability category 1 sometimes scores non-whole numbers. can modeled follows: numeric categories specified explicitly, probabilities remain :","code":"err ~ c(p0, p1, p2) err ~ c(p0=0, p1=1, p2=2, 3)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"general-table-of-supported-residual-distributions","dir":"Articles","previous_headings":"Residual functions when using rxode2 functions","what":"General table of supported residual distributions","title":"rxode2 ODE solving syntax","text":"general supported following table (available rxode2::rxResidualError)","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"bugs-andor-deficiencies","dir":"Articles","previous_headings":"","what":"Bugs and/or deficiencies","title":"rxode2 ODE solving syntax","text":"modulo operator %% currently unsupported.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"note","dir":"Articles","previous_headings":"","what":"Note","title":"rxode2 ODE solving syntax","text":"ODE specification mini-language parsed help open source tool , Plevyak (2015).","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"example-model","dir":"Articles","previous_headings":"","what":"Example model","title":"rxode2 ODE solving syntax","text":"commented example quickly show capabilities rxode2 syntax.","code":""},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"example","dir":"Articles","previous_headings":"","what":"Example","title":"rxode2 ODE solving syntax","text":"","code":"f <- function() {   ini({        })   model({     # An rxode2 model specification (this line is a comment).      if(comed==0) {  # concomitant medication (con-med)?       F <- 1.0     # full bioavailability w.o. con-med     } else {       F <- 0.80    # 20% reduced bioavailability     }      C2 <- centr/V2  # concentration in the central compartment     C3 <- peri/V3   # concentration in the peripheral compartment      # ODE describing the PK and PD      d/dt(depot) <- -KA*depot     d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                      Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff     eff(0)      <- 1   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-syntax.html","id":"interface-and-data-handling-between-r-and-the-generated-c-code","dir":"Articles","previous_headings":"","what":"Interface and data handling between R and the generated C code","title":"rxode2 ODE solving syntax","text":"Users specify variables dynamic system’s state variables via d/dt(identifier) operator part model specification, model parameters via params= argument rxode2 solve() method: values variables pre-specified time points saved model fitting/integration returned part fitted values (see function et(), define set time points capture values variables) returned part modeling output. ODE specification mini-language parsed help open source tool DParser, Plevyak (2015).","code":"m1 <- rxode2(model = ode, modName = \"m1\")  # model parameters -- a named vector is required theta <-     c(KA=0.29, CL=18.6, V2=40.2, Q=10.5, V3=297, Kin=1, Kout=1, EC50=200)  # state variables and their amounts at time 0 (the use of names is # encouraged, but not required) inits <- c(depot=0, centr=0, peri=0, eff=1)  # qd1 is an eventTable specification with a set of dosing and sampling  # records (code not shown here)  solve(theta, event = qd1, inits = inits)"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-tutorials.html","id":"interactive-tutorials","dir":"Articles","previous_headings":"","what":"Interactive tutorials","title":"Interactive Tutorials","text":"rxode2 comes interactive tutorials can run R session; built Rstudio 1.3 can run R session :","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` ## rxode2 Syntax learnr::run_tutorial(\"rxode2-00-syntax\", \"rxode2\")  ## rxode2 event tables   learnr::run_tutorial(\"rxode2-00-events\", \"rxode2\")"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"creating-a-2-compartment-model-in-rxode2","dir":"Articles","previous_headings":"","what":"Creating a 2-compartment model in rxode2","title":"Weight Based Dosing in rxode2","text":"","code":"library(rxode2) #> rxode2 2.1.3 using 2 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()` ## Note the time covariate is not included in the simulation m1 <- function() {   model({     CL ~ (1-0.2*SEX)*(0.807+0.00514*(CRCL-91.2))*exp(eta.cl)     V1 ~ 4.8*exp(eta.v1)     Q ~ (3.46+0.0593*(WT-75.1))*exp(eta.q);     V2 ~ 1.93*(3.13+0.0458*(WT-75.1))*exp(eta.v2)     A1 ~ centr;     A2 ~ peri;     d/dt(centr) <- - A1*(CL/V1 + Q/V1) + A2*Q/V2;     d/dt(peri) <- A1*Q/V1 - A2*Q/V2;     DV = centr / V1 * (1 + prop.err)   }) }"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"simulating-covariates","dir":"Articles","previous_headings":"","what":"Simulating Covariates","title":"Weight Based Dosing in rxode2","text":"simulation correlates age, sex, weight. Since using weight based dosing, needs simulated first","code":"set.seed(42) rxSetSeed(42) library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union nsub=30 # Simulate Weight based on age and gender AGE<-round(runif(nsub,min=18,max=70)) SEX<-round(runif(nsub,min=0,max=1)) HTm<-round(rnorm(nsub,176.3,0.17*sqrt(4482)),digits=1) HTf<-round(rnorm(nsub,162.2,0.16*sqrt(4857)),digits=1) WTm<-round(exp(3.28+1.92*log(HTm/100))*exp(rnorm(nsub,0,0.14)),digits=1) WTf<-round(exp(3.49+1.45*log(HTf/100))*exp(rnorm(nsub,0,0.17)),digits=1) WT<-ifelse(SEX==1,WTf,WTm) CRCL<-round(runif(nsub,30,140)) ## id is in lower case to match the event table cov.df <- tibble(id=seq_along(AGE), AGE=AGE, SEX=SEX, WT=WT, CRCL=CRCL) print(cov.df) #> # A tibble: 30 x 5 #>       id   AGE   SEX    WT  CRCL #>    <int> <dbl> <dbl> <dbl> <dbl> #>  1     1    66     1  49.4    83 #>  2     2    67     1  52.5    79 #>  3     3    33     0  97.9    37 #>  4     4    61     1  63.8    66 #>  5     5    51     0  71.8   127 #>  6     6    45     1  69.6   132 #>  7     7    56     0  61      73 #>  8     8    25     0  57.7    47 #>  9     9    52     1  58.7    65 #> 10    10    55     1  73.1    64 #> # i 20 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"creating-weight-based-event-table","dir":"Articles","previous_headings":"","what":"Creating weight based event table","title":"Weight Based Dosing in rxode2","text":"","code":"s<-c(0,0.25,0.5,0.75,1,1.5,seq(2,24,by=1)) s <- lapply(s, function(x){.x <- 0.1 * x; c(x - .x, x + .x)})  e <- et() %>%     ## Specify the id and weight based dosing from covariate data.frame     ## This requires rxode2 XXX      et(id=cov.df$id, amt=6*cov.df$WT, rate=6 * cov.df$WT) %>%     ## Sampling is added for each ID     et(s) %>%     as.data.frame %>%     ## Merge the event table with the covarite information     merge(cov.df, by=\"id\") %>%     as_tibble   e #> # A tibble: 900 x 12 #>       id    low  time   high cmt         amt  rate  evid   AGE   SEX    WT  CRCL #>    <int>  <dbl> <dbl>  <dbl> <chr>     <dbl> <dbl> <int> <dbl> <dbl> <dbl> <dbl> #>  1     1  0     0      0     (obs)       NA    NA      0    66     1  49.4    83 #>  2     1 NA     0     NA     (default)  296.  296.     1    66     1  49.4    83 #>  3     1  0.225 0.246  0.275 (obs)       NA    NA      0    66     1  49.4    83 #>  4     1  0.45  0.516  0.55  (obs)       NA    NA      0    66     1  49.4    83 #>  5     1  0.675 0.729  0.825 (obs)       NA    NA      0    66     1  49.4    83 #>  6     1  0.9   0.921  1.1   (obs)       NA    NA      0    66     1  49.4    83 #>  7     1  1.35  1.42   1.65  (obs)       NA    NA      0    66     1  49.4    83 #>  8     1  1.8   1.82   2.2   (obs)       NA    NA      0    66     1  49.4    83 #>  9     1  2.7   2.97   3.3   (obs)       NA    NA      0    66     1  49.4    83 #> 10     1  3.6   3.87   4.4   (obs)       NA    NA      0    66     1  49.4    83 #> # i 890 more rows"},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"solving-daptomycin-simulation","dir":"Articles","previous_headings":"","what":"Solving Daptomycin simulation","title":"Weight Based Dosing in rxode2","text":"","code":"data <- rxSolve(m1, e,              ## Lotri uses lower-triangular matrix rep. for named matrix              omega=lotri(eta.cl ~ .306,                           eta.q ~0.0652,                          eta.v1 ~.567,                          eta.v2 ~ .191),              sigma=lotri(prop.err ~ 0.15),              addDosing = TRUE, addCov = TRUE) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ print(data) #> -- Solved rxode2 object -- #> -- Parameters ($params): -- #> # A tibble: 30 x 5 #>    id    eta.cl  eta.v1      eta.q  eta.v2 #>    <fct>  <dbl>   <dbl>      <dbl>   <dbl> #>  1 1      0.343 -1.25   -0.0936    -0.253  #>  2 2      0.444 -0.0581  0.187      0.233  #>  3 3      0.142  0.203   0.0617     1.03   #>  4 4      0.274 -0.733  -0.0000116 -0.321  #>  5 5      0.505 -0.104  -0.406     -0.518  #>  6 6      0.472  1.48   -0.537      0.210  #>  7 7      0.156 -0.0295  0.112     -0.226  #>  8 8     -0.747  0.519  -0.347      0.476  #>  9 9      0.326 -0.373   0.340      0.470  #> 10 10     0.428 -1.28    0.131     -0.0240 #> # i 20 more rows #> -- Initial Conditions ($inits): -- #> centr  peri  #>     0     0  #> -- First part of data (object): -- #> # A tibble: 900 x 12 #>      id  evid   cmt   amt  rate  time    DV centr  peri   SEX    WT  CRCL #>   <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1     1     1     1  296.  296. 0       0     0    0        1  49.4    83 #> 2     1     0    NA   NA    NA  0       0     0    0        1  49.4    83 #> 3     1     0    NA   NA    NA  0.246  41.3  58.7  9.45     1  49.4    83 #> 4     1     0    NA   NA    NA  0.516  31.0 100.  34.0      1  49.4    83 #> 5     1     0    NA   NA    NA  0.729  72.2 124.  58.9      1  49.4    83 #> 6     1     0    NA   NA    NA  0.921  82.0 140.  83.1      1  49.4    83 #> # i 894 more rows plot(data, log=\"y\") #> Warning in transformation$transform(x): NaNs produced #> Warning in ggplot2::scale_y_log10(..., breaks = breaks, minor_breaks = #> minor_breaks, : log-10 transformation introduced infinite #> values."},{"path":"https://nlmixr2.github.io/rxode2/articles/rxode2-wt.html","id":"daptomycin-reference","dir":"Articles","previous_headings":"","what":"Daptomycin Reference","title":"Weight Based Dosing in rxode2","text":"weight-based simulation adapted Daptomycin article : Dvorchik B, Arbeit RD, Chung J, Liu S, Knebel W, Kastrissios H. Population pharmacokinetics daptomycin. Antimicrob Agents Che mother 2004; 48: 2799-2807. doi:(10.1128/AAC.48.8.2799-2807.2004)[https://dx.doi.org/10.1128%2FAAC.48.8.2799-2807.2004] simulation example made available work Sherwin Sy modifications Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors","text":"Matthew L. Fidler (core team/developer/manual) Melissa Hallow (tutorial writer) Wenping Wang (core team/developer)","code":""},{"path":"https://nlmixr2.github.io/rxode2/authors.html","id":"contributors","dir":"","previous_headings":"","what":"Contributors","title":"Authors","text":"Zufar Mulyukov – Wrote initial version rxShiny() modifications Matthew Fidler Alan Hindmarsh – Lsoda author Awad H. Al-Mohy – Al-Mohy matrix exponential author Ernst Hairer – dop853 author Gerhard Wanner – dop853 author Goro Fuji – Timsort author Hadley Wickham – Author original findLhs RxODE, also original author .s3register (used permission anyone, changed Matthew Fidler) Jack Dongarra – LApack author Linda Petzold – LSODA Martin Maechler – expm author, used routines inductive linearization Morwenn – Timsort author Nicholas J. Higham – Author Al-mohy matrix exponential Roger B. Sidje – expokit matrix exponential author Simon Frost – thread safe C implementation liblsoda Kevin Ushey – Original author fast factor, modified Matthew Filder Yu Feng – thread safe liblsoda Cleve Moler – LApack author David Cooley – Author fast_factor modified now used RxODE quickly create factors IDs without sorting like R Drew Schmidt – Drew Schmidt author edits exponential matrix utility taken R package expm Matt Dowle & Arun Srinivasan – data.table authors openMp framework based upon","code":""},{"path":"https://nlmixr2.github.io/rxode2/authors.html","id":"rxode-acknowledgments","dir":"","previous_headings":"","what":"RxODE acknowledgments:","title":"Authors","text":"Sherwin Sy – Weight based dosing example Justin Wilkins – Documentation updates, logo testing Emma Schwager – R IJK distribution author J Coligne – dop853 fortran author Bill Denney – Documentation updates, manual minor bug fixes Tim Waterhouse – Fixed one bug mac working directories Richard Upton – Helped solving ADVAN linCmt() solutions Dirk Eddelbuettel – Made fixes Rcpp changes require R strict headers Ross Ihaka – R author Robert Gentleman – R author R core team – R authors","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"overview","dir":"","previous_headings":"","what":"Overview","title":"Facilities for Simulating from ODE-Based Models","text":"rxode2 R package solving simulating ode-based models. models convert rxode2 mini-language C create compiled dll fast solving. ODE solving using rxode2 key parts: rxode2() creates C code fast ODE solving based simple syntax related Leibnitz notation. NONMEM deSolve compatible data frame, created et() eventTable() easy simulation events data frame can augmented adding time-varying adding individual covariates (iCov= needed) multiple subject data, may parallelized. single subject output data frame adaptive Covariances metrics uncertanty can used simulate solving","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"installation","dir":"","previous_headings":"","what":"Installation","title":"Facilities for Simulating from ODE-Based Models","text":"can install released version rxode2 CRAN : fastest way install development version rxode2 use r-universe service. service compiles binaries development version MacOS Windows don’t wait package compilation: doesn’t work install development version rxode2 build models rxode2, need working c compiler. use parallel threaded solving rxode2, c compiler needs support open-mp. can check see R working c compiler can check : toolchain, can set described platform information :","code":"install.packages(\"rxode2\") install.packages(c(\"dparser\", \"rxode2ll\", \"rxode2parse\",                    \"rxode2random\", \"rxode2et\", \"rxode2\"),                  repos=c(nlmixr2=\"https://nlmixr2.r-universe.dev\",                          CRAN=\"https://cloud.r-project.org\")) devtools::install_github(\"nlmixr2/rxode2parse\") devtools::install_github(\"nlmixr2/rxode2random\") devtools::install_github(\"nlmixr2/rxode2et\") devtools::install_github(\"nlmixr2/rxode2ll\") devtools::install_github(\"nlmixr2/rxode2\") ## install.packages(\"pkgbuild\") pkgbuild::has_build_tools(debug = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"windows","dir":"","previous_headings":"Installation","what":"Windows","title":"Facilities for Simulating from ODE-Based Models","text":"windows may simply use installr install rtools: Alternatively can download install rtools directly.","code":"install.packages(\"installr\") library(installr) install.rtools()"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"mac-osx","dir":"","previous_headings":"Installation","what":"Mac OSX","title":"Facilities for Simulating from ODE-Based Models","text":"get speed need OpenMP enabled compile rxode2 compiler. various options date discussion likely data.table installation FAQ MacOS. last thing keep mind rxode2 uses code similar original lsoda requires gfortran compiler setup well OpenMP compilers. going using rxode2 nlmixr together older mac computer, suggest trying following: crashes R session binary work Mac machine. able run nlmixr, need compile package manually. proceed assuming homebrew installed system. system terminal need install dependencies compile symengine: installing dependencies, need re-install symengine:","code":"library(symengine) brew install cmake gmp mpfr libmpc install.packages(\"symengine\", type=\"source\") library(symengine)"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"linux","dir":"","previous_headings":"Installation","what":"Linux","title":"Facilities for Simulating from ODE-Based Models","text":"install linux make sure install gcc (openmp support) gfortran using distribution’s package manager. also install system dependencies like udunits symengine dependencies complete installation work linux. also done system packages package manager add appropriate repositories. point r2u project.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"r-versions-40-and-41","dir":"","previous_headings":"Installation","what":"R versions 4.0 and 4.1","title":"Facilities for Simulating from ODE-Based Models","text":"installation R versions 4.0.x 4.1.x, please see instructions install symengine nlmixr2 installation instructions: https://github.com/nlmixr2/nlmixr2#r-package-installation","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"development-version","dir":"","previous_headings":"","what":"Development version","title":"Facilities for Simulating from ODE-Based Models","text":"Since development version rxode2 uses StanHeaders, need make sure compiler setup support C++14, described rstan setup page. R 4.0, believe requires modifying windows toolchain longer (much easier setup). C++ toolchain setup appropriately, can install development version GitHub :","code":"# install.packages(\"devtools\") devtools::install_github(\"nlmixr2/rxode2parse\") devtools::install_github(\"nlmixr2/rxode2random\") devtools::install_github(\"nlmixr2/rxode2et\") devtools::install_github(\"nlmixr2/rxode2ll\") devtools::install_github(\"nlmixr2/rxode2\")"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"illustrated-example","dir":"","previous_headings":"","what":"Illustrated Example","title":"Facilities for Simulating from ODE-Based Models","text":"model equations can specified text string, model file R expression. differential algebraic equations permitted. Differential equations specified d/dt(var_name) =. equation can separated semicolon. load rxode2 package compile model: Model parameters may specified ini({}) model block, initial conditions can specified within model cmt(0)= X, like model eff(0) <- 1. may also specify subject variability initial conditions residual error components just like nlmixr2. allows single interface nlmixr2/rxode2 models. Also note, classic rxode2 interface still works just like past (don’t worry breaking code time). fact, can get classic rxode2 model $simulationModel ui object:","code":"library(rxode2) #> rxode2 2.1.2.9000 using 8 threads (see ?getRxThreads) #>   no cache: create with `rxCreateCache()`  mod1 <- function() {   ini({     # central      KA=2.94E-01     CL=1.86E+01     V2=4.02E+01     # peripheral     Q=1.05E+01     V3=2.97E+02     # effects     Kin=1     Kout=1     EC50=200    })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <- Q*C2 - Q*C3     eff(0) <- 1     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff   }) } mod1 <- mod1() # create the ui object (can also use `rxode2(mod1)`) mod1  summary(mod1$simulationModel)"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"specify-dosing-and-sampling-in-rxode2","dir":"","previous_headings":"","what":"Specify Dosing and sampling in rxode2","title":"Facilities for Simulating from ODE-Based Models","text":"rxode2 provides simple flexible way specify dosing sampling functions generate event table. First, empty event table generated “et()” function. interface similar NONMEM event tables: can see code, can dose compartment named rxode2 model. slight deviation NONMEM can reduce need compartment renumbering. events can also combined expanded (multi-subject events complex regimens) rbind, c, seq, rep. information creating complex dosing regimens using rxode2 see rxode2 events vignette.","code":"ev  <- et(amountUnits=\"mg\", timeUnits=\"hours\") %>%   et(amt=10000, addl=9,ii=12,cmt=\"depot\") %>%   et(time=120, amt=2000, addl=4, ii=14, cmt=\"depot\") %>%   et(0:240) # Add sampling"},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"solving-odes","dir":"","previous_headings":"","what":"Solving ODEs","title":"Facilities for Simulating from ODE-Based Models","text":"ODE can now solved using rxSolve: returns modified data frame. can see compartment values plot :  ,  Note labels automatically labeled units initial event table. rxode2 extracts units label plot (present).","code":"x <- mod1 %>% rxSolve(ev) #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> In file included from /usr/share/R/include/R.h:71, #>                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, #>                  from /tmp/RtmpMKFh3q/temp_libpath2b2b3c8645df/rxode2/include/rxode2.h:9, #>                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, #>                  from rx_f5e1c8d22ddcc3e6890ad688a5f37e59_.c:117: #> /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] #>    80 |     }; #>       |      ^ x #> ── Solved rxode2 object ── #> ── Parameters (x$params): ── #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #> ── Initial Conditions (x$inits): ── #> depot centr  peri   eff  #>     0     0     0     1  #> ── First part of data (object): ── #> # A tibble: 241 × 7 #>   time    C2    C3  depot centr  peri   eff #>    [h] <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1    0   0   0     10000     0     0   1    #> 2    1  44.4 0.920  7453. 1784.  273.  1.08 #> 3    2  54.9 2.67   5554. 2206.  794.  1.18 #> 4    3  51.9 4.46   4140. 2087. 1324.  1.23 #> 5    4  44.5 5.98   3085. 1789. 1776.  1.23 #> 6    5  36.5 7.18   2299. 1467. 2132.  1.21 #> # ℹ 235 more rows library(ggplot2) plot(x,C2) + ylab(\"Central Concentration\") plot(x,eff)  + ylab(\"Effect\")"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"ode-solving","dir":"","previous_headings":"","what":"ODE solving","title":"Facilities for Simulating from ODE-Based Models","text":"brief comparison pharmacometric ODE solving R packages rxode2. several R packages differential equations. popular deSolve. However pharmacometrics-specific ODE solving, 2 packages rxode2 released CRAN. uses compiled code faster ODE solving. mrgsolve, uses C++ lsoda solver solve ODE systems. user required write hybrid R/C++ code create mrgsolve model translated C++ solving. contrast, rxode2 R-like mini-language parsed C code solves ODE system. Unlike rxode2, mrgsolve currently support symbolic manipulation ODE systems, like automatic Jacobian calculation forward sensitivity calculation (rxode2 currently supports basis nlmixr2’s FOCEi algorithm) dMod, uses unique syntax create “reactions”. reactions create underlying ODEs created c code compiled deSolve model. contrast rxode2 defines ODE systems lower level. rxode2’s parsing mini-language comes C, whereas dMod’s parsing comes R. Like rxode2, dMod supports symbolic manipulation ODE systems calculates forward sensitivities adjoint sensitivities systems. Unlike rxode2, dMod thread-safe since deSolve yet thread-safe. PKPDsim defines models R-like syntax converts system compiled code. Like mrgsolve, PKPDsim currently support symbolic manipulation ODE systems. PKPDsim thread-safe. open pharmacometrics open source community fairly friendly, rxode2 maintainers positive interactions ODE-solving pharmacometric projects listed.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"pk-solved-systems","dir":"","previous_headings":"","what":"PK Solved systems","title":"Facilities for Simulating from ODE-Based Models","text":"rxode2 supports 1-3 compartment models gradients (using stan math’s auto-differentiation). currently uses equations PKADVAN allow time-varying covariates. rxode2 can mix ODEs solved systems.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"the-following-packages-for-solved-pk-systems-are-on-cran","dir":"","previous_headings":"PK Solved systems","what":"The following packages for solved PK systems are on CRAN","title":"Facilities for Simulating from ODE-Based Models","text":"mrgsolve currently 1-2 compartment (poly-exponential models) models built-. solved systems ODEs currently mixed. pmxTools currently 1-3 compartment (super-positioning) models built-. R-implementation. PKPDsim uses 1-3 “ADVAN” solutions using non-superpositioning. PKPDmodels one-compartment model gradients.","code":""},{"path":"https://nlmixr2.github.io/rxode2/index.html","id":"non-cran-libraries","dir":"","previous_headings":"PK Solved systems","what":"Non-CRAN libraries:","title":"Facilities for Simulating from ODE-Based Models","text":"PKADVAN Provides 1-3 compartment models using non-superpositioning. allows time-varying covariates.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":null,"dir":"Reference","previous_headings":"","what":"Turn into an ini block for initialization — as.ini","title":"Turn into an ini block for initialization — as.ini","text":"Turn ini block initialization","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Turn into an ini block for initialization — as.ini","text":"","code":"as.ini(x)  # S3 method for character as.ini(x)  # S3 method for data.frame as.ini(x)  # S3 method for call as.ini(x)  # S3 method for lotriFix as.ini(x)  # S3 method for matrix as.ini(x)  # S3 method for default as.ini(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Turn into an ini block for initialization — as.ini","text":"x Item convert rxode2/nlmixr2 ui ini expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Turn into an ini block for initialization — as.ini","text":"rxode2 ini expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Turn into an ini block for initialization — as.ini","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.ini.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Turn into an ini block for initialization — as.ini","text":"","code":"ini <- quote(ini({   tka <- log(1.57)   tcl <- log(2.72)   tv <- log(31.5)   eta.ka ~ 0.6   eta.cl ~ 0.3   eta.v ~ 0.1   add.sd <- 0.7 }))  as.ini(ini) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  l <- quote(lotri({   tka <- log(1.57)   tcl <- log(2.72)   tv <- log(31.5)   eta.ka ~ 0.6   eta.cl ~ 0.3   eta.v ~ 0.1   add.sd <- 0.7  }))  as.ini(l) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  m <- lotri({    eta.ka ~ 0.6    eta.cl ~ 0.3    eta.v ~ 0.1 })  as.ini(m) #> ini({ #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  one.compartment <- function() {   ini({     tka <- log(1.57)     tcl <- log(2.72)     tv <- log(31.5)     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  as.ini(one.compartment) #>   #>   #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- c(0, 0.7) #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ui <- one.compartment()  as.ini(ui) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- c(0, 0.7) #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ui$iniDf #>   ntheta neta1 neta2   name lower       est upper   fix label backTransform #> 1      1    NA    NA    tka  -Inf 0.4510756   Inf FALSE  <NA>          <NA> #> 2      2    NA    NA    tcl  -Inf 1.0006319   Inf FALSE  <NA>          <NA> #> 3      3    NA    NA     tv  -Inf 3.4499875   Inf FALSE  <NA>          <NA> #> 4      4    NA    NA add.sd     0 0.7000000   Inf FALSE  <NA>          <NA> #> 5     NA     1     1 eta.ka  -Inf 0.6000000   Inf FALSE  <NA>          <NA> #> 6     NA     2     2 eta.cl  -Inf 0.3000000   Inf FALSE  <NA>          <NA> #> 7     NA     3     3  eta.v  -Inf 0.1000000   Inf FALSE  <NA>          <NA> #>   condition  err #> 1      <NA> <NA> #> 2      <NA> <NA> #> 3      <NA> <NA> #> 4        cp  add #> 5        id <NA> #> 6        id <NA> #> 7        id <NA>  as.ini(ui$iniDf) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- c(0, 0.7) #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ini <- c(\"ini({\",           \"tka <- log(1.57)\",           \"tcl <- log(2.72)\",           \"tv <- log(31.5)\",           \"eta.ka ~ 0.6\",           \"eta.cl ~ 0.3\",           \"eta.v ~ 0.1\",           \"add.sd <- 0.7\",           \"})\")  as.ini(ini) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })  ini <- paste(ini, collapse=\"\\n\")  as.ini(ini) #> ini({ #>     tka <- 0.451075619360217 #>     tcl <- 1.00063188030791 #>     tv <- 3.44998754583159 #>     add.sd <- 0.7 #>     eta.ka ~ 0.6 #>     eta.cl ~ 0.3 #>     eta.v ~ 0.1 #> })"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":null,"dir":"Reference","previous_headings":"","what":"Turn into a model expression — as.model","title":"Turn into a model expression — as.model","text":"Turn model expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Turn into a model expression — as.model","text":"","code":"as.model(x)  # S3 method for character as.model(x)  # S3 method for call as.model(x)  # S3 method for list as.model(x)  # S3 method for default as.model(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Turn into a model expression — as.model","text":"x item convert model({}) expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Turn into a model expression — as.model","text":"model expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Turn into a model expression — as.model","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Turn into a model expression — as.model","text":"","code":"model <- quote(model({   ka <- exp(tka + eta.ka)   cl <- exp(tcl + eta.cl)   v <- exp(tv + eta.v)   d/dt(depot) = -ka * depot   d/dt(center) = ka * depot - cl / v * center   cp = center / v   cp ~ add(add.sd) }))   as.model(model) #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })  one.compartment <- function() {   ini({     tka <- log(1.57)     tcl <- log(2.72)     tv <- log(31.5)     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  as.model(one.compartment) #>   #>   #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })   ui <- one.compartment()  as.model(ui) #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })  model <- c(\"model({\",            \"ka <- exp(tka + eta.ka)\",            \"cl <- exp(tcl + eta.cl)\",            \"v <- exp(tv + eta.v)\",            \"d/dt(depot) = -ka * depot\",            \"d/dt(center) = ka * depot - cl / v * center\",            \"cp = center / v\",            \"cp ~ add(add.sd)\",            \"})\")  as.model(model) #> model({ #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     d/dt(depot) = -ka * depot #>     d/dt(center) = ka * depot - cl/v * center #>     cp = center/v #>     cp ~ add(add.sd) #> })  model <- paste(model, collapse=\"\\n\")  as.model(model) #> model({ #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> })"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":null,"dir":"Reference","previous_headings":"","what":"As rxode2 ui — as.rxUi","title":"As rxode2 ui — as.rxUi","text":"rxode2 ui","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"As rxode2 ui — as.rxUi","text":"","code":"as.rxUi(x)  # S3 method for rxode2 as.rxUi(x)  # S3 method for rxode2tos as.rxUi(x)  # S3 method for rxModelVars as.rxUi(x)  # S3 method for `function` as.rxUi(x)  # S3 method for rxUi as.rxUi(x)  # S3 method for default as.rxUi(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"As rxode2 ui — as.rxUi","text":"x Object convert rxUi object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"As rxode2 ui — as.rxUi","text":"rxUi object (error converted)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"As rxode2 ui — as.rxUi","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/as.rxUi.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"As rxode2 ui — as.rxUi","text":"","code":"mod1 <- function() {  ini({    # central     KA=2.94E-01    CL=1.86E+01    V2=4.02E+01    # peripheral    Q=1.05E+01    V3=2.97E+02    # effects    Kin=1    Kout=1    EC50=200   })  model({    C2 <- centr/V2    C3 <- peri/V3    d/dt(depot) <- -KA*depot    d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3    d/dt(peri)  <- Q*C2 - Q*C3    eff(0) <- 1    d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff  }) }  as.rxUi(mod1) #>   #>   #>  ── rxode2-based free-form 4-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2            centr #> 3                  3             peri #> 4                  4              eff #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         KA <- 0.294 #>         CL <- 18.6 #>         V2 <- 40.2 #>         Q <- 10.5 #>         V3 <- 297 #>         Kin <- 1 #>         Kout <- 1 #>         EC50 <- 200 #>     }) #>     model({ #>         C2 <- centr/V2 #>         C3 <- peri/V3 #>         d/dt(depot) <- -KA * depot #>         d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3 #>         d/dt(peri) <- Q * C2 - Q * C3 #>         eff(0) <- 1 #>         d/dt(eff) <- Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":null,"dir":"Reference","previous_headings":"","what":"Assert properties of the rxUi models — assertRxUi","title":"Assert properties of the rxUi models — assertRxUi","text":"Assert properties rxUi models","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assert properties of the rxUi models — assertRxUi","text":"","code":"assertRxUi(model, extra = \"\", .var.name = .vname(model))  assertRxUiPrediction(model, extra = \"\", .var.name = .vname(model))  assertRxUiSingleEndpoint(model, extra = \"\", .var.name = .vname(model))  assertRxUiTransformNormal(model, extra = \"\", .var.name = .vname(model))  assertRxUiNormal(model, extra = \"\", .var.name = .vname(model))  assertRxUiMuRefOnly(model, extra = \"\", .var.name = .vname(model))  assertRxUiEstimatedResiduals(model, extra = \"\", .var.name = .vname(model))  assertRxUiPopulationOnly(model, extra = \"\", .var.name = .vname(model))  assertRxUiMixedOnly(model, extra = \"\", .var.name = .vname(model))  assertRxUiRandomOnIdOnly(model, extra = \"\", .var.name = .vname(model))"},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assert properties of the rxUi models — assertRxUi","text":"model Model check extra Extra text append error message (like \"focei\") .var.name [character(1)] Name checked object print assertions. Defaults heuristic implemented vname.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assert properties of the rxUi models — assertRxUi","text":"rxUi model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Assert properties of the rxUi models — assertRxUi","text":"functions different types assertions assertRxUi -- Make sure proper rxode2 model (throw error) assertRxUiSingleEndpoint -- Make sure rxode2 model single endpoint model (throw error) assertRxUiTransformNormal -- Make sure model residual distribution normal transformably normal assertRxUiNormal -- Make sure model residual distribution normal assertRxUiEstimatedResiduals -- Make sure residual error parameters estimated (modeled). assertRxUiPopulationOnly -- Make sure model population model (mixed effects) assertRxUiMixedOnly -- Make sure model mixed effect model (population effect, ) assertRxUiPrediction -- Make sure model predictions assertRxUiMuRefOnly -- Make sure parameters mu-referenced assertRxUiRandomOnIdOnly -- Make sure random effects ID level","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assert properties of the rxUi models — assertRxUi","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/assertRxUi.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Assert properties of the rxUi models — assertRxUi","text":"","code":"# \\donttest{ one.cmt <- function() {  ini({    tka <- 0.45; label(\"Ka\")    tcl <- log(c(0, 2.7, 100)); label(\"Cl\")    tv <- 3.45; label(\"V\")    eta.ka ~ 0.6    eta.cl ~ 0.3    eta.v ~ 0.1    add.sd <- 0.7  })  model({    ka <- exp(tka + eta.ka)    cl <- exp(tcl + eta.cl)    v <- exp(tv + eta.v)    linCmt() ~ add(add.sd)  }) }  assertRxUi(one.cmt) #>   #>   # assertRxUi(rnorm) # will fail  assertRxUiSingleEndpoint(one.cmt) #>   #>   # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"meant perform way quantile() can drop replacement code using quantile() using distributional assumptions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"","code":"binomProbs(x, ...)  # S3 method for default binomProbs(   x,   probs = c(0.025, 0.05, 0.5, 0.95, 0.975),   na.rm = FALSE,   names = TRUE,   onlyProbs = TRUE,   n = 0L,   m = 0L,   pred = FALSE,   piMethod = c(\"lim\"),   M = 5e+05,   tol = .Machine$double.eps^0.25,   ciMethod = c(\"wilson\", \"wilsonCorrect\", \"agrestiCoull\", \"wald\", \"wc\", \"ac\"),   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"x numeric vector whose mean probability based confidence values wanted, NA NaN values allowed numeric vectors unless na.rm TRUE. ... Arguments passed default method, allows many different methods applied. probs numeric vector probabilities values interval 0 1, inclusive. 0, represents maximum observed, 1, represents maximum observed. 0.5 represents expected probability (mean). na.rm logical; true, NA NaN's removed x quantiles computed. names logical; true, result names attribute. onlyProbs logical; true, return probability based confidence interval/prediction interval estimates, otherwise return extra statistics. n integer/integerish; n used calculate prediction confidence interval.  n=0 (default) use number non-NA observations.  calculating prediction interval, represents number observations used input (\"true\") distribution. m integer.  using prediction interval represents number samples observed future prediction interval. pred Use prediction interval instead confidence interval.  default FALSE. piMethod gives prediction interval method (currently lim) Lu 2020 M number simulations run LIM PI. tol tolerance root finding LIM prediction interval ciMethod gives method calculating confidence interval. Can : \"argestiCoull\" \"ac\" -- Agresti-Coull method. 95\\ interval, method use concept   \"adding 2 successes 2 failures,\" rather uses formulas explicitly described following link: https://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval#Agresti-Coull_Interval. \"wilson\" -- Wilson Method \"wilsonCorrect\" \"wc\" -- Wilson method continuity correction \"wald\" -- Wald confidence interval standard z approximation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"default return probabilities names (named) points expected distribution located given sampling mean standard deviation. onlyProbs=FALSE prepend mean, variance, standard deviation, minimum, maximum number non-NA observations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"used confidence intervals rxode2 solved objects using confint(mean=\"binom\")","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"Newcombe, R. G. (1998). \"Two-sided confidence intervals single proportion: comparison seven methods\". Statistics Medicine. 17 (8): 857–872. doi:10.1002/(SICI)1097-0258(19980430)17:8<857::AID-SIM777>3.0.CO;2-E. PMID 9595616. Hezhi Lu, Hua Jin, new prediction interval binomial random variable based inferential models, Journal Statistical Planning Inference, Volume 205, 2020, Pages 156-174, ISSN 0378-3758, https://doi.org/10.1016/j.jspi.2019.07.001.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/binomProbs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate expected confidence bands with binomial sampling distribution — binomProbs","text":"","code":"x<- rbinom(7001, p=0.375, size=1) binomProbs(x) #>      2.5%        5%       50%       95%     97.5%  #> 0.3716257 0.3734362 0.3829453 0.3925448 0.3943933   # you can also use the prediction interval # \\donttest{ binomProbs(x, pred=TRUE) #>      2.5%        5%       50%       95%     97.5%  #> 0.3668047 0.3695186 0.3829453 0.3965148 0.3990858  # }  # Can get some extra statistics if you request onlyProbs=FALSE binomProbs(x, onlyProbs=FALSE) #>         mean          var           sd            n         2.5%           5%  #>    0.3829453    0.2362982    0.4861051 7001.0000000    0.3716257    0.3734362  #>          50%          95%        97.5%  #>    0.3829453    0.3925448    0.3943933   x[2] <- NA_real_  binomProbs(x, onlyProbs=FALSE) #>  mean   var    sd     n  2.5%    5%   50%   95% 97.5%  #>    NA    NA    NA    NA    NA    NA    NA    NA    NA   binomProbs(x, na.rm=TRUE) #>      2.5%        5%       50%       95%     97.5%  #> 0.3715373 0.3733478 0.3828571 0.3924570 0.3943055"},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the rxode2 coefficients — coef.rxode2","title":"Return the rxode2 coefficients — coef.rxode2","text":"returns parameters , state variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the rxode2 coefficients — coef.rxode2","text":"","code":"# S3 method for rxode2 coef(object, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the rxode2 coefficients — coef.rxode2","text":"object rxode2 object ... ignored arguments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the rxode2 coefficients — coef.rxode2","text":"rxCoef object following params  list strings parameters rxode2 object state list strings names state rxode2 object. ini model specified default values parameters. rxode2 referring rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/coef.rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the rxode2 coefficients — coef.rxode2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":null,"dir":"Reference","previous_headings":"","what":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"returns new expression, old expression, new variable old variable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"","code":".assertRenameErrorModelLine(line, vars)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"line quoted call information line vars Variables contained within mdel","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"list(new, old, newChar, oldChar)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-assertRenameErrorModelLine.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This function asserts the requested rename makes sense — .assertRenameErrorModelLine","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":null,"dir":"Reference","previous_headings":"","what":"This copies the rxode2 UI object so it can be modified — .copyUi","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"copies rxode2 UI object can modified","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"","code":".copyUi(ui)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"ui Original UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"Copied UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-copyUi.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This copies the rxode2 UI object so it can be modified — .copyUi","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":null,"dir":"Reference","previous_headings":"","what":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"Handle single error normal t distributions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"","code":".handleSingleErrTypeNormOrTFoceiBase(   env,   pred1,   errNum = 1L,   rxPredLlik = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"env Environment parsed model pred1 data.frame current error errNum number error specification nlmixr2 model rxPredLlik boolean indicating log likelihood calculated non-normal distributions.  default TRUE.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"list lines added.  lines contain rx_yj_ integer corresponds transformation type. rx_lambda_ transformation lambda rx_low_ lower boundary transformation rx_hi_ upper boundary transformation rx_pred_f_ prediction function rx_pred_ transformed prediction function rx_r_ transformed variance","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-handleSingleErrTypeNormOrTFoceiBase.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Handle the single error for normal or t distributions — .handleSingleErrTypeNormOrTFoceiBase","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":null,"dir":"Reference","previous_headings":"","what":"This gets the append arg for the ini() piping — .iniGetAppendArg","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"gets append arg ini() piping","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"","code":".iniGetAppendArg(f, s)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"f try(force(append)) argument, s .character(substitute(append)) argument","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"corrected ini piping argument exported creating new ini methods requirements piping","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniGetAppendArg.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This gets the append arg for the ini() piping — .iniGetAppendArg","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":null,"dir":"Reference","previous_headings":"","what":"Reorder rows in iniDf — .iniHandleAppend","title":"Reorder rows in iniDf — .iniHandleAppend","text":"Reorder rows iniDf","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reorder rows in iniDf — .iniHandleAppend","text":"","code":".iniHandleAppend(expr, rxui, envir, append)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reorder rows in iniDf — .iniHandleAppend","text":"expr Expression parsing rxui User interface function envir Environment parsing append Reorder theta parameters.  NULL means change parameter order.  parameter name (character string) means put new parameter named parameter.  number less equal zero means put parameter beginning list.  number greater last parameter number means put parameter end list.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleAppend.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Reorder rows in iniDf — .iniHandleAppend","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":null,"dir":"Reference","previous_headings":"","what":"Update the iniDf of a model — .iniHandleLine","title":"Update the iniDf of a model — .iniHandleLine","text":"Update iniDf model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update the iniDf of a model — .iniHandleLine","text":"","code":".iniHandleLine(expr, rxui, envir = parent.frame(), append = NULL)  .iniHandleFixOrUnfix(expr, rxui, envir = parent.frame(), append = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update the iniDf of a model — .iniHandleLine","text":"expr Expression parsing rxui User interface function envir Environment parsing append Reorder theta parameters.  NULL means change parameter order.  parameter name (character string) means put new parameter named parameter.  number less equal zero means put parameter beginning list.  number greater last parameter number means put parameter end list.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update the iniDf of a model — .iniHandleLine","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-iniHandleLine.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Update the iniDf of a model — .iniHandleLine","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":null,"dir":"Reference","previous_headings":"","what":"Check if a language object matches a template language object — .matchesLangTemplate","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"template == str2lang(\".\"), match anything. template == str2lang(\".name\"), match name. template == str2lang(\".call()\"), match call.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"","code":".matchesLangTemplate(x, template)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"x object check template template object match","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"TRUE matches, FALSE, otherwise","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-matchesLangTemplate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Check if a language object matches a template language object — .matchesLangTemplate","text":"","code":".matchesLangTemplate(str2lang(\"d/dt(foo)\"), str2lang(\"d/dt(.name)\")) #> [1] TRUE .matchesLangTemplate(str2lang(\"d/dt(foo)\"), str2lang(\"d/foo(.name)\")) #> [1] FALSE .matchesLangTemplate(str2lang(\"d/dt(foo)\"), str2lang(\"d/.\")) #> [1] TRUE"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal messaging statements — .minfo","title":"Internal messaging statements — .minfo","text":"Internal messaging statements","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal messaging statements — .minfo","text":"","code":".minfo(text, ..., .envir = parent.frame())  .malert(text, ..., .envir = parent.frame())  .mwarn(text, ..., .envir = parent.frame())  .msuccess(text, ..., .envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal messaging statements — .minfo","text":"text Text ... arguments .envir Environment evaluate ","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal messaging statements — .minfo","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-minfo.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal messaging statements — .minfo","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":null,"dir":"Reference","previous_headings":"","what":"Handle model lines — .modelHandleModelLines","title":"Handle model lines — .modelHandleModelLines","text":"Handle model lines","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Handle model lines — .modelHandleModelLines","text":"","code":".modelHandleModelLines(   modelLines,   rxui,   modifyIni = FALSE,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir )"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Handle model lines — .modelHandleModelLines","text":"modelLines model lines considered rxui rxode2 UI object modifyIni ini() considered append boolean determine lines appended piping.  possible values : TRUE lines appended model instead replaced FALSE lines replaced model (default) NA lines pre-pended model instead replaced lhs expression, append lines last observed line expression lhs auto boolean tells piping automatically selects parameters characterized population parameter, subject variability, covariate.  TRUE automatic selection occurs.  FALSE automatic selection turned everything added covariate (can promoted parameter ini statement). default TRUE, can changed options(rxode2.autoVarPiping=FALSE). cov character vector variables assumed covariates.  override automatic promotion population parameter estimate (eta) envir Environment evaluation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Handle model lines — .modelHandleModelLines","text":"New UI","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-modelHandleModelLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Handle model lines — .modelHandleModelLines","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":null,"dir":"Reference","previous_headings":"","what":"Returns quoted call information — .quoteCallInfoLines","title":"Returns quoted call information — .quoteCallInfoLines","text":"Returns quoted call information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Returns quoted call information — .quoteCallInfoLines","text":"","code":".quoteCallInfoLines(callInfo, envir = parent.frame(), iniDf = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Returns quoted call information — .quoteCallInfoLines","text":"callInfo Call information envir Environment evaluation (needed) iniDf parent model iniDf piping ini block (NULL otherwise)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Returns quoted call information — .quoteCallInfoLines","text":"Quote call information.  name=expression, change name<-expression quoted call list. expressions within brackets ie {}, unlist brackets called one single sequence.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-quoteCallInfoLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Returns quoted call information — .quoteCallInfoLines","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":null,"dir":"Reference","previous_headings":"","what":"Print out a table in the documentation — .rxDocTable","title":"Print out a table in the documentation — .rxDocTable","text":"Print table documentation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print out a table in the documentation — .rxDocTable","text":"","code":".rxDocTable(table, caption = \"none\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print out a table in the documentation — .rxDocTable","text":"table data frame caption character vector representing caption latex table","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print out a table in the documentation — .rxDocTable","text":"based knitr context: output kableExtra::kbl latex output output DT::datatable html output otherwise output knitr::kable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print out a table in the documentation — .rxDocTable","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxDocTable.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Print out a table in the documentation — .rxDocTable","text":"","code":".rxDocTable(rxReservedKeywords) #>  #>  #> |Reserved Name  |Meaning                                                                                                                                                                           |Alias  | #> |:--------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------| #> |time           |solver time                                                                                                                                                                       |t      | #> |podo           |In Transit compartment models, last dose amount                                                                                                                                   |       | #> |tlast          |Time of Last dose                                                                                                                                                                 |       | #> |M_E            |Exp(1)                                                                                                                                                                            |       | #> |M_LOG2E        |log2(e)                                                                                                                                                                           |       | #> |M_LOG10E       |log10(e)                                                                                                                                                                          |       | #> |M_LN2          |log(2)                                                                                                                                                                            |       | #> |M_LN10         |log(10)                                                                                                                                                                           |       | #> |M_PI           |pi                                                                                                                                                                                |       | #> |M_PI_2         |pi/2                                                                                                                                                                              |       | #> |M_PI_4         |pi/4                                                                                                                                                                              |       | #> |M_1_PI         |1/pi                                                                                                                                                                              |       | #> |M_2_PI         |2/pi                                                                                                                                                                              |       | #> |M_2_SQRTPI     |2/sqrt(pi)                                                                                                                                                                        |       | #> |M_SQRT2        |sqrt(2)                                                                                                                                                                           |       | #> |M_SQRT1_2      |1/sqrt(2)                                                                                                                                                                         |       | #> |M_SQRT_3       |sqrt(3)                                                                                                                                                                           |       | #> |M_SQRT_32      |sqrt(32)                                                                                                                                                                          |       | #> |M_LOG10_2      |Log10(2)                                                                                                                                                                          |       | #> |M_2PI          |2*pi                                                                                                                                                                              |       | #> |M_SQRT_PI      |sqrt(pi)                                                                                                                                                                          |       | #> |M_1_SQRT_2PI   |1/(sqrt(2*pi))                                                                                                                                                                    |       | #> |M_LN_SQRT_PI   |log(sqrt(pi))                                                                                                                                                                     |       | #> |M_LN_SQRT_2PI  |log(sqrt(2*pi))                                                                                                                                                                   |       | #> |M_LN_SQRT_PId2 |log(sqrt(pi/2))                                                                                                                                                                   |       | #> |pi             |pi                                                                                                                                                                                |       | #> |NA             |R's NA value                                                                                                                                                                      |       | #> |NaN            |Not a Number Value                                                                                                                                                                |       | #> |Inf            |Infinite Value                                                                                                                                                                    |       | #> |newind         |1: First record of individual; 2: Subsequent record of individual                                                                                                                 |NEWIND | #> |rxFlag         |Flag for what part of the rxode2 model is being run; 1: ddt; 2: jac; 3: ini; 4: F; 5: lag; 6: rate; 7: dur; 8: mtime; 9: matrix exponential; 10: inductive linearization; 11: lhs |       |"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"Get upper boundary condition transformation ","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"","code":".rxGetHiBoundaryPred1AndIni(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"env Environment parsed model pred1 data.frame current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"Upper Boundary","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetHiBoundaryPred1AndIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the upper boundary condition when the transformation it — .rxGetHiBoundaryPred1AndIni","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"Get lambda value based pred information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"","code":".rxGetLambdaFromPred1AndIni(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"env Environment environment pred1 Single error data frame","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"Lambda expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLambdaFromPred1AndIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the lambda value based on the pred information — .rxGetLambdaFromPred1AndIni","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"Get lower boundary condition transformation requires ","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"","code":".rxGetLowBoundaryPred1AndIni(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"env Environment parsed model pred1 data.frame current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"Lower Boundary","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetLowBoundaryPred1AndIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the lower boundary condition when the transformation requires it — .rxGetLowBoundaryPred1AndIni","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the Observation transformation — .rxGetPredictionDVTransform","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"Get Observation transformation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"","code":".rxGetPredictionDVTransform(env, pred1, yj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"env Environment parsed model pred1 data.frame current error yj transformation number current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"transformation expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionDVTransform.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the Observation transformation — .rxGetPredictionDVTransform","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the prediction name — .rxGetPredictionF","title":"Get the prediction name — .rxGetPredictionF","text":"Get prediction name","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the prediction name — .rxGetPredictionF","text":"","code":".rxGetPredictionF(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the prediction name — .rxGetPredictionF","text":"env Environment parsed model pred1 data.frame current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the prediction name — .rxGetPredictionF","text":"prediction symbol","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionF.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the prediction name — .rxGetPredictionF","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the prediction transformation — .rxGetPredictionFTransform","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"Get prediction transformation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"","code":".rxGetPredictionFTransform(env, pred1, yj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"env Environment parsed model pred1 data.frame current error yj transformation number current error","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"transformation expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetPredictionFTransform.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the prediction transformation — .rxGetPredictionFTransform","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":null,"dir":"Reference","previous_headings":"","what":"Get Variance for error type — .rxGetVarianceForErrorType","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"Get Variance error type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"","code":".rxGetVarianceForErrorType(env, pred1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"env Environment pred1 Pred one end-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"Variance error type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxGetVarianceForErrorType.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get Variance for error type — .rxGetVarianceForErrorType","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"Internal function tell linCmt() model variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"","code":".rxIsLinCmt()"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"0 1","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxIsLinCmt.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal function to tell if the linCmt() is the model variables — .rxIsLinCmt","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function for calculating the Jacobian — .rxJacobian","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"Internal function calculating Jacobian","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"","code":".rxJacobian(model, vars = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"model symengine environment vars Variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"Jacobian information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxJacobian.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal function for calculating the Jacobian — .rxJacobian","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"Internal function generate model variables linCmt() model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"","code":".rxLinCmtGen(lenState, vars)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"lenState Length state vars Variables model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"Model variables expanded linCmt model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxLinCmtGen.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal function to generate the model variables for a linCmt() model — .rxLinCmtGen","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal Pruning function — .rxPrune","title":"Internal Pruning function — .rxPrune","text":"Internal Pruning function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal Pruning function — .rxPrune","text":"","code":".rxPrune(x, envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal Pruning function — .rxPrune","text":"x List quoted lines envir Environment information stored","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal Pruning function — .rxPrune","text":"Pruned model code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxPrune.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Internal Pruning function — .rxPrune","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":null,"dir":"Reference","previous_headings":"","what":"Sensitivity for model — .rxSens","title":"Sensitivity for model — .rxSens","text":"Sensitivity model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Sensitivity for model — .rxSens","text":"","code":".rxSens(model, vars, vars2)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Sensitivity for model — .rxSens","text":"model symengine model environment vars Variables single sensitivity vars2 present, 2 parameter sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Sensitivity for model — .rxSens","text":"Sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxSens.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Sensitivity for model — .rxSens","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":null,"dir":"Reference","previous_headings":"","what":"Temporarily set options then restore them while running code — .rxWithOptions","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"Temporarily set options restore running code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"","code":".rxWithOptions(ops, code)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"ops list options temporarily set code code code run sink","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"value code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithOptions.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Temporarily set options then restore them while running code — .rxWithOptions","text":"","code":".rxWithOptions(list(digits = 21), {   print(pi) }) #> [1] 3.141592653589793116  print(pi) #> [1] 3.141593"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":null,"dir":"Reference","previous_headings":"","what":"With one sink, then release — .rxWithSink","title":"With one sink, then release — .rxWithSink","text":"one sink, release","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"With one sink, then release — .rxWithSink","text":"","code":".rxWithSink(file, code)  .rxWithSinkBoth(file, code)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"With one sink, then release — .rxWithSink","text":"file path file sink running code code code run sink","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"With one sink, then release — .rxWithSink","text":"return results code section","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"With one sink, then release — .rxWithSink","text":".rxWithSink captures output cat .rxWithSinkBoth captures output cat message","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"With one sink, then release — .rxWithSink","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithSink.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"With one sink, then release — .rxWithSink","text":"","code":"t <- tempfile() .rxWithSink(t, cat(\"message\\n\")) cat(\"cat2\\n\") # now you can see the cat2 #> cat2 lines <- readLines(t) unlink(t)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":null,"dir":"Reference","previous_headings":"","what":"Temporarily set options then restore them while running code — .rxWithWd","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"Temporarily set options restore running code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"","code":".rxWithWd(wd, code)"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"wd working directory temporarily set system evaluating code code code run sink","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"value code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-rxWithWd.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Temporarily set options then restore them while running code — .rxWithWd","text":"","code":".rxWithWd(tempdir(), {   getwd() }) #> [1] \"/tmp/RtmpPXkf2x\"  getwd() #> [1] \"/home/runner/work/rxode2/rxode2/docs/reference\""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-useUtf.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to figure out if this session supports Unicode — .useUtf","title":"Internal function to figure out if this session supports Unicode — .useUtf","text":"Internal function figure session supports Unicode","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-useUtf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to figure out if this session supports Unicode — .useUtf","text":"","code":".useUtf()"},{"path":"https://nlmixr2.github.io/rxode2/reference/dot-useUtf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Internal function to figure out if this session supports Unicode — .useUtf","text":"boolean indicating session supports Unicode","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":null,"dir":"Reference","previous_headings":"","what":"Error function — erf","title":"Error function — erf","text":"Error function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Error function — erf","text":"","code":"erf(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Error function — erf","text":"x vector real values","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Error function — erf","text":"erf x","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Error function — erf","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/erf.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Error function — erf","text":"","code":"erf(1.0) #> [1] 0.8427008"},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":null,"dir":"Reference","previous_headings":"","what":"Event translation for rxode2 — etTrans","title":"Event translation for rxode2 — etTrans","text":"Event translation rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Event translation for rxode2 — etTrans","text":"","code":"etTrans(   inData,   obj,   addCmt = FALSE,   dropUnits = FALSE,   allTimeVar = FALSE,   keepDosingOnly = FALSE,   combineDvid = NULL,   keep = character(0),   addlKeepsCov = FALSE,   addlDropSs = TRUE,   ssAtDoseTime = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Event translation for rxode2 — etTrans","text":"inData Data frame translate obj Model translate data addCmt Add compartment data frame (default FALSE). dropUnits Boolean drop units (default FALSE). allTimeVar Treat covariates time-varying keepDosingOnly keep individuals dosing records trailing dosing records last observation. combineDvid boolean indicating rxode2 use DVID observation records change cmt value; Useful multiple-endpoint nlmixr models.  default determined option(\"rxode2.combine.dvid\") option set, TRUE. typically affect rxode2 simulations. keep named vector items want keep final rxode2 dataset. added rxode2 event records (seen), last observation carried forward used. addlKeepsCov determines additional dosing items repeats dose (FALSE) keeps covariates record dose (TRUE) addlDropSs steady state doses addl specification steady state flag dropped repeated doses (TRUE) retained (FALSE) ssAtDoseTime Boolean TRUE back calculates steady concentration actual time dose, otherwise FALSE doses shifted","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/etTrans.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Event translation for rxode2 — etTrans","text":"Object solving rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":null,"dir":"Reference","previous_headings":"","what":"Find the assignments in R expression — findLhs","title":"Find the assignments in R expression — findLhs","text":"Find assignments R expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Find the assignments in R expression — findLhs","text":"","code":"findLhs(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Find the assignments in R expression — findLhs","text":"x R expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Find the assignments in R expression — findLhs","text":"list assigned parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/findLhs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Find the assignments in R expression — findLhs","text":"Hadley Wickham Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":null,"dir":"Reference","previous_headings":"","what":"Gammap: normalized lower incomplete gamma function — gammap","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"gamma_p boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"","code":"gammap(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"numeric '' parameter normalized lower incomplete gamma z numeric 'z' parameter normalized lower incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"gammap results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"gamma p function given : gammap = lowergamma(, z)/gamma()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammap.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Gammap: normalized lower incomplete gamma function — gammap","text":"","code":"gammap(1, 3) #> [1] 0.9502129 gammap(1:3, 3) #> [1] 0.9502129 0.8008517 0.5768099 gammap(1, 1:3) #> [1] 0.6321206 0.8646647 0.9502129"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":null,"dir":"Reference","previous_headings":"","what":"gammapDer: derivative of gammap — gammapDer","title":"gammapDer: derivative of gammap — gammapDer","text":"gamma_p_derivative boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"gammapDer: derivative of gammap — gammapDer","text":"","code":"gammapDer(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"gammapDer: derivative of gammap — gammapDer","text":"numeric '' parameter upper incomplete gamma z numeric 'z' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"gammapDer: derivative of gammap — gammapDer","text":"lowergamma results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gammapDer: derivative of gammap — gammapDer","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapDer.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"gammapDer: derivative of gammap — gammapDer","text":"","code":"gammapDer(1:3, 3) #> [1] 0.04978707 0.14936121 0.22404181  gammapDer(1, 1:3) #> [1] 0.36787944 0.13533528 0.04978707"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":null,"dir":"Reference","previous_headings":"","what":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"gammapInv gammapInva:  Inverses normalized gammap function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"","code":"gammapInv(a, p)  gammapInva(x, p)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"numeric '' parameter upper incomplete gamma p numeric 'p' parameter upper incomplete gamma x numeric 'x' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"inverse gammap results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"equation: p = gammap(, x) 'gammapInv' function returns value 'x' satisfies equation 'gammapInva' function returns value 'q' satisfies equation NOTE: gammapInva slow","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammapInv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"gammapInv and gammapInva: Inverses of normalized gammap function — gammapInv","text":"","code":"gammapInv(1:3, 0.5) #> [1] 0.6931472 1.6783470 2.6740603  gammapInv(1, 1:3 / 3.1) #> [1] 0.3894648 1.0360919 3.4339872  gammapInv(1:3, 1:3 / 3.1) #> [1] 0.3894648 2.1988984 6.8872929  gammapInva(1:3, 1:3 / 3.1) #> [1] 0.3894648 2.1988984 6.8872929"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":null,"dir":"Reference","previous_headings":"","what":"Gammaq: normalized upper incomplete gamma function — gammaq","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"gamma_q boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"","code":"gammaq(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"numeric '' parameter normalized upper incomplete gamma z numeric 'z' parameter normalized upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"gammaq results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"gamma q function given : gammaq = uppergamma(, z)/gamma()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaq.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Gammaq: normalized upper incomplete gamma function — gammaq","text":"","code":"gammaq(1, 3) #> [1] 0.04978707 gammaq(1:3, 3) #> [1] 0.04978707 0.19914827 0.42319008 gammaq(1, 1:3) #> [1] 0.36787944 0.13533528 0.04978707"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":null,"dir":"Reference","previous_headings":"","what":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"gammaqInv gammaqInva:  Inverses normalized gammaq function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"","code":"gammaqInv(a, q)  gammaqInva(x, q)"},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"numeric '' parameter upper incomplete gamma q numeric 'q' parameter upper incomplete gamma x numeric 'x' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"inverse gammaq results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"equation: q = gammaq(, x) 'gammaqInv' function returns value 'x' satisfies equation 'gammaqInva' function returns value '' satisfies equation NOTE: gammaqInva slow","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/gammaqInv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"gammaqInv and gammaqInva: Inverses of normalized gammaq function — gammaqInv","text":"","code":"gammaqInv(1:3, 0.5) #> [1] 0.6931472 1.6783470 2.6740603  gammaqInv(1, 1:3 / 3) #> [1] 1.0986123 0.4054651 0.0000000  gammaqInv(1:3, 1:3 / 3.1) #> [1] 1.131402 1.248533 0.684289  gammaqInva(1:3, 1:3 / 3.1) #> [1] 1.131402 1.248533 0.684289"},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"Create complete shiny application exploring dosing regimens given (hardcoded) PK/PD model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"","code":"genShinyApp.template(   appDir = \"shinyExample\",   verbose = TRUE,   ODE.config = list(ode = \"model\", params = c(KA = 0.294), inits = c(eff = 1), method =     \"lsoda\", atol = 1e-08, rtol = 1e-06) )  write.template.server(appDir)  write.template.ui(appDir, statevars)"},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"appDir string directory store shiny app, default \"shinyExample\". directory appDir created exist. verbose logical specifying whether write messages shiny app generated. Defaults TRUE. ODE.config model name compiled list parameters sent rxSolve(). statevars List statevars passed write.template.ui() function.  usually called directly. PK/PD model defined using rxode2(), set parameters initial values defined.  appropriate R scripts shiny's user interface ui.R server logic server.R created directory appDir. function evaluates following PK/PD model default: can changed ODE.config parameter. launch shiny app, simply issue runApp(appDir) R command.","code":"C2 = centr/V2;     C3 = peri/V3;     d/dt(depot) =-KA*depot;     d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  =                    Q*C2 - Q*C3;     d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff;"},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"None, functions used side effects.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"functions create simple, working example dosing regimen simulation web application. Users may want modify code experiment creating shiny applications specific rxode2 models.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/genShinyApp.template.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Generate an example (template) of a dosing regimen shiny app — genShinyApp.template","text":"","code":"# \\donttest{ # remove myapp when the example is complete on.exit(unlink(\"myapp\", recursive = TRUE, force = TRUE)) # create the shiny app example (template) genShinyApp.template(appDir = \"myapp\") #>  #> Generating an example (template) for a dosing regimen shiny app #>  #> Using the following PK/PD model: #>    C2 = centr/V2; #>    C3 = peri/V3; #>    d/dt(depot) =-KA*depot; #>    d/dt(centr) = KA*depot - CL*C2 - Q*C2 + Q*C3; #>    d/dt(peri)  =                    Q*C2 - Q*C3; #>    d/dt(eff)  = Kin - Kout*(1-C2/(EC50+C2))*eff; #>      #> Translating the PK/PD ODE model into C, compiling, etc. #>  #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #>  #> Parameters and their values: #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.294  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #>  #> Initial values in each compartment: #> depot centr  pari   eff  #>     0     0     0     1  #>  #> Warning: 2 arguments not used by format ' #>       # #>       # Dosing regimen template generated by rxode2::genShinyApp.template() #>       # #>  #>       debug = TRUE #>       #wd = sprintf(\"%%s/../\", getwd()) #>       #setwd(wd) #>  #>       # Server inputs: Dose, dosing regimen, dosing frequency, #>       # dosing cycle definition, number of dosing cycles #>  #>       library(shiny) #>       library(rxode2) #>  #>       # read objects from \"rx_shiny_data.rda\" in the  AppDir folder, #>       # objects include, mod1, params, inits, method, atol, rtol.] #>  #>       load(\"./rx_shiny_data.rda\") #>       if (!rxDynLoad(mod1)) mod1 <- rxode2(mod1, modName=\"mod1\") #>       # Define server logic #>       shinyServer(function(input, output) { #>  #>         get.cp <- reactive({ #>           ds <- input$Dose #>           reg <- switch(input$regimen, \"QD\"=1, \"BID\"=2) #>           cyc <- switch(input$cycle, #>               \"continous\"=c(7,0), #>               \"1wkon 1wkoff\"=c(7,7), #>               \"2wkon 1wkoff\"=c(14,7), #>               \"3wkon 1wkoff\"=c(21,7) #>           ) #>           cyc <- [... truncated] #> Shiny files (ui.R, server.R) plus R data saved. #>  #> To launch the Shiny app, type the following two R commands: #>  #> \tlibrary(shiny) #> \trunApp(\"myapp\") #>  # run the shiny app if (requireNamespace(\"shiny\", quietly=TRUE)) {   library(shiny)   # runApp(\"myapp\") # Won't launch in environments without browsers } # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"Get base simulation model simulation inis underlying rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"","code":"getBaseIniSimModel(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"obj Fit Object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"Simulation object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseIniSimModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the base simulation model for simulation with inis in the underlying rxode2 model — getBaseIniSimModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the base simulation model for simulation — getBaseSimModel","title":"Get the base simulation model for simulation — getBaseSimModel","text":"Get base simulation model simulation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the base simulation model for simulation — getBaseSimModel","text":"","code":"getBaseSimModel(obj)  # S3 method for default getBaseSimModel(obj)  # S3 method for default getBaseIniSimModel(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the base simulation model for simulation — getBaseSimModel","text":"obj Fit Object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the base simulation model for simulation — getBaseSimModel","text":"Simulation object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSimModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the base simulation model for simulation — getBaseSimModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"Get symengine loading symengine rxS()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"","code":"getBaseSymengineModel(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"obj Object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"Simulation model ready load symeng","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getBaseSymengineModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the symengine for loading into symengine with rxS() — getBaseSymengineModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":null,"dir":"Reference","previous_headings":"","what":"Get/Set the number of threads that rxode2 uses — getRxThreads","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"Get/Set number threads rxode2 uses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"","code":"getRxThreads(verbose = FALSE)  setRxThreads(threads = NULL, percent = NULL, throttle = NULL)  rxCores(verbose = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"verbose Display value relevant OpenMP settings threads NULL (default) rereads environment variables. 0 means use logical CPUs available. Otherwise number >= 1 percent provided number 2 100; percentage logical CPUs use. default startup, 50 percent. throttle 2 (default) means , roughly speaking, single thread used number subjects solved <=2, 2 threads number points <=4, etc. throttle speed small data tasks (especially repeated many times) incurring overhead managing multiple threads. throttle also suppress sorting ID solved first (nsubject solved)*throttle <= nthreads.  rxode2 sorting occurs minimize time waiting another thread finish. last item solved long solving time, solving wait last costly solving occur. items likely take time solved first, wait less likely impact overall solving time. rxode2 IDs sorted individual number solving points (largest first). also C interface allows IDs resorted total time spent solving equation.  allows packages like nlmixr sort solving time needed. Overall number threads throttled (restricted) small tasks sorting IDs suppressed.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/getRxThreads.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get/Set the number of threads that rxode2 uses — getRxThreads","text":"number threads rxode2 uses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":null,"dir":"Reference","previous_headings":"","what":"Empty Guide — guide_none","title":"Empty Guide — guide_none","text":"empty guide draws nothing; included rxode2 compatibility ggplot 3.2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Empty Guide — guide_none","text":"","code":"guide_none(title = waiver(), position = waiver())"},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Empty Guide — guide_none","text":"title character string expression indicating title guide. NULL, title shown. default (waiver()), name scale object name specified labs() used title. position guide drawn: one top, bottom, left, right.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/guide_none.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Empty Guide — guide_none","text":"nothing, simply included compatible ggplot 3.2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign the ini block in the rxode2 related object — ini<-","title":"Assign the ini block in the rxode2 related object — ini<-","text":"Assign ini block rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign the ini block in the rxode2 related object — ini<-","text":"","code":"ini(x, envir = environment(x)) <- value"},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign the ini block in the rxode2 related object — ini<-","text":"x rxode2 related object envir Environment assignment occurs value Value object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign the ini block in the rxode2 related object — ini<-","text":"rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini-set.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign the ini block in the rxode2 related object — ini<-","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":null,"dir":"Reference","previous_headings":"","what":"Ini block for rxode2/nlmixr models — ini.rxUi","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"ini block controls initial conditions 'theta' (fixed effects), 'omega' (random effects), 'sigma' (residual error) elements model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"","code":"# S3 method for rxUi ini(x, ..., envir = parent.frame(), append = NULL)  # S3 method for default ini(x, ..., envir = parent.frame(), append = NULL)  ini(x, ..., envir = parent.frame(), append = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"x expression ... expressions ini() function envir environment unevaluated model expressions evaluated.  May also NULL, list, data frame, pairlist integer specified sys.call. append Reorder theta parameters.  NULL means change parameter order.  parameter name (character string) means put new parameter named parameter.  number less equal zero means put parameter beginning list.  number greater last parameter number means put parameter end list.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"ini block","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"ini() function used two different ways.  main way used set initial conditions associated attributes (described ) model.  way used updating initial conditions model, often using pipe operator. 'theta' 'sigma' can set using either <- = tvCL <- 1 equivalently tvCL = 1.  'omega' can set ~ etaCL ~ 0.1. Parameters can named unnamed (though named parameters preferred). named parameter set using name left assignment unnamed parameters set without assignment operator.  tvCL <- 1 set named parameter tvCL 1.  Unnamed parameters set using just value, 1. estimation methods, lower upper bounds can set 'theta' 'sigma' values.  set lower /upper bound, use vector values.  vector c(lower, estimate, upper).  vector may given just estimate (estimate), lower bound estimate (c(lower, estimate)), three (c(lower, estimate, upper)).  set estimate upper bound without lower bound, set lower bound -Inf, c(-Inf, estimate, upper).  estimation method support bounds, bounds ignored warning. 'omega' values can set single value values lower-triangular matrix.  values may set either variance-covariance matrix (default) correlation matrix -diagonals standard deviations diagonals.  Names may set left side ~.  set variance-covariance matrix variance values 2 3 covariance -2.5 use ~c(2, 2.5, 3).  set matrix names iivKa iivCL, use iivKa + iivCL~c(2, 2.5, 3).  set correlation matrix standard deviations diagonal, use cor() like iivKa + iivCL~cor(2, -0.5, 3). Values may fixed (therefore estimated) using either name fixed end assignment calling fixed() function value fix.  'theta' 'sigma', either estimate full definition (including lower upper bounds) may included fixed setting.  example, following effectively equivalent set 'theta' 'sigma' fixed value (lower upper bounds ignored fixed value): tvCL <- fixed(1), tvCL <- fixed(0, 1), tvCL <- fixed(0, 1, 2), tvCL <- c(0, fixed(1), 2), tvCL <- c(0, 1, fixed).  'omega' assignment, full block none block must set fixed.  Examples setting 'omega' value fixed : iivKa~fixed(1), iivKa + iivCL~fixed(1, 2, 3), iivKa + iivCL~c(1, 2, 3, fixed).  Anywhere fixed used, FIX, FIXED, fix may used equivalently. value, standard mathematical operators functions may used define value.  example, log(2) 24*30 may used define value anywhere number can used (e.g. lower bound, estimate, upper bound, variance, etc.). Values may labeled using label() function assignment. Labels used make reporting easier giving human-readable description parameter, labels effect estimation.  typical way set label parameter tvCL label \"Typical Value Clearance (L/hr)\" tvCL <- 1; label(\"Typical Value Clearance (L/hr)\"). rxode2/nlmixr2 attempt determine back-transformations user.  example, CL <- exp(tvCL) detect tvCL must back-transformed exp() easier interpretation.  want control back-transformation, can specify back-transformation using backTransform() assignment.  example, set back-transformation exp(), can use tvCL <- 1; backTransform(exp()).","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/ini.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Ini block for rxode2/nlmixr models — ini.rxUi","text":"","code":"# Set the ini() block in a model one.compartment <- function() {   ini({     tka <- log(1.57); label(\"Ka\")     tcl <- log(2.72); label(\"Cl\")     tv <- log(31.5); label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  # Use piping to update initial conditions one.compartment %>% ini(tka <- log(2)) #>   #>   #> ℹ change initial estimate of `tka` to `0.693147180559945` #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.6931472 1.0006319 3.4499875 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.693147180559945 #>         label(\"Ka\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } one.compartment %>% ini(tka <- label(\"Absorption rate, Ka (1/hr)\")) #>   #>   #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.4510756 1.0006319 3.4499875 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.451075619360217 #>         label(\"Absorption rate, Ka (1/hr)\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } # Move the tka parameter to be just below the tv parameter (affects parameter # summary table, only) one.compartment %>% ini(tka <- label(\"Absorption rate, Ka (1/hr)\"), append = \"tv\") #>   #>   #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tcl        tv       tka    add.sd  #> 1.0006319 3.4499875 0.4510756 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         tka <- 0.451075619360217 #>         label(\"Absorption rate, Ka (1/hr)\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> } # When programming with rxode2/nlmixr2, it may be easier to pass strings in # to modify the ini one.compartment %>% ini(\"tka <- log(2)\") #>   #>   #> ℹ change initial estimate of `tka` to `0.693147180559945` #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.6931472 1.0006319 3.4499875 0.7000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.693147180559945 #>         label(\"Ka\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- c(0, 0.7) #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Check to see if this is an rxSolve object. — is.rxSolve","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"Check see rxSolve object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"","code":"is.rxSolve(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"x object check see rxSolve rxSolve object expired strip rxSolve information.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"boolean indicating rxSolve object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/is.rxSolve.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Check to see if this is an rxSolve object. — is.rxSolve","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"Calculate log likelihood binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"","code":"llikBeta(x, shape1, shape2, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"x Observation shape1, shape2 non-negative parameters Beta distribution. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"data frame fx log pdf value dShape1 dShape2 derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"rxode2() model, can use llikBeta() use arguments.  can also get derivative shape1 shape2 llikBetaDshape1() llikBetaDshape2().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBeta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBeta","text":"","code":"# \\donttest{  x <- seq(1e-4, 1 - 1e-4, length.out = 21)  llikBeta(x, 0.5, 0.5) #>             fx     dShape1     dShape2 #> 1   3.46049030 -7.82404601  1.38619436 #> 2   0.37793108 -1.60763953  1.33490633 #> 3   0.05888752 -0.91549105  1.28084495 #> 4  -0.11510253 -0.51035907  1.22369308 #> 5  -0.22855163 -0.22284360  1.16307581 #> 6  -0.30780832  0.00019998  1.09854562 #> 7  -0.36444410  0.18245488  1.02956227 #> 8  -0.40444714  0.33655795  0.95546529 #> 9  -0.43118004  0.47005363  0.87543540 #> 10 -0.44655956  0.58780889  0.78843918 #> 11 -0.45158271  0.69314718  0.69314718 #> 12 -0.44655956  0.78843918  0.58780889 #> 13 -0.43118004  0.87543540  0.47005363 #> 14 -0.40444714  0.95546529  0.33655795 #> 15 -0.36444410  1.02956227  0.18245488 #> 16 -0.30780832  1.09854562  0.00019998 #> 17 -0.22855163  1.16307581 -0.22284360 #> 18 -0.11510253  1.22369308 -0.51035907 #> 19  0.05888752  1.28084495 -0.91549105 #> 20  0.37793108  1.33490633 -1.60763953 #> 21  3.46049030  1.38619436 -7.82404601  llikBeta(x, 1, 3, TRUE) #>          x shape1 shape2           fx     dShape1     dShape2 #> 1  0.00010      1      3   1.09841228 -7.37700704  0.33323333 #> 2  0.05009      1      3   0.99583622 -1.16060056  0.28194530 #> 3  0.10008      1      3   0.88771347 -0.46845208  0.22788392 #> 4  0.15007      1      3   0.77340972 -0.06332009  0.17073205 #> 5  0.20006      1      3   0.65217518  0.22419538  0.11011478 #> 6  0.25005      1      3   0.52311481  0.44723895  0.04558459 #> 7  0.30004      1      3   0.38514811  0.62949385 -0.02339876 #> 8  0.35003      1      3   0.23695415  0.78359692 -0.09749574 #> 9  0.40002      1      3   0.07689437  0.91709260 -0.17752562 #> 10 0.45001      1      3  -0.09709808  1.03484786 -0.26452185 #> 11 0.50000      1      3  -0.28768207  1.14018615 -0.35981385 #> 12 0.54999      1      3  -0.49835866  1.23547815 -0.46515214 #> 13 0.59998      1      3  -0.73386918  1.32247438 -0.58290740 #> 14 0.64997      1      3  -1.00086054  1.40250426 -0.71640308 #> 15 0.69996      1      3  -1.30906667  1.47660124 -0.87050615 #> 16 0.74995      1      3  -1.67357647  1.54558459 -1.05276105 #> 17 0.79994      1      3  -2.11966363  1.61011478 -1.27580462 #> 18 0.84993      1      3  -2.69469457  1.67073205 -1.56332009 #> 19 0.89992      1      3  -3.50495854  1.72788392 -1.96845208 #> 20 0.94991      1      3  -4.88925549  1.78194530 -2.66060056 #> 21 0.99990      1      3 -17.32206846  1.83323333 -8.87700704  et <- et(seq(1e-4, 1-1e-4, length.out=21)) et$shape1 <- 0.5 et$shape2 <- 1.5  model <- function() {   model({     fx <- llikBeta(time, shape1, shape2)     dShape1 <- llikBetaDshape1(time, shape1, shape2)     dShape2 <- llikBetaDshape2(time, shape1, shape2)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 21 × 6 #>     time     fx dShape1 dShape2 shape1 shape2 #>    <dbl>  <dbl>   <dbl>   <dbl>  <dbl>  <dbl> #> 1 0.0001 4.15   -6.82    0.386     0.5    1.5 #> 2 0.0501 1.02   -0.608   0.335     0.5    1.5 #> 3 0.100  0.647   0.0845  0.281     0.5    1.5 #> 4 0.150  0.415   0.490   0.224     0.5    1.5 #> 5 0.200  0.241   0.777   0.163     0.5    1.5 #> 6 0.250  0.0976  1.00    0.0985    0.5    1.5 #> # ℹ 15 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"Calculate log likelihood binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"","code":"llikBinom(x, size, prob, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"x Number successes size Size trial prob probability success full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"data frame fx pdf value dProb derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"rxode2() model, can use llikBinom() use arguments.  can also get derivative prob llikBinomDprob()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikBinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the binomial function (and its derivatives) — llikBinom","text":"","code":"# \\donttest{ llikBinom(46:54, 100, 0.5) #>          fx dProb #> 1 -2.848030   -16 #> 2 -2.709193   -12 #> 3 -2.610102    -8 #> 4 -2.550679    -4 #> 5 -2.530876     0 #> 6 -2.550679     4 #> 7 -2.610102     8 #> 8 -2.709193    12 #> 9 -2.848030    16  llikBinom(46:54, 100, 0.5, TRUE) #>    x size prob        fx dProb #> 1 46  100  0.5 -2.848030   -16 #> 2 47  100  0.5 -2.709193   -12 #> 3 48  100  0.5 -2.610102    -8 #> 4 49  100  0.5 -2.550679    -4 #> 5 50  100  0.5 -2.530876     0 #> 6 51  100  0.5 -2.550679     4 #> 7 52  100  0.5 -2.610102     8 #> 8 53  100  0.5 -2.709193    12 #> 9 54  100  0.5 -2.848030    16  # In rxode2 you can use:  et <- et(46:54) et$size <- 100 et$prob <-0.5  model <- function() {   model({     fx <- llikBinom(time, size, prob)     dProb <- llikBinomDprob(time, size, prob)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 9 × 5 #>    time    fx dProb  size  prob #>   <dbl> <dbl> <dbl> <dbl> <dbl> #> 1    46 -2.85   -16   100   0.5 #> 2    47 -2.71   -12   100   0.5 #> 3    48 -2.61    -8   100   0.5 #> 4    49 -2.55    -4   100   0.5 #> 5    50 -2.53     0   100   0.5 #> 6    51 -2.55     4   100   0.5 #> # ℹ 3 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"log likelihood Cauchy distribution derivatives (stan)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"","code":"llikCauchy(x, location = 0, scale = 1, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"x Observation location, scale location scale parameters. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"data frame fx log pdf value dLocation dScale derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"rxode2() model, can use llikCauchy() use arguments.  can also get derivative location scale llikCauchyDlocation() llikCauchyDscale().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikCauchy.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood of Cauchy distribution and it's derivatives (from stan) — llikCauchy","text":"","code":"# \\donttest{ x <- seq(-3, 3, length.out = 21)  llikCauchy(x, 0, 1) #>           fx  dLocation     dScale #> 1  -3.447315 -0.6000000  0.8000000 #> 2  -3.259780 -0.6513872  0.7587455 #> 3  -3.055753 -0.7100592  0.7041420 #> 4  -2.832979 -0.7763401  0.6303142 #> 5  -2.589293 -0.8490566  0.5283019 #> 6  -2.323385 -0.9230769  0.3846154 #> 7  -2.036728 -0.9836066  0.1803279 #> 8  -1.738057 -0.9944751 -0.1049724 #> 9  -1.452215 -0.8823529 -0.4705882 #> 10 -1.230908 -0.5504587 -0.8348624 #> 11 -1.144730  0.0000000 -1.0000000 #> 12 -1.230908  0.5504587 -0.8348624 #> 13 -1.452215  0.8823529 -0.4705882 #> 14 -1.738057  0.9944751 -0.1049724 #> 15 -2.036728  0.9836066  0.1803279 #> 16 -2.323385  0.9230769  0.3846154 #> 17 -2.589293  0.8490566  0.5283019 #> 18 -2.832979  0.7763401  0.6303142 #> 19 -3.055753  0.7100592  0.7041420 #> 20 -3.259780  0.6513872  0.7587455 #> 21 -3.447315  0.6000000  0.8000000  llikCauchy(x, 3, 1, full=TRUE) #>       x location scale        fx  dLocation     dScale #> 1  -3.0        3     1 -4.755648 -0.3243243  0.9459459 #> 2  -2.7        3     1 -4.655977 -0.3404001  0.9402807 #> 3  -2.4        3     1 -4.551246 -0.3580902  0.9336870 #> 4  -2.1        3     1 -4.440937 -0.3776379  0.9259534 #> 5  -1.8        3     1 -4.324449 -0.3993344  0.9168053 #> 6  -1.5        3     1 -4.201087 -0.4235294  0.9058824 #> 7  -1.2        3     1 -4.070040 -0.4506438  0.8927039 #> 8  -0.9        3     1 -3.930358 -0.4811845  0.8766194 #> 9  -0.6        3     1 -3.780926 -0.5157593  0.8567335 #> 10 -0.3        3     1 -3.620428 -0.5550883  0.8317914 #> 11  0.0        3     1 -3.447315 -0.6000000  0.8000000 #> 12  0.3        3     1 -3.259780 -0.6513872  0.7587455 #> 13  0.6        3     1 -3.055753 -0.7100592  0.7041420 #> 14  0.9        3     1 -2.832979 -0.7763401  0.6303142 #> 15  1.2        3     1 -2.589293 -0.8490566  0.5283019 #> 16  1.5        3     1 -2.323385 -0.9230769  0.3846154 #> 17  1.8        3     1 -2.036728 -0.9836066  0.1803279 #> 18  2.1        3     1 -1.738057 -0.9944751 -0.1049724 #> 19  2.4        3     1 -1.452215 -0.8823529 -0.4705882 #> 20  2.7        3     1 -1.230908 -0.5504587 -0.8348624 #> 21  3.0        3     1 -1.144730  0.0000000 -1.0000000  et <- et(-3, 3, length.out=10) et$location <- 0 et$scale <- 1  model <- function() {   model({     fx <- llikCauchy(time, location, scale)     dLocation <- llikCauchyDlocation(time, location, scale)     dScale <- llikCauchyDscale(time, location, scale)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> Warning:  #> with negative times, compartments initialize at first negative observed time #> with positive times, compartments initialize at time zero #> use 'rxSetIni0(FALSE)' to initialize at first observed time #> this warning is displayed once per session #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>     time    fx dLocation dScale location scale #>    <dbl> <dbl>     <dbl>  <dbl>    <dbl> <dbl> #> 1 -3     -3.45    -0.6    0.8          0     1 #> 2 -2.33  -3.01    -0.724  0.690        0     1 #> 3 -1.67  -2.47    -0.882  0.471        0     1 #> 4 -1     -1.84    -1      0            0     1 #> 5 -0.333 -1.25    -0.6   -0.8          0     1 #> 6  0.333 -1.25     0.600 -0.8          0     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for chi-squared distribution — llikChisq","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"log likelihood derivatives chi-squared distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"","code":"llikChisq(x, df, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"x variable distributed chi-squared distribution df degrees freedom (non-negative, can non-integer). full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"data frame fx log pdf value dDf derivatives respect df parameter observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"rxode2() model, can use llikChisq() use x df arguments.  can also get derivative df llikChisqDdf().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikChisq.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for chi-squared distribution — llikChisq","text":"","code":"# \\donttest{ llikChisq(1, df = 1:3, full=TRUE) #>   x df        fx         dDf #> 1 1  1 -1.418939  0.63518142 #> 2 1  2 -1.193147 -0.05796576 #> 3 1  3 -1.418939 -0.36481858  llikChisq(1, df = 6:9) #>          fx        dDf #> 1 -3.272589 -0.8079658 #> 2 -4.126989 -0.8981519 #> 3 -5.064348 -0.9746324 #> 4 -6.072899 -1.0410091  et <- et(1:3) et$x <- 1  model <- function() {   model({    fx <- llikChisq(x, time)    dDf <- llikChisqDdf(x, time)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 3 × 4 #>    time    fx     dDf     x #>   <dbl> <dbl>   <dbl> <dbl> #> 1     1 -1.42  0.635      1 #> 2     2 -1.19 -0.0580     1 #> 3     3 -1.42 -0.365      1 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for exponential distribution — llikExp","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"log likelihood derivatives exponential distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"","code":"llikExp(x, rate, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"x variable distributed exponential distribution rate vector rates. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"data frame fx log pdf value dRate derivatives respect rate parameter observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"rxode2() model, can use llikExp() use x rate arguments.  can also get derivative rate llikExpDrate().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikExp.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for exponential distribution — llikExp","text":"","code":"# \\donttest{ llikExp(1, 1:3) #>          fx      dRate #> 1 -1.000000  0.0000000 #> 2 -1.306853 -0.5000000 #> 3 -1.901388 -0.6666667  llikExp(1, 1:3, full=TRUE) #>   x rate        fx      dRate #> 1 1    1 -1.000000  0.0000000 #> 2 1    2 -1.306853 -0.5000000 #> 3 1    3 -1.901388 -0.6666667  # You can use rxode2 for these too:  et <- et(1:3) et$x <- 1  model <- function() {   model({     fx <- llikExp(x, time)     dRate <- llikExpDrate(x, time)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 3 × 4 #>    time    fx  dRate     x #>   <dbl> <dbl>  <dbl> <dbl> #> 1     1 -1     0         1 #> 2     2 -1.31 -0.5       1 #> 3     3 -1.90 -0.667     1 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for F distribution — llikF","title":"log likelihood and derivatives for F distribution — llikF","text":"log likelihood derivatives F distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for F distribution — llikF","text":"","code":"llikF(x, df1, df2, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for F distribution — llikF","text":"x variable distributed f distribution df1, df2 degrees freedom.  Inf allowed. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for F distribution — llikF","text":"data frame fx log pdf value dDf1 dDf2 derivatives respect df1/df2 parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for F distribution — llikF","text":"rxode2() model, can use llikF() use x rate arguments.  can also get derivative df1 df2 llikFDdf1() llikFDdf2().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for F distribution — llikF","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikF.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for F distribution — llikF","text":"","code":"# \\donttest{ x <- seq(0.001, 5, length.out = 100)  llikF(x^2, 1, 5) #>               fx         dDf1          dDf2 #> 1    5.939135090 -5.769327755  0.0098138672 #> 2    1.996061339 -1.829698617  0.0098667834 #> 3    1.308027230 -1.151720596  0.0100203751 #> 4    0.898151814 -0.758467992  0.0102713542 #> 5    0.601410048 -0.484842238  0.0106143384 #> 6    0.365567397 -0.278507461  0.0110419510 #> 7    0.167275143 -0.115987118  0.0115449567 #> 8   -0.005928407  0.015333168  0.0121124287 #> 9   -0.161467803  0.123053577  0.0127319420 #> 10  -0.304110944  0.212138450  0.0133897879 #> 11  -0.437089645  0.286034532  0.0140712037 #> 12  -0.562684604  0.347253504  0.0147606108 #> 13  -0.682555520  0.397699568  0.0154418570 #> 14  -0.797939098  0.438864917  0.0160984557 #> 15  -0.909773848  0.471951996  0.0167138175 #> 16  -1.018782007  0.497952966  0.0172714700 #> 17  -1.125525217  0.517703019  0.0177552614 #> 18  -1.230443500  0.531917117  0.0181495454 #> 19  -1.333883243  0.541215913  0.0184393449 #> 20  -1.436117763  0.546144408  0.0186104926 #> 21  -1.537362702  0.547185650  0.0186497493 #> 22  -1.637787764  0.544770962  0.0185448973 #> 23  -1.737525805  0.539287723  0.0182848126 #> 24  -1.836679954  0.531085385  0.0178595139 #> 25  -1.935329266  0.520480216  0.0172601925 #> 26  -2.033533263  0.507759100  0.0164792220 #> 27  -2.131335592  0.493182655  0.0155101530 #> 28  -2.228767013  0.476987816  0.0143476919 #> 29  -2.325847849  0.459390036  0.0129876676 #> 30  -2.422589998  0.440585182  0.0114269877 #> 31  -2.518998603  0.420751193  0.0096635860 #> 32  -2.615073425  0.400049565  0.0076963639 #> 33  -2.710810000  0.378626673  0.0055251263 #> 34  -2.806200581  0.356614982  0.0031505147 #> 35  -2.901234940  0.334134152  0.0005739382 #> 36  -2.995901017  0.311292055 -0.0022024960 #> 37  -3.090185467  0.288185720 -0.0051760516 #> 38  -3.184074111  0.264902202 -0.0083434311 #> 39  -3.277552304  0.241519400 -0.0117008417 #> 40  -3.370605236  0.218106810 -0.0152440579 #> 41  -3.463218177  0.194726229 -0.0189684823 #> 42  -3.555376677  0.171432411 -0.0228692014 #> 43  -3.647066718  0.148273673 -0.0269410395 #> 44  -3.738274839  0.125292465 -0.0311786073 #> 45  -3.828988227  0.102525891 -0.0355763478 #> 46  -3.919194789  0.080006197 -0.0401285781 #> 47  -4.008883199  0.057761220 -0.0448295271 #> 48  -4.098042931  0.035814802 -0.0496733702 #> 49  -4.186664276  0.014187176 -0.0546542601 #> 50  -4.274738350 -0.007104684 -0.0597663547 #> 51  -4.362257091 -0.028046739 -0.0650038410 #> 52  -4.449213248 -0.048627585 -0.0703609577 #> 53  -4.535600366 -0.068838184 -0.0758320130 #> 54  -4.621412763 -0.088671614 -0.0814114015 #> 55  -4.706645505 -0.108122840 -0.0870936184 #> 56  -4.791294380 -0.127188508 -0.0928732707 #> 57  -4.875355868 -0.145866755 -0.0987450876 #> 58  -4.958827105 -0.164157039 -0.1047039284 #> 59  -5.041705859 -0.182059980 -0.1107447891 #> 60  -5.123990491 -0.199577223 -0.1168628074 #> 61  -5.205679923 -0.216711307 -0.1230532668 #> 62  -5.286773611 -0.233465550 -0.1293115988 #> 63  -5.367271506 -0.249843943 -0.1356333846 #> 64  -5.447174028 -0.265851061 -0.1420143563 #> 65  -5.526482033 -0.281491973 -0.1484503963 #> 66  -5.605196787 -0.296772169 -0.1549375368 #> 67  -5.683319932 -0.311697491 -0.1614719584 #> 68  -5.760853464 -0.326274073 -0.1680499885 #> 69  -5.837799706 -0.340508283 -0.1746680988 #> 70  -5.914161282 -0.354406680 -0.1813229026 #> 71  -5.989941093 -0.367975967 -0.1880111521 #> 72  -6.065142296 -0.381222952 -0.1947297352 #> 73  -6.139768282 -0.394154517 -0.2014756717 #> 74  -6.213822655 -0.406777586 -0.2082461105 #> 75  -6.287309215 -0.419099097 -0.2150383254 #> 76  -6.360231938 -0.431125983 -0.2218497115 #> 77  -6.432594959 -0.442865149 -0.2286777813 #> 78  -6.504402560 -0.454323453 -0.2355201612 #> 79  -6.575659147 -0.465507693 -0.2423745875 #> 80  -6.646369247 -0.476424593 -0.2492389023 #> 81  -6.716537483 -0.487080791 -0.2561110501 #> 82  -6.786168572 -0.497482830 -0.2629890742 #> 83  -6.855267306 -0.507637151 -0.2698711122 #> 84  -6.923838547 -0.517550083 -0.2767553934 #> 85  -6.991887212 -0.527227840 -0.2836402345 #> 86  -7.059418267 -0.536676516 -0.2905240366 #> 87  -7.126436719 -0.545902081 -0.2974052813 #> 88  -7.192947604 -0.554910380 -0.3042825282 #> 89  -7.258955983 -0.563707126 -0.3111544109 #> 90  -7.324466935 -0.572297907 -0.3180196342 #> 91  -7.389485548 -0.580688177 -0.3248769714 #> 92  -7.454016916 -0.588883261 -0.3317252608 #> 93  -7.518066132 -0.596888352 -0.3385634034 #> 94  -7.581638284 -0.604708517 -0.3453903600 #> 95  -7.644738447 -0.612348689 -0.3522051484 #> 96  -7.707371685 -0.619813679 -0.3590068413 #> 97  -7.769543041 -0.627108167 -0.3657945633 #> 98  -7.831257537 -0.634236713 -0.3725674892 #> 99  -7.892520171 -0.641203751 -0.3793248413 #> 100 -7.953335909 -0.648013598 -0.3860658874  model <- function(){   model({     fx <- llikF(time, df1, df2)     dMean <- llikFDdf1(time, df1, df2)     dSd <- llikFDdf2(time, df1, df2)   }) }  et <- et(x) et$df1 <- 1 et$df2 <- 5  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 100 × 6 #>     time     fx   dMean     dSd   df1   df2 #>    <dbl>  <dbl>   <dbl>   <dbl> <dbl> <dbl> #> 1 0.001   2.48  -2.32   0.00983     1     5 #> 2 0.0515  0.484 -0.380  0.0108      1     5 #> 3 0.102   0.112 -0.0731 0.0117      1     5 #> 4 0.152  -0.118  0.0943 0.0125      1     5 #> 5 0.203  -0.291  0.204  0.0133      1     5 #> 6 0.253  -0.431  0.283  0.0140      1     5 #> # ℹ 94 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Gamma distribution — llikGamma","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"log likelihood derivatives Gamma distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"","code":"llikGamma(x, shape, rate, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"x variable distributed gamma distribution shape distribution's shape parameter. Must positive. rate distribution's rate parameters.  Must positive. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"rxode2() model, can use llikGamma() use x rate arguments.  can also get derivative shape rate llikGammaDshape() llikGammaDrate().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Gamma distribution — llikGamma","text":"","code":"# \\donttest{  llikGamma(1, 1, 10) #>          fx   dShape dRate #> 1 -7.697415 2.879801  -0.9  # You can use this in `rxode2` too:  et  <- et(seq(0.001, 1, length.out=10)) et$shape <- 1 et$rate <- 10   model <- function() {   model({     fx <- llikGamma(time, shape, rate)     dShape<- llikGammaDshape(time, shape, rate)     dRate <- llikGammaDrate(time, shape, rate)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>    time      fx dShape  dRate  rate shape #>   <dbl>   <dbl>  <dbl>  <dbl> <dbl> <dbl> #> 1 0.001  2.29   -4.03   0.099    10     1 #> 2 0.112  1.18    0.691 -0.012    10     1 #> 3 0.223  0.0726  1.38  -0.123    10     1 #> 4 0.334 -1.04    1.78  -0.234    10     1 #> 5 0.445 -2.15    2.07  -0.345    10     1 #> 6 0.556 -3.26    2.29  -0.456    10     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Geom distribution — llikGeom","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"log likelihood derivatives Geom distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"","code":"llikGeom(x, prob, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"x variable distributed geom distribution prob probability success trial. 0 < prob <= 1. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"rxode2() model, can use llikGeom() use x rate arguments.  can also get derivative prob llikGeomDprob().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikGeom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Geom distribution — llikGeom","text":"","code":"# \\donttest{  llikGeom(1:10, 0.2) #>           fx dProb #> 1  -1.832581  3.75 #> 2  -2.055725  2.50 #> 3  -2.278869  1.25 #> 4  -2.502012  0.00 #> 5  -2.725156 -1.25 #> 6  -2.948299 -2.50 #> 7  -3.171443 -3.75 #> 8  -3.394586 -5.00 #> 9  -3.617730 -6.25 #> 10 -3.840873 -7.50  et  <- et(1:10) et$prob <- 0.2   model <- function() {   model({     fx <- llikGeom(time, prob)     dProb <- llikGeomDprob(time, prob)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time    fx dProb  prob #>   <dbl> <dbl> <dbl> <dbl> #> 1     1 -1.83  3.75   0.2 #> 2     2 -2.06  2.5    0.2 #> 3     3 -2.28  1.25   0.2 #> 4     4 -2.50  0      0.2 #> 5     5 -2.73 -1.25   0.2 #> 6     6 -2.95 -2.5    0.2 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"Calculate log likelihood negative binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"","code":"llikNbinom(x, size, prob, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"x Number successes size Size trial prob probability success full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"data frame fx pdf value dProb derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"rxode2() model, can use llikNbinom() use arguments.  can also get derivative prob llikNbinomDprob()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinom","text":"","code":"# \\donttest{ llikNbinom(46:54, 100, 0.5) #>          fx dProb #> 1 -13.25200   108 #> 2 -12.81168   106 #> 3 -12.38560   104 #> 4 -11.97335   102 #> 5 -11.57458   100 #> 6 -11.18892    98 #> 7 -10.81603    96 #> 8 -10.45559    94 #> 9 -10.10728    92  llikNbinom(46:54, 100, 0.5, TRUE) #>    x size prob        fx dProb #> 1 46  100  0.5 -13.25200   108 #> 2 47  100  0.5 -12.81168   106 #> 3 48  100  0.5 -12.38560   104 #> 4 49  100  0.5 -11.97335   102 #> 5 50  100  0.5 -11.57458   100 #> 6 51  100  0.5 -11.18892    98 #> 7 52  100  0.5 -10.81603    96 #> 8 53  100  0.5 -10.45559    94 #> 9 54  100  0.5 -10.10728    92  # In rxode2 you can use:  et <- et(46:54) et$size <- 100 et$prob <-0.5  model <- function() {   model({     fx <- llikNbinom(time, size, prob)     dProb <- llikNbinomDprob(time, size, prob)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 9 × 5 #>    time    fx dProb  size  prob #>   <dbl> <dbl> <dbl> <dbl> <dbl> #> 1    46 -13.3   108   100   0.5 #> 2    47 -12.8   106   100   0.5 #> 3    48 -12.4   104   100   0.5 #> 4    49 -12.0   102   100   0.5 #> 5    50 -11.6   100   100   0.5 #> 6    51 -11.2    98   100   0.5 #> # ℹ 3 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"Calculate log likelihood negative binomial function (derivatives)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"","code":"llikNbinomMu(x, size, mu, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"x Number successes size Size trial mu mu parameter negative binomial full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"data frame fx pdf value dProb derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"rxode2() model, can use llikNbinomMu() use arguments.  can also get derivative mu llikNbinomMuDmu()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNbinomMu.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the log likelihood of the negative binomial function (and its derivatives) — llikNbinomMu","text":"","code":"# \\donttest{ llikNbinomMu(46:54, 100, 40) #>          fx       dMu #> 1 -3.326828 0.1071429 #> 2 -3.446132 0.1250000 #> 3 -3.579663 0.1428571 #> 4 -3.727034 0.1607143 #> 5 -3.887874 0.1785714 #> 6 -4.061827 0.1964286 #> 7 -4.248554 0.2142857 #> 8 -4.447728 0.2321429 #> 9 -4.659037 0.2500000  llikNbinomMu(46:54, 100, 40, TRUE) #>    x size mu        fx       dMu #> 1 46  100 40 -3.326828 0.1071429 #> 2 47  100 40 -3.446132 0.1250000 #> 3 48  100 40 -3.579663 0.1428571 #> 4 49  100 40 -3.727034 0.1607143 #> 5 50  100 40 -3.887874 0.1785714 #> 6 51  100 40 -4.061827 0.1964286 #> 7 52  100 40 -4.248554 0.2142857 #> 8 53  100 40 -4.447728 0.2321429 #> 9 54  100 40 -4.659037 0.2500000  et <- et(46:54) et$size <- 100 et$mu <- 40  model <- function() {   model({     fx <- llikNbinomMu(time, size, mu)     dProb <- llikNbinomMuDmu(time, size, mu)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 9 × 5 #>    time    fx dProb  size    mu #>   <dbl> <dbl> <dbl> <dbl> <dbl> #> 1    46 -3.33 0.107   100    40 #> 2    47 -3.45 0.125   100    40 #> 3    48 -3.58 0.143   100    40 #> 4    49 -3.73 0.161   100    40 #> 5    50 -3.89 0.179   100    40 #> 6    51 -4.06 0.196   100    40 #> # ℹ 3 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":null,"dir":"Reference","previous_headings":"","what":"Log likelihood for normal distribution — llikNorm","title":"Log likelihood for normal distribution — llikNorm","text":"Log likelihood normal distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Log likelihood for normal distribution — llikNorm","text":"","code":"llikNorm(x, mean = 0, sd = 1, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Log likelihood for normal distribution — llikNorm","text":"x Observation mean Mean likelihood sd Standard deviation likelihood full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Log likelihood for normal distribution — llikNorm","text":"data frame fx pdf value dMean dSd derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Log likelihood for normal distribution — llikNorm","text":"rxode2() model, can use llikNorm() use arguments.  can also get derivatives llikNormDmean() llikNormDsd()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Log likelihood for normal distribution — llikNorm","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikNorm.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Log likelihood for normal distribution — llikNorm","text":"","code":"# \\donttest{  llikNorm(0) #>           fx dMean dSd #> 1 -0.9189385     0  -1  llikNorm(seq(-2,2,length.out=10), full=TRUE) #>             x mean sd         fx      dMean        dSd #> 1  -2.0000000    0  1 -2.9189385 -2.0000000  3.0000000 #> 2  -1.5555556    0  1 -2.1288151 -1.5555556  1.4197531 #> 3  -1.1111111    0  1 -1.5362225 -1.1111111  0.2345679 #> 4  -0.6666667    0  1 -1.1411608 -0.6666667 -0.5555556 #> 5  -0.2222222    0  1 -0.9436299 -0.2222222 -0.9506173 #> 6   0.2222222    0  1 -0.9436299  0.2222222 -0.9506173 #> 7   0.6666667    0  1 -1.1411608  0.6666667 -0.5555556 #> 8   1.1111111    0  1 -1.5362225  1.1111111  0.2345679 #> 9   1.5555556    0  1 -2.1288151  1.5555556  1.4197531 #> 10  2.0000000    0  1 -2.9189385  2.0000000  3.0000000  # With rxode2 you can use:   et <- et(-3, 3, length.out=10) et$mu <- 0 et$sigma <- 1  model <- function(){   model({     fx <- llikNorm(time, mu, sigma)     dMean <- llikNormDmean(time, mu, sigma)     dSd <- llikNormDsd(time, mu, sigma)   })  }  ret <- rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ret #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>     time     fx  dMean    dSd    mu sigma #>    <dbl>  <dbl>  <dbl>  <dbl> <dbl> <dbl> #> 1 -3     -5.42  -3      8         0     1 #> 2 -2.33  -3.64  -2.33   4.44      0     1 #> 3 -1.67  -2.31  -1.67   1.78      0     1 #> 4 -1     -1.42  -1      0         0     1 #> 5 -0.333 -0.974 -0.333 -0.889     0     1 #> 6  0.333 -0.974  0.333 -0.889     0     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":null,"dir":"Reference","previous_headings":"","what":"log-likelihood for the Poisson distribution — llikPois","title":"log-likelihood for the Poisson distribution — llikPois","text":"log-likelihood Poisson distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log-likelihood for the Poisson distribution — llikPois","text":"","code":"llikPois(x, lambda, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log-likelihood for the Poisson distribution — llikPois","text":"x non negative integers lambda non-negative means full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log-likelihood for the Poisson distribution — llikPois","text":"data frame fx pdf value dLambda derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log-likelihood for the Poisson distribution — llikPois","text":"rxode2() model, can use llikPois() use arguments.  can also get derivatives llikPoisDlambda()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log-likelihood for the Poisson distribution — llikPois","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikPois.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log-likelihood for the Poisson distribution — llikPois","text":"","code":"# \\donttest{ llikPois(0:7, lambda = 1) #>          fx dLambda #> 1 -1.000000      -1 #> 2 -1.000000       0 #> 3 -1.693147       1 #> 4 -2.791759       2 #> 5 -4.178054       3 #> 6 -5.787492       4 #> 7 -7.579251       5 #> 8 -9.525161       6  llikPois(0:7, lambda = 4, full=TRUE) #>   x lambda        fx dLambda #> 1 0      4 -4.000000   -1.00 #> 2 1      4 -2.613706   -0.75 #> 3 2      4 -1.920558   -0.50 #> 4 3      4 -1.632876   -0.25 #> 5 4      4 -1.632876    0.00 #> 6 5      4 -1.856020    0.25 #> 7 6      4 -2.261485    0.50 #> 8 7      4 -2.821101    0.75  # In rxode2 you can use:  et <- et(0:10) et$lambda <- 0.5  model <- function() {   model({     fx <- llikPois(time, lambda)     dLambda <- llikPoisDlambda(time, lambda)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 11 × 4 #>    time    fx dLambda lambda #>   <dbl> <dbl>   <dbl>  <dbl> #> 1     0 -0.5       -1    0.5 #> 2     1 -1.19       1    0.5 #> 3     2 -2.58       3    0.5 #> 4     3 -4.37       5    0.5 #> 5     4 -6.45       7    0.5 #> 6     5 -8.75       9    0.5 #> # ℹ 5 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":null,"dir":"Reference","previous_headings":"","what":"Log likelihood of T and it's derivatives (from stan) — llikT","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"Log likelihood T derivatives (stan)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"","code":"llikT(x, df, mean = 0, sd = 1, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"x Observation df degrees freedom (\\(> 0\\), maybe non-integer).  df       = Inf allowed. mean Mean likelihood sd Standard deviation likelihood full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"data frame fx log pdf value dDf dMean dSd derivatives respect parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"rxode2() model, can use llikT() use arguments.  can also get derivative df, mean sd llikTDdf(), llikTDmean() llikTDsd().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikT.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Log likelihood of T and it's derivatives (from stan) — llikT","text":"","code":"# \\donttest{ x <- seq(-3, 3, length.out = 21)  llikT(x, 7, 0, 1) #>            fx          dDf      dMean        dSd #> 1  -4.2612484 -0.086858773 -1.5000000  3.5000000 #> 2  -3.8091335 -0.060260433 -1.5115465  3.0811756 #> 3  -3.3561547 -0.037201674 -1.5047022  2.6112853 #> 4  -2.9088542 -0.018379169 -1.4723926  2.0920245 #> 5  -2.4761000 -0.004340221 -1.4062500  1.5312500 #> 6  -2.0693878  0.004691380 -1.2972973  0.9459459 #> 7  -1.7028228  0.009010785 -1.1374408  0.3649289 #> 8  -1.3925134  0.009569214 -0.9218950 -0.1702945 #> 9  -1.1551333  0.007927361 -0.6521739 -0.6086957 #> 10 -1.0056349  0.005918024 -0.3385049 -0.8984485 #> 11 -0.9545342  0.005051942  0.0000000 -1.0000000 #> 12 -1.0056349  0.005918024  0.3385049 -0.8984485 #> 13 -1.1551333  0.007927361  0.6521739 -0.6086957 #> 14 -1.3925134  0.009569214  0.9218950 -0.1702945 #> 15 -1.7028228  0.009010785  1.1374408  0.3649289 #> 16 -2.0693878  0.004691380  1.2972973  0.9459459 #> 17 -2.4761000 -0.004340221  1.4062500  1.5312500 #> 18 -2.9088542 -0.018379169  1.4723926  2.0920245 #> 19 -3.3561547 -0.037201674  1.5047022  2.6112853 #> 20 -3.8091335 -0.060260433  1.5115465  3.0811756 #> 21 -4.2612484 -0.086858773  1.5000000  3.5000000  llikT(x, 15, 0, 1, full=TRUE) #>            fx           dDf      dMean        dSd #> 1  -4.6956220 -0.0338931511 -2.0000000  5.0000000 #> 2  -4.1042965 -0.0225073150 -1.9380888  4.2328398 #> 3  -3.5354158 -0.0134033865 -1.8497110  3.4393064 #> 4  -2.9974985 -0.0065857823 -1.7310665  2.6352396 #> 5  -2.5001272 -0.0019378861 -1.5789474  1.8421053 #> 6  -2.0536885  0.0007929097 -1.3913043  1.0869565 #> 7  -1.6689304  0.0019903977 -1.1678832  0.4014599 #> 8  -1.3563325  0.0021369166 -0.9108159 -0.1802657 #> 9  -1.1253251  0.0017504002 -0.6250000 -0.6250000 #> 10 -0.9834495  0.0012985422 -0.3180915 -0.9045726 #> 11 -0.9355929  0.0011086635  0.0000000 -1.0000000 #> 12 -0.9834495  0.0012985422  0.3180915 -0.9045726 #> 13 -1.1253251  0.0017504002  0.6250000 -0.6250000 #> 14 -1.3563325  0.0021369166  0.9108159 -0.1802657 #> 15 -1.6689304  0.0019903977  1.1678832  0.4014599 #> 16 -2.0536885  0.0007929097  1.3913043  1.0869565 #> 17 -2.5001272 -0.0019378861  1.5789474  1.8421053 #> 18 -2.9974985 -0.0065857823  1.7310665  2.6352396 #> 19 -3.5354158 -0.0134033865  1.8497110  3.4393064 #> 20 -4.1042965 -0.0225073150  1.9380888  4.2328398 #> 21 -4.6956220 -0.0338931511  2.0000000  5.0000000  et <- et(-3, 3, length.out=10) et$nu <- 7 et$mean <- 0 et$sd <- 1  model <- function() {   model({     fx <- llikT(time, nu, mean, sd)     dDf <- llikTDdf(time, nu, mean, sd)     dMean <- llikTDmean(time, nu, mean, sd)     dSd   <- llikTDsd(time, nu, mean, sd)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 8 #>     time    fx       dDf  dMean    dSd    nu  mean    sd #>    <dbl> <dbl>     <dbl>  <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1 -3     -4.26 -0.0869   -1.5    3.5       7     0     1 #> 2 -2.33  -3.26 -0.0326   -1.5    2.5       7     0     1 #> 3 -1.67  -2.29  0.000289 -1.36   1.27      7     0     1 #> 4 -1     -1.49  0.00971  -1      0         7     0     1 #> 5 -0.333 -1.02  0.00611  -0.375 -0.875     7     0     1 #> 6  0.333 -1.02  0.00611   0.375 -0.875     7     0     1 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Unif distribution — llikUnif","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"log likelihood derivatives Unif distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"","code":"llikUnif(x, alpha, beta, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"x variable distributed uniform distribution alpha lower limit uniform distribution beta upper limit distribution full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"rxode2() model, can use llikUnif() use x rate arguments.  can also get derivative alpha beta llikUnifDalpha() llikUnifDbeta().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikUnif.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Unif distribution — llikUnif","text":"","code":"# \\donttest{  llikUnif(1, -2, 2) #>          fx dAlpha dBeta #> 1 -1.386294   0.25 -0.25  et  <- et(seq(1,1, length.out=4)) et$alpha <- -2 et$beta <- 2   model <- function() {   model({     fx <- llikUnif(time, alpha, beta)     dAlpha<- llikUnifDalpha(time, alpha, beta)     dBeta <- llikUnifDbeta(time, alpha, beta)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 4 × 6 #>    time    fx dAlpha dBeta alpha  beta #>   <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> #> 1     1 -1.39   0.25 -0.25    -2     2 #> 2     1 -1.39   0.25 -0.25    -2     2 #> 3     1 -1.39   0.25 -0.25    -2     2 #> 4     1 -1.39   0.25 -0.25    -2     2 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":null,"dir":"Reference","previous_headings":"","what":"log likelihood and derivatives for Weibull distribution — llikWeibull","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"log likelihood derivatives Weibull distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"","code":"llikWeibull(x, shape, scale, full = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"x variable distributed Weibull distribution shape, scale shape scale parameters, latter defaulting 1. full Add data frame showing x, mean, sd well fx derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"data frame fx log pdf value dProb derivatives respect prob parameters observation time-point","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"rxode2() model, can use llikWeibull() use x rate arguments.  can also get derivative shape scale llikWeibullDshape() llikWeibullDscale().","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/llikWeibull.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"log likelihood and derivatives for Weibull distribution — llikWeibull","text":"","code":"# \\donttest{ llikWeibull(1, 1, 10) #>          fx    dShape dScale #> 1 -2.402585 -1.072327  -0.09  # rxode2 can use this too:  et  <- et(seq(0.001, 1, length.out=10)) et$shape <- 1 et$scale <- 10   model <- function() {   model({     fx <- llikWeibull(time, shape, scale)     dShape<- llikWeibullDshape(time, shape, scale)     dScale <- llikWeibullDscale(time, shape, scale)   }) }  rxSolve(model, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 6 #>    time    fx dShape  dScale shape scale #>   <dbl> <dbl>  <dbl>   <dbl> <dbl> <dbl> #> 1 0.001 -2.30  -8.21 -0.100      1    10 #> 2 0.112 -2.31  -3.44 -0.0989     1    10 #> 3 0.223 -2.32  -2.72 -0.0978     1    10 #> 4 0.334 -2.34  -2.29 -0.0967     1    10 #> 5 0.445 -2.35  -1.97 -0.0956     1    10 #> 6 0.556 -2.36  -1.73 -0.0944     1    10 #> # ℹ 4 more rows # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":null,"dir":"Reference","previous_headings":"","what":"logit and inverse logit (expit) functions — logit","title":"logit and inverse logit (expit) functions — logit","text":"logit inverse logit (expit) functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"logit and inverse logit (expit) functions — logit","text":"","code":"logit(x, low = 0, high = 1)  expit(alpha, low = 0, high = 1)  logitNormInfo(mean = 0, sd = 1, low = 0, high = 1, abs.tol = 1e-06, ...)  probitNormInfo(mean = 0, sd = 1, low = 0, high = 1, abs.tol = 1e-06, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"logit and inverse logit (expit) functions — logit","text":"x Input value(s) range [low,high] translate -Inf Inf low Lowest value range high Highest value range alpha Infinite value(s) translate range [low, high] mean logit-scale mean sd logit-scale standard deviation abs.tol absolute accuracy requested. ... parameters passed integrate()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"logit and inverse logit (expit) functions — logit","text":"values logit expit","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"logit and inverse logit (expit) functions — logit","text":"logit given : logit(p) = -log(1/p-1) : p = x-low/high-low expit given : expit(p, low, high) = (high-low)/(1+exp(-alpha)) + low logitNormInfo() gives mean, variance coefficient variability untransformed scale.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/logit.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"logit and inverse logit (expit) functions — logit","text":"","code":"logit(0.25) #> [1] -1.098612  expit(-1.09) #> [1] 0.2516183  logitNormInfo(logit(0.25), sd = 0.1) #>         mean          var           cv  #> 0.2504672899 0.0003515538 0.0748591440   logitNormInfo(logit(1, 0, 10), sd = 1, low = 0, high = 10) #>      mean       var        cv  #> 1.3386966 1.2662276 0.8405697"},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":null,"dir":"Reference","previous_headings":"","what":"lowergamma: upper incomplete gamma function — lowergamma","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"tgamma_lower boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"","code":"lowergamma(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"numeric '' parameter upper incomplete gamma z numeric 'z' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"lowergamma results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"lowergamma function given : $$lowergamma(, z) = \\int_{0}^{z}t^{-1}\\cdot e^{-t} dt$$","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/lowergamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"lowergamma: upper incomplete gamma function — lowergamma","text":"","code":"lowergamma(1, 3) #> [1] 0.9502129  lowergamma(1:3, 3) #> [1] 0.9502129 0.8008517 1.1536198  lowergamma(1, 1:3) #> [1] 0.6321206 0.8646647 0.9502129"},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"generic function meanProbs produces expected confidence bands either t distribution normal sampling distribution. uses qnorm() qt() mean standard deviation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"","code":"meanProbs(x, ...)  # S3 method for default meanProbs(   x,   probs = seq(0, 1, 0.25),   na.rm = FALSE,   names = TRUE,   useT = TRUE,   onlyProbs = TRUE,   pred = FALSE,   n = 0L,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"x numeric vector whose mean probability based confidence values wanted, NA NaN values allowed numeric vectors unless ‘na.rm’ ‘TRUE’. ... Arguments passed default method, allows many different methods applied. probs numeric vector probabilities values interval 0 1 . na.rm logical; true, NA NaN's removed x quantiles computed. names logical; true, result names attribute. useT logical; true, use t-distribution calculate confidence-based estimates. false use normal distribution calculate confidence based estimates. onlyProbs logical; true, return probability based confidence interval estimates, otherwise return pred logical; true use prediction interval instead confidence interval n integer/integerish; n used calculate prediction confidence interval.  n=0 (default) use number non-NA observations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"default return probabilities names (named) points expected distribution located given sampling mean standard deviation. onlyProbs=FALSE prepend mean, variance, standard deviation, minimum, maximum number non-NA observations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"single probability, p, uses either: mean + qt(p, df=n)*sd/sqrt(n) mean + qnorm(p)*sd/sqrt(n) smallest observation corresponds probability 0 largest probability 1 mean corresponds 0.5. mean standard deviation sample calculated based Welford's method single pass. meant perform way quantile() can drop replacement code using quantile() using distributional assumptions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/meanProbs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate expected confidence bands or prediction intreval with normal or t sampling distribution — meanProbs","text":"","code":"quantile(x<- rnorm(1001)) #>           0%          25%          50%          75%         100%  #> -3.403225585 -0.642325341 -0.001837139  0.656154718  2.780013343  meanProbs(x) #>           0%          25%          50%          75%         100%  #> -3.403225585 -0.023025596 -0.002431956  0.018161685  2.780013343   # Can get some extra statistics if you request onlyProbs=FALSE meanProbs(x, onlyProbs=FALSE) #>          mean           var            sd           min           max  #> -2.431956e-03  9.324674e-01  9.656435e-01 -3.403226e+00  2.780013e+00  #>             n            0%           25%           50%           75%  #>  1.001000e+03 -3.403226e+00 -2.302560e-02 -2.431956e-03  1.816168e-02  #>          100%  #>  2.780013e+00   x[2] <- NA_real_  meanProbs(x, onlyProbs=FALSE) #> mean  var   sd  min  max    n   0%  25%  50%  75% 100%  #>   NA   NA   NA   NA   NA   NA   NA   NA   NA   NA   NA   quantile(x<- rnorm(42)) #>         0%        25%        50%        75%       100%  #> -2.1838542 -0.8532155  0.1476959  0.8088171  1.5847511   meanProbs(x) #>          0%         25%         50%         75%        100%  #> -2.18385418 -0.18647291 -0.07532481  0.03582330  1.58475107   meanProbs(x, useT=FALSE) #>          0%         25%         50%         75%        100%  #> -2.18385418 -0.18548788 -0.07532481  0.03483827  1.58475107"},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign the model block in the rxode2 related object — model<-","title":"Assign the model block in the rxode2 related object — model<-","text":"Assign model block rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign the model block in the rxode2 related object — model<-","text":"","code":"model(x, envir = environment(x)) <- value"},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign the model block in the rxode2 related object — model<-","text":"x rxode2 related object envir Environment assignment occurs value Value object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign the model block in the rxode2 related object — model<-","text":"rxode2 related object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model-set.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign the model block in the rxode2 related object — model<-","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":null,"dir":"Reference","previous_headings":"","what":"Model block for rxode2/nlmixr models — model.function","title":"Model block for rxode2/nlmixr models — model.function","text":"Model block rxode2/nlmixr models","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Model block for rxode2/nlmixr models — model.function","text":"","code":"# S3 method for `function` model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for rxUi model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for rxode2 model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for rxModelVars model(   x,   ...,   append = NULL,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  model(   x,   ...,   append = FALSE,   auto = getOption(\"rxode2.autoVarPiping\", TRUE),   cov = NULL,   envir = parent.frame() )  # S3 method for default model(x, ..., append = FALSE, cov = NULL, envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Model block for rxode2/nlmixr models — model.function","text":"x model expression ... arguments append boolean determine lines appended piping.  possible values : TRUE lines appended model instead replaced FALSE lines replaced model (default) NA lines pre-pended model instead replaced lhs expression, append lines last observed line expression lhs auto boolean tells piping automatically selects parameters characterized population parameter, subject variability, covariate.  TRUE automatic selection occurs.  FALSE automatic selection turned everything added covariate (can promoted parameter ini statement). default TRUE, can changed options(rxode2.autoVarPiping=FALSE). cov character vector variables assumed covariates.  override automatic promotion population parameter estimate (eta) envir environment unevaluated model expressions evaluated.  May also NULL, list, data frame, pairlist integer specified sys.call.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Model block for rxode2/nlmixr models — model.function","text":"Model block ini information included.  ini must called model block","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/model.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Model block for rxode2/nlmixr models — model.function","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract model lines from a rxui model — modelExtract","title":"Extract model lines from a rxui model — modelExtract","text":"Extract model lines rxui model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract model lines from a rxui model — modelExtract","text":"","code":"modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for `function` modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for rxUi modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for rxode2 modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for rxModelVars modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )  # S3 method for default modelExtract(   x,   ...,   expression = FALSE,   endpoint = FALSE,   lines = FALSE,   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract model lines from a rxui model — modelExtract","text":"x model extract lines ... variables extract. missing, extract entire model (conditioned endpoint option ) expression return expressions (TRUE) strings (FALSE) endpoint include endpoint.  can : NA    -- Missing means include endpoint non-endpoint lines TRUE  -- include endpoint lines FALSE -- include non-endpoint lines lines boolean.  TRUE add lines attribute output value ie attr(, \"lines\") envir Environment evaluating variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract model lines from a rxui model — modelExtract","text":"expressions strings extracted lines. Note duplicated lhs expression line, return lines","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Extract model lines from a rxui model — modelExtract","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/modelExtract.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract model lines from a rxui model — modelExtract","text":"","code":"one.compartment <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v  <- exp(tv + eta.v)     d/dt(depot)  <- -ka * depot     d/dt(center) <-  ka * depot - cl / v * center     cp <- center / v     cp ~ add(add.sd)   })  }   f <- one.compartment()   modelExtract(f, cp) #> [1] \"cp <- center/v\"   modelExtract(one.compartment, d/dt(depot)) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #> [1] \"d/dt(depot) <- -ka * depot\"   # from variable  var <- \"d/dt(depot)\"   modelExtract(one.compartment, var) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #> [1] \"d/dt(depot) <- -ka * depot\"   modelExtract(f, endpoint=NA, lines=TRUE, expression=TRUE) #> [[1]] #> ka <- exp(tka + eta.ka) #>  #> [[2]] #> cl <- exp(tcl + eta.cl) #>  #> [[3]] #> v <- exp(tv + eta.v) #>  #> [[4]] #> d/dt(depot) <- -ka * depot #>  #> [[5]] #> d/dt(center) <- ka * depot - cl/v * center #>  #> [[6]] #> cp <- center/v #>  #> [[7]] #> cp ~ add(add.sd) #>  #> attr(,\"lines\") #> [1] 1 2 3 4 5 6 7"},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":null,"dir":"Reference","previous_headings":"","what":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"NULL given method, choices returned named vector.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"","code":"odeMethodToInt(method = c(\"liblsoda\", \"lsoda\", \"dop853\", \"indLin\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"method method solving ODEs.  Currently supports: \"liblsoda\" thread safe lsoda.  supports parallel thread-based solving, ignores user Jacobian specification. \"lsoda\" -- LSODA solver.  support parallel thread-based solving, allows user Jacobian specification. \"dop853\" -- DOP853 solver.  support parallel thread-based solving user Jacobian specification \"indLin\" -- Solving inductive linearization.  rxode2 dll must setup specially use solving routine.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/odeMethodToInt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Conversion between character and integer ODE integration methods for rxode2 — odeMethodToInt","text":"integer method (unless input NULL, case, see details)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot rxode2 objects — plot.rxSolve","title":"Plot rxode2 objects — plot.rxSolve","text":"Plot rxode2 objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot rxode2 objects — plot.rxSolve","text":"","code":"# S3 method for rxSolve plot(x, y, ..., log = \"\", xlab = \"Time\", ylab = \"\")  # S3 method for rxSolveConfint1 plot(x, y, ..., xlab = \"Time\", ylab = \"\", log = \"\")  # S3 method for rxSolveConfint2 plot(x, y, ..., xlab = \"Time\", ylab = \"\", log = \"\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot rxode2 objects — plot.rxSolve","text":"x rxode2 object plot y Compartments left-hand-side values plot either bare name character vector ... Ignored log \"\" (neither x y), \"x\", \"y\", \"xy\" (\"yx\") log-scale? xlab, ylab x y axis labels","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plot.rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot rxode2 objects — plot.rxSolve","text":"ggplot2 object","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Update Solved object with '+' — +.rxSolve","title":"Update Solved object with '+' — +.rxSolve","text":"Update Solved object '+'","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update Solved object with '+' — +.rxSolve","text":"","code":"# S3 method for rxSolve +(solved, new)"},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update Solved object with '+' — +.rxSolve","text":"solved Solved object new New information added table.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update Solved object with '+' — +.rxSolve","text":"new solved object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/plus-.rxSolve.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Update Solved object with '+' — +.rxSolve","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":null,"dir":"Reference","previous_headings":"","what":"Print the rxCoef object — print.rxCoef","title":"Print the rxCoef object — print.rxCoef","text":"prints user supplied arguments rxCoef object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print the rxCoef object — print.rxCoef","text":"","code":"# S3 method for rxCoef print(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print the rxCoef object — print.rxCoef","text":"x rxCoef object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print the rxCoef object — print.rxCoef","text":"original object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxCoef.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print the rxCoef object — print.rxCoef","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":null,"dir":"Reference","previous_headings":"","what":"Print rxDll object — print.rxDll","title":"Print rxDll object — print.rxDll","text":"tells rxDll loaded, ready /deleted.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print rxDll object — print.rxDll","text":"","code":"# S3 method for rxDll print(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print rxDll object — print.rxDll","text":"original object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxDll.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print rxDll object — print.rxDll","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Print information about the rxode2 object. — print.rxode2","title":"Print information about the rxode2 object. — print.rxode2","text":"prints model name status able solved","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print information about the rxode2 object. — print.rxode2","text":"","code":"# S3 method for rxode2 print(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print information about the rxode2 object. — print.rxode2","text":"x rxode2 object ... Ignored parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print information about the rxode2 object. — print.rxode2","text":"original object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/print.rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print information about the rxode2 object. — print.rxode2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":null,"dir":"Reference","previous_headings":"","what":"probit and inverse probit functions — probit","title":"probit and inverse probit functions — probit","text":"probit inverse probit functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"probit and inverse probit functions — probit","text":"","code":"probit(x, low = 0, high = 1)  probitInv(x, low = 0, high = 1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"probit and inverse probit functions — probit","text":"x Input value(s) range [low,high] translate -Inf Inf low Lowest value range high Highest value range","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"probit and inverse probit functions — probit","text":"values probit, probitInv probitNormInfo","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/probit.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"probit and inverse probit functions — probit","text":"","code":"probit(0.25) #> [1] -0.6744898  probitInv(-0.674) #> [1] 0.2501557  probitNormInfo(probit(0.25), sd = 0.1) #>       mean        var         cv  #> 0.25106491 0.00100658 0.12636827   probitNormInfo(probit(1, 0, 10), sd = 1, low = 0, high = 10) #>     mean      var       cv  #> 1.824166 4.307824 1.137796"},{"path":"https://nlmixr2.github.io/rxode2/reference/reexports.html","id":null,"dir":"Reference","previous_headings":"","what":"Objects exported from other packages — reexports","title":"Objects exported from other packages — reexports","text":"objects imported packages. Follow links see documentation. ggplot2 aes, expand_limits, facet_wrap, geom_line, ggplot, label_both, label_context, label_context, label_value, label_wrap_gen, scale_type, scale_x_continuous, scale_x_date, scale_x_discrete, scale_y_continuous, scale_y_date, scale_y_discrete, waiver, xlab, ylab lotri lotri magrittr %>% rxode2et .clearPipe, .collectWarnings, .s3register, add.dosing, add.sampling, .et, .rxEvid, .rxRateDur, et, etExpand, etRbind, etRep, etSeq, eventTable, rxCbindStudyIndividual, rxEtDispatchSolve, rxEvid, rxRateDur, rxReq, rxStack, toTrialDuration rxode2parse .getLastIdLvl, forderForceBase, rxDerived, rxSetIni0 rxode2random .cbindOme, .expandPars, .vecDf, cvPost, invWR1d, phi, rinvchisq, rLKJ1, rxGetSeed, rxGetSeed, rxRmvn, rxSeedEng, rxSetSeed, rxSetSeed, rxSetSeed, rxWithPreserveSeed, rxWithSeed, rxWithSeed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/reexports.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Objects exported from other packages — reexports","text":"Inherited parent routine","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":null,"dir":"Reference","previous_headings":"","what":"Allow unloading of dlls — rxAllowUnload","title":"Allow unloading of dlls — rxAllowUnload","text":"Allow unloading dlls","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Allow unloading of dlls — rxAllowUnload","text":"","code":"rxAllowUnload(allow)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Allow unloading of dlls — rxAllowUnload","text":"allow boolean indicating garbage collection unload rxode2 dlls.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Allow unloading of dlls — rxAllowUnload","text":"Boolean allow; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Allow unloading of dlls — rxAllowUnload","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAllowUnload.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Allow unloading of dlls — rxAllowUnload","text":"","code":"# Garbage collection will not unload un-used rxode2 dlls rxAllowUnload(FALSE); #> [1] FALSE  # Garbage collection will unload unused rxode2 dlls rxAllowUnload(TRUE); #> [1] TRUE"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Append two rxui models together — rxAppendModel","title":"Append two rxui models together — rxAppendModel","text":"Append two rxui models together","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Append two rxui models together — rxAppendModel","text":"","code":"rxAppendModel(..., common = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Append two rxui models together — rxAppendModel","text":"... models append together common boolean determines need common value bind","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Append two rxui models together — rxAppendModel","text":"New model models appended together","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Append two rxui models together — rxAppendModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAppendModel.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Append two rxui models together — rxAppendModel","text":"","code":"# \\donttest{  ocmt <- function() {   ini({     tka <- exp(0.45) # Ka     tcl <- exp(1) # Cl     tv <- exp(3.45); # log V     ## the label(\"Label name\") works with all models     add.sd <- 0.7   })   model({     ka <- tka     cl <- tcl     v <- tv     d/dt(depot) <- -ka * depot     d/dt(center) <- ka * depot - cl / v * center     cp <- center / v     cp ~ add(add.sd)   }) }  idr <- function() {   ini({     tkin <- log(1)     tkout <- log(1)     tic50 <- log(10)     gamma <- fix(1)     idr.sd <- 1   })   model({     kin <- exp(tkin)     kout <- exp(tkout)     ic50 <- exp(tic50)     d/dt(eff) <- kin - kout*(1-ceff^gamma/(ic50^gamma+ceff^gamma))     eff ~ add(idr.sd)   }) }  rxAppendModel(ocmt %>% model(ceff=cp,append=TRUE), idr) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #>   #>   #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd      tkin     tkout     tic50     gamma  #>  1.568312  2.718282 31.500392  0.700000  0.000000  0.000000  2.302585  1.000000  #>    idr.sd  #>  1.000000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #> 3                  3              eff #>  ── Multiple Endpoint Model ($multipleEndpoint): ──   #>   variable                cmt                dvid* #> 1   cp ~ …  cmt='cp' or cmt=4  dvid='cp' or dvid=1 #> 2  eff ~ … cmt='eff' or cmt=3 dvid='eff' or dvid=2 #>   * If dvids are outside this range, all dvids are re-numered sequentially, ie 1,7, 10 becomes 1,2,3 etc #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>         tkin <- 0 #>         tkout <- 0 #>         tic50 <- 2.30258509299405 #>         gamma <- fix(1) #>         idr.sd <- c(0, 1) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.sd) #>         ceff <- cp #>         kin <- exp(tkin) #>         kout <- exp(tkout) #>         ic50 <- exp(tic50) #>         d/dt(eff) <- kin - kout * (1 - ceff^gamma/(ic50^gamma +  #>             ceff^gamma)) #>         eff ~ add(idr.sd) #>     }) #> }  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign Control Variable — rxAssignControlValue","title":"Assign Control Variable — rxAssignControlValue","text":"Assign Control Variable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign Control Variable — rxAssignControlValue","text":"","code":"rxAssignControlValue(ui, option, value)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign Control Variable — rxAssignControlValue","text":"ui rxode2 ui function option Option name control modify value Value control modify","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign Control Variable — rxAssignControlValue","text":"Nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignControlValue.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign Control Variable — rxAssignControlValue","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign pointer based on model variables — rxAssignPtr","title":"Assign pointer based on model variables — rxAssignPtr","text":"Assign pointer based model variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign pointer based on model variables — rxAssignPtr","text":"","code":"rxAssignPtr(object = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign pointer based on model variables — rxAssignPtr","text":"object rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxAssignPtr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign pointer based on model variables — rxAssignPtr","text":"nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates a logical matrix for block matrixes. — rxBlockZeros","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"Creates logical matrix block matrixes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"","code":"rxBlockZeros(mat, i)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"mat Matrix Row/column block matrix setup.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxBlockZeros.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates a logical matrix for block matrixes. — rxBlockZeros","text":"logical matrix returning elements zero.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the C file associated with the rxode2 object — rxC","title":"Return the C file associated with the rxode2 object — rxC","text":"return C code generating rxode2 DLL.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the C file associated with the rxode2 object — rxC","text":"","code":"rxC(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the C file associated with the rxode2 object — rxC","text":"obj rxode2 family objects character string model specification location file model specification.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the C file associated with the rxode2 object — rxC","text":"path library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxC.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the C file associated with the rxode2 object — rxC","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":null,"dir":"Reference","previous_headings":"","what":"Use cat when rxode2.verbose is TRUE — rxCat","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"Use cat rxode2.verbose TRUE","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"","code":"rxCat(a, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"... Parameters sent cat","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCat.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Use cat when rxode2.verbose is TRUE — rxCat","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":null,"dir":"Reference","previous_headings":"","what":"rxChain Chain or add item to solved system of equations — rxChain","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"Add item solved system equations","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"","code":"rxChain(obj1, obj2)  # S3 method for solveRxDll +(obj1, obj2)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"obj1 Solved object. obj2 New object added/piped/chained solved object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"newObject event table, return new solved object new event table.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxChain Chain or add item to solved system of equations — rxChain","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":null,"dir":"Reference","previous_headings":"","what":"Second command in chaining commands — rxChain2","title":"Second command in chaining commands — rxChain2","text":"s3 method called internally + \\%>\\% operators.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Second command in chaining commands — rxChain2","text":"","code":"rxChain2(obj, solvedObject)  # S3 method for default rxChain2(obj, solvedObject)  # S3 method for EventTable rxChain2(obj, solvedObject)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Second command in chaining commands — rxChain2","text":"obj object added/chained/piped solved object solvedObject solved object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Second command in chaining commands — rxChain2","text":"chained operation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxChain2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Second command in chaining commands — rxChain2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":null,"dir":"Reference","previous_headings":"","what":"Cleanup anonymous DLLs by unloading them — rxClean","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"cleans rxode2 loaded DLLs","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"","code":"rxClean(wd)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"wd directory cleaned; (DEPRECIATED), longer anything. unloads rxode2 anonymous dlls.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"TRUE successful","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxClean.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Cleanup anonymous DLLs by unloading them — rxClean","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":null,"dir":"Reference","previous_headings":"","what":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"Combine Error Lines create rxode2 expression","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"","code":"rxCombineErrorLines(   uiModel,   errLines = NULL,   prefixLines = NULL,   paramsLine = NULL,   modelVars = FALSE,   cmtLines = TRUE,   dvidLine = TRUE,   lstExpr = NULL,   useIf = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"uiModel UI model errLines Error lines; missing, get error lines rxGetDistributionSimulationLines() prefixLines Prefix lines, param statement paramsLine Params line, present. modelVars Return model vars instead rxode2 statement cmtLines Include trailing cmt lines dvidLine Include trailing dvid() specification lstExpr list expressions model, NULL.  NULL defaults model expressions accessible uiModel$lstExpr. useIf Use (CMT == X) endpoints","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"quoted expression can evaluated compiled rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"exported allow functions mangle error lines make types estimation methods (needed)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCombineErrorLines.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Combine Error Lines and create rxode2 expression — rxCombineErrorLines","text":"","code":"# \\donttest{  one.cmt <- function() {   ini({     ## You may label each parameter with a comment     tka <- 0.45 # Log Ka     tcl <- log(c(0, 2.7, 100)) # Log Cl     ## This works with interactive models     ## You may also label the preceding line with label(\"label text\")     tv <- 3.45; label(\"log V\")     ## the label(\"Label name\") works with all models     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     linCmt() ~ add(add.sd)   }) }  f <- rxode2(one.cmt) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()'  # You can get the simulation model easily by rxCombineErrorLines(f) #> rxode2({ #>     params(tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v) #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     rx_yj_ ~ 2 #>     rx_lambda_ ~ 1 #>     rx_low_ ~ 0 #>     rx_hi_ ~ 1 #>     rx_pred_f_ ~ linCmt() #>     rx_pred_ ~ rx_pred_f_ #>     rx_r_ ~ (add.sd)^2 #>     ipredSim <- rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>         rx_hi_) #>     sim <- rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.rxLinCmt, rx_lambda_,  #>         rx_yj_, rx_low_, rx_hi_) #>     cmt(rxLinCmt) #>     dvid(1) #> })  # You can then get the compiled model by simply evaluting the model: r <- eval(rxCombineErrorLines(f)) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # This also works with multile endpoint models: pk.turnover.emax <- function() {   ini({     tktr <- log(1)     tka <- log(1)     tcl <- log(0.1)     tv <- log(10)     ##     eta.ktr ~ 1     eta.ka ~ 1     eta.cl ~ 2     eta.v ~ 1     prop.err <- 0.1     pkadd.err <- 0.1     ##     temax <- logit(0.8)     tec50 <- log(0.5)     tkout <- log(0.05)     te0 <- log(100)     ##     eta.emax ~ .5     eta.ec50  ~ .5     eta.kout ~ .5     eta.e0 ~ .5     ##     pdadd.err <- 10   })   model({     ktr <- exp(tktr + eta.ktr)     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     ##     emax=expit(temax+eta.emax)     ec50 =  exp(tec50 + eta.ec50)     kout = exp(tkout + eta.kout)     e0 = exp(te0 + eta.e0)     ##     DCP = center/v     PD=1-emax*DCP/(ec50+DCP)     ##     effect(0) = e0     kin = e0*kout     ##     d/dt(depot) = -ktr * depot     d/dt(gut) =  ktr * depot -ka * gut     d/dt(center) =  ka * gut - cl / v * center     d/dt(effect) = kin*PD -kout*effect     ##     cp = center / v     cp ~ prop(prop.err) + add(pkadd.err)     effect ~ add(pdadd.err)   }) }  f <- rxode2(pk.turnover.emax) #>   #>   rxCombineErrorLines(f) #> rxode2({ #>     params(tktr, tka, tcl, tv, prop.err, pkadd.err, temax, tec50,  #>         tkout, te0, pdadd.err, eta.ktr, eta.ka, eta.cl, eta.v,  #>         eta.emax, eta.ec50, eta.kout, eta.e0) #>     ktr <- exp(tktr + eta.ktr) #>     ka <- exp(tka + eta.ka) #>     cl <- exp(tcl + eta.cl) #>     v <- exp(tv + eta.v) #>     emax = expit(temax + eta.emax) #>     ec50 = exp(tec50 + eta.ec50) #>     kout = exp(tkout + eta.kout) #>     e0 = exp(te0 + eta.e0) #>     DCP = center/v #>     PD = 1 - emax * DCP/(ec50 + DCP) #>     effect(0) = e0 #>     kin = e0 * kout #>     d/dt(depot) = -ktr * depot #>     d/dt(gut) = ktr * depot - ka * gut #>     d/dt(center) = ka * gut - cl/v * center #>     d/dt(effect) = kin * PD - kout * effect #>     cp = center/v #>     if (CMT == 5) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ cp #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (pkadd.err)^2 + (rx_pred_f_)^2 * (prop.err)^2 #>         ipredSim <- rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim <- rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.cp, rx_lambda_,  #>             rx_yj_, rx_low_, rx_hi_) #>     } #>     if (CMT == 4) { #>         rx_yj_ ~ 2 #>         rx_lambda_ ~ 1 #>         rx_low_ ~ 0 #>         rx_hi_ ~ 1 #>         rx_pred_f_ ~ effect #>         rx_pred_ ~ rx_pred_f_ #>         rx_r_ ~ (pdadd.err)^2 #>         ipredSim <- rxTBSi(rx_pred_, rx_lambda_, rx_yj_, rx_low_,  #>             rx_hi_) #>         sim <- rxTBSi(rx_pred_ + sqrt(rx_r_) * rxerr.effect,  #>             rx_lambda_, rx_yj_, rx_low_, rx_hi_) #>     } #>     cmt(cp) #>     dvid(5, 4) #> })  # Note that in the parsed form, you can also get the compiled rxode2 # model with $simulationModel  f$simulationModel #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> rxode2 2.1.3 model named rx_8d950c43531f6895b2b69a6751925bb9 model (✔ ready).  #> $state: depot, gut, center, effect #> $stateExtra: cp #> $params: tktr, tka, tcl, tv, prop.err, pkadd.err, temax, tec50, tkout, te0, pdadd.err, eta.ktr, eta.ka, eta.cl, eta.v, eta.emax, eta.ec50, eta.kout, eta.e0, CMT, rxerr.cp, rxerr.effect #> $lhs: ktr, ka, cl, v, emax, ec50, kout, e0, DCP, PD, kin, cp, ipredSim, sim  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":null,"dir":"Reference","previous_headings":"","what":"Compile a model if needed — rxCompile","title":"Compile a model if needed — rxCompile","text":"compilation workhorse creating rxode2 model DLL files.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compile a model if needed — rxCompile","text":"","code":"rxCompile(   model,   dir,   prefix,   force = FALSE,   modName = NULL,   package = NULL,   ... )  # S3 method for rxModelVars rxCompile(   model,   dir = NULL,   prefix = NULL,   force = FALSE,   modName = NULL,   package = NULL,   ... )  # S3 method for character rxCompile(   model,   dir = NULL,   prefix = NULL,   force = FALSE,   modName = NULL,   package = NULL,   ... )  # S3 method for rxDll rxCompile(model, ...)  # S3 method for rxode2 rxCompile(model, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compile a model if needed — rxCompile","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . dir model directory C file stored compiling.   prefix string indicating prefix use C based functions.  missing, calculated based file name, md5 parsed model. force boolean stating (re)compile forced rxode2 detects models already generated. modName string used model name. string used naming various aspects computations, including generating C symbol names, dynamic libraries, etc. Therefore, necessary modName consists simple ASCII alphanumeric characters starting letter. package Package name pre-compiled binaries. ... arguments sent rxTrans() function.","code":"If unspecified, the C code is stored in a temporary directory, then the model is compiled and moved to the current directory. Afterwards the C code is removed.  If specified, the C code is stored in the specified directory and then compiled in that directory.  The C code is not removed after the DLL is created in the same directory.  This can be useful to debug the c-code outputs."},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compile a model if needed — rxCompile","text":"rxDll object following components dll DLL path model model specification .c function call C code correct context DLL using .C() function. .call function call C code correct context DLL using .Call() function. args list arguments used create rxDll object.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCompile.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Compile a model if needed — rxCompile","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":null,"dir":"Reference","previous_headings":"","what":"Current Condition for rxode2 object — rxCondition","title":"Current Condition for rxode2 object — rxCondition","text":"Current Condition rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Current Condition for rxode2 object — rxCondition","text":"","code":"rxCondition(obj, condition = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Current Condition for rxode2 object — rxCondition","text":"obj rxode2 object condition specified one conditions rxode2 object (determined rxExpandIfElse()), assign rxode2 current condition parameter.  condition one known condition, condition set NULL, implying conditioning currently used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Current Condition for rxode2 object — rxCondition","text":"Current condition rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCondition.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Current Condition for rxode2 object — rxCondition","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":null,"dir":"Reference","previous_headings":"","what":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"assumes normal ODE equations first equations ODE expanded forward sensitivities type sensitivity (like adjoint)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"","code":"rxControlUpdateSens(rxControl, sensCmt = NULL, ncmt = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"rxControl Input list rxControl type list sensCmt Number sensitivity compartments ncmt Number compartments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"Updated rxControl $atol, $rtol, $ssAtol $ssRtol updated different sensitivities normal ODEs (first) different sensitivity larger compartments (sensitivities).","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxControlUpdateSens.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"This updates the tolerances based on the sensitivity equations — rxControlUpdateSens","text":"","code":"tmp <- rxControl()  tmp2 <- rxControlUpdateSens(tmp, 3, 6)  tmp2$atol #> [1] 1e-08 1e-08 1e-08 1e-08 1e-08 1e-08 tmp2$rtol #> [1] 1e-06 1e-06 1e-06 1e-06 1e-06 1e-06 tmp2$ssAtol #> [1] 1e-08 1e-08 1e-08 1e-08 1e-08 1e-08 tmp2$ssRtol #> [1] 1e-06 1e-06 1e-06 1e-06 1e-06 1e-06"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":null,"dir":"Reference","previous_headings":"","what":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"run, R_user_dir rxode2's cache present, create cache","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"","code":"rxCreateCache()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxCreateCache.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This will create the cache directory for rxode2 to save between sessions — rxCreateCache","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":null,"dir":"Reference","previous_headings":"","what":"Add to rxode2's derivative tables — rxD","title":"Add to rxode2's derivative tables — rxD","text":"Add rxode2's derivative tables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add to rxode2's derivative tables — rxD","text":"","code":"rxD(name, derivatives)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add to rxode2's derivative tables — rxD","text":"name Function Name derivatives list functions. function takes number arguments original function.  first function construct derivative respect first argument; second function construct derivitive respect second argument, .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add to rxode2's derivative tables — rxD","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Add to rxode2's derivative tables — rxD","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxD.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add to rxode2's derivative tables — rxD","text":"","code":"## Add an arbitrary list of derivative functions ## In this case the fun(x,y) is assumed to be 0.5*x^2+0.5*y^2  rxD(\"fun\", list(   function(x, y) {     return(x)   },   function(x, y) {     return(y)   } ))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":null,"dir":"Reference","previous_headings":"","what":"Delete the DLL for the model — rxDelete","title":"Delete the DLL for the model — rxDelete","text":"function deletes DLL, delete model information object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Delete the DLL for the model — rxDelete","text":"","code":"rxDelete(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Delete the DLL for the model — rxDelete","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Delete the DLL for the model — rxDelete","text":"boolean stating operation successful.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDelete.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Delete the DLL for the model — rxDelete","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":null,"dir":"Reference","previous_headings":"","what":"Demote the error type — rxDemoteAddErr","title":"Demote the error type — rxDemoteAddErr","text":"Demote error type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Demote the error type — rxDemoteAddErr","text":"","code":"rxDemoteAddErr(errType)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Demote the error type — rxDemoteAddErr","text":"errType Error type factor","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Demote the error type — rxDemoteAddErr","text":"Demoted Error Type","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Demote the error type — rxDemoteAddErr","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDemoteAddErr.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Demote the error type — rxDemoteAddErr","text":"","code":"rxErrTypeCombine(\"add\") %>%   rxErrTypeCombine(\"prop\") #> $transform #> [1] untransformed #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] add + prop #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] untransformed #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\"  # This removes the internal additive error rxErrTypeCombine(\"add\") %>%   rxErrTypeCombine(\"prop\") %>%   rxDemoteAddErr() #> $transform #> [1] untransformed #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] prop #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] untransformed #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\"  # This is used for logitNorm(NA), the additive portion is stripped"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":null,"dir":"Reference","previous_headings":"","what":"Jacobian and parameter derivatives — rxDfdy","title":"Jacobian and parameter derivatives — rxDfdy","text":"Return Jacobain parameter derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Jacobian and parameter derivatives — rxDfdy","text":"","code":"rxDfdy(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Jacobian and parameter derivatives — rxDfdy","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Jacobian and parameter derivatives — rxDfdy","text":"list jacobian parameters defined rxode2 object.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDfdy.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Jacobian and parameter derivatives — rxDfdy","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the DLL associated with the rxode2 object — rxDll","title":"Return the DLL associated with the rxode2 object — rxDll","text":"return dynamic load library shared object used run C code rxode2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the DLL associated with the rxode2 object — rxDll","text":"","code":"rxDll(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the DLL associated with the rxode2 object — rxDll","text":"obj rxode2 family objects character string model specification location file model specification.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the DLL associated with the rxode2 object — rxDll","text":"path library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDll.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the DLL associated with the rxode2 object — rxDll","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":null,"dir":"Reference","previous_headings":"","what":"Load rxode2 object — rxDynLoad","title":"Load rxode2 object — rxDynLoad","text":"Load rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load rxode2 object — rxDynLoad","text":"","code":"rxDynLoad(obj)  rxLoad(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load rxode2 object — rxDynLoad","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load rxode2 object — rxDynLoad","text":"Boolean returning rxode2 library loaded.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynLoad.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Load rxode2 object — rxDynLoad","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":null,"dir":"Reference","previous_headings":"","what":"Unload rxode2 object — rxDynUnload","title":"Unload rxode2 object — rxDynUnload","text":"Unload rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Unload rxode2 object — rxDynUnload","text":"","code":"rxDynUnload(obj)  rxUnload(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Unload rxode2 object — rxDynUnload","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Unload rxode2 object — rxDynUnload","text":"Boolean returning rxode2 library loaded.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxDynUnload.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Unload rxode2 object — rxDynUnload","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":null,"dir":"Reference","previous_headings":"","what":"Combine transformations and error structures — rxErrTypeCombine","title":"Combine transformations and error structures — rxErrTypeCombine","text":"Combine error information figure transformation applied current endpoint","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Combine transformations and error structures — rxErrTypeCombine","text":"","code":"rxErrTypeCombine(oldErrType, newErrType)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Combine transformations and error structures — rxErrTypeCombine","text":"oldErrType old transformation, default zero representing prior transformation. parameter first allow piping. parameter addTransform missing oldErrType character value, functions swaps oldErrType addTransform assigns oldErrType zero assuming prior distribution. newErrType new distribution \"added\" current transformation.  assumes inputs preferred distribution name, determined rxPreferredDistributionName()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Combine transformations and error structures — rxErrTypeCombine","text":"new transformation factor","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Combine transformations and error structures — rxErrTypeCombine","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxErrTypeCombine.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Combine transformations and error structures — rxErrTypeCombine","text":"","code":"rxErrTypeCombine(\"probitNorm\") #> $transform #> [1] probit #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] add #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] none #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\"  rxErrTypeCombine(\"probitNorm\") %>%   rxErrTypeCombine(\"boxCox\") #> $transform #> [1] probit + boxCox #> 10 Levels: boxCox yeoJohnson untransformed lnorm logit ... probit + boxCox #>  #> $errType #> [1] add #> Levels: add prop pow add + prop add + pow none #>  #> $errTypeF #> [1] none #> Levels: untransformed transformed f none #>  #> $addProp #> [1] default #> Levels: combined1 combined2 default #>  #> attr(,\"class\") #> [1] \"rxCombinedErrorList\""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand d(f)/d(eta) — rxExpandFEta_","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"Expand d(f)/d(eta)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"","code":"rxExpandFEta_(state, neta, pred, isTheta = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"state state expand neta number etas pred type prediction isTheta logical, expansion actually thetas instead etas","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandFEta_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand d(f)/d(eta) — rxExpandFEta_","text":"String symengine expressions evaluate calculate df/deta","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":null,"dir":"Reference","previous_headings":"","what":"Faster expand.grid — rxExpandGrid","title":"Faster expand.grid — rxExpandGrid","text":"support x y characters right now","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Faster expand.grid — rxExpandGrid","text":"","code":"rxExpandGrid(x, y, type = 0L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Faster expand.grid — rxExpandGrid","text":"x first element (must character) y second element (must character) type Internal type=0L traditional expand grid type=1L Jacobian expand grid (adds symbols)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Faster expand.grid — rxExpandGrid","text":"Expand grid","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Faster expand.grid — rxExpandGrid","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Faster expand.grid — rxExpandGrid","text":"","code":"## rxExpandGrid(letters, letters) #>     Var1 Var2 #> 1      a    a #> 2      b    a #> 3      c    a #> 4      d    a #> 5      e    a #> 6      f    a #> 7      g    a #> 8      h    a #> 9      i    a #> 10     j    a #> 11     k    a #> 12     l    a #> 13     m    a #> 14     n    a #> 15     o    a #> 16     p    a #> 17     q    a #> 18     r    a #> 19     s    a #> 20     t    a #> 21     u    a #> 22     v    a #> 23     w    a #> 24     x    a #> 25     y    a #> 26     z    a #> 27     a    b #> 28     b    b #> 29     c    b #> 30     d    b #> 31     e    b #> 32     f    b #> 33     g    b #> 34     h    b #> 35     i    b #> 36     j    b #> 37     k    b #> 38     l    b #> 39     m    b #> 40     n    b #> 41     o    b #> 42     p    b #> 43     q    b #> 44     r    b #> 45     s    b #> 46     t    b #> 47     u    b #> 48     v    b #> 49     w    b #> 50     x    b #> 51     y    b #> 52     z    b #> 53     a    c #> 54     b    c #> 55     c    c #> 56     d    c #> 57     e    c #> 58     f    c #> 59     g    c #> 60     h    c #> 61     i    c #> 62     j    c #> 63     k    c #> 64     l    c #> 65     m    c #> 66     n    c #> 67     o    c #> 68     p    c #> 69     q    c #> 70     r    c #> 71     s    c #> 72     t    c #> 73     u    c #> 74     v    c #> 75     w    c #> 76     x    c #> 77     y    c #> 78     z    c #> 79     a    d #> 80     b    d #> 81     c    d #> 82     d    d #> 83     e    d #> 84     f    d #> 85     g    d #> 86     h    d #> 87     i    d #> 88     j    d #> 89     k    d #> 90     l    d #> 91     m    d #> 92     n    d #> 93     o    d #> 94     p    d #> 95     q    d #> 96     r    d #> 97     s    d #> 98     t    d #> 99     u    d #> 100    v    d #> 101    w    d #> 102    x    d #> 103    y    d #> 104    z    d #> 105    a    e #> 106    b    e #> 107    c    e #> 108    d    e #> 109    e    e #> 110    f    e #> 111    g    e #> 112    h    e #> 113    i    e #> 114    j    e #> 115    k    e #> 116    l    e #> 117    m    e #> 118    n    e #> 119    o    e #> 120    p    e #> 121    q    e #> 122    r    e #> 123    s    e #> 124    t    e #> 125    u    e #> 126    v    e #> 127    w    e #> 128    x    e #> 129    y    e #> 130    z    e #> 131    a    f #> 132    b    f #> 133    c    f #> 134    d    f #> 135    e    f #> 136    f    f #> 137    g    f #> 138    h    f #> 139    i    f #> 140    j    f #> 141    k    f #> 142    l    f #> 143    m    f #> 144    n    f #> 145    o    f #> 146    p    f #> 147    q    f #> 148    r    f #> 149    s    f #> 150    t    f #> 151    u    f #> 152    v    f #> 153    w    f #> 154    x    f #> 155    y    f #> 156    z    f #> 157    a    g #> 158    b    g #> 159    c    g #> 160    d    g #> 161    e    g #> 162    f    g #> 163    g    g #> 164    h    g #> 165    i    g #> 166    j    g #> 167    k    g #> 168    l    g #> 169    m    g #> 170    n    g #> 171    o    g #> 172    p    g #> 173    q    g #> 174    r    g #> 175    s    g #> 176    t    g #> 177    u    g #> 178    v    g #> 179    w    g #> 180    x    g #> 181    y    g #> 182    z    g #> 183    a    h #> 184    b    h #> 185    c    h #> 186    d    h #> 187    e    h #> 188    f    h #> 189    g    h #> 190    h    h #> 191    i    h #> 192    j    h #> 193    k    h #> 194    l    h #> 195    m    h #> 196    n    h #> 197    o    h #> 198    p    h #> 199    q    h #> 200    r    h #> 201    s    h #> 202    t    h #> 203    u    h #> 204    v    h #> 205    w    h #> 206    x    h #> 207    y    h #> 208    z    h #> 209    a    i #> 210    b    i #> 211    c    i #> 212    d    i #> 213    e    i #> 214    f    i #> 215    g    i #> 216    h    i #> 217    i    i #> 218    j    i #> 219    k    i #> 220    l    i #> 221    m    i #> 222    n    i #> 223    o    i #> 224    p    i #> 225    q    i #> 226    r    i #> 227    s    i #> 228    t    i #> 229    u    i #> 230    v    i #> 231    w    i #> 232    x    i #> 233    y    i #> 234    z    i #> 235    a    j #> 236    b    j #> 237    c    j #> 238    d    j #> 239    e    j #> 240    f    j #> 241    g    j #> 242    h    j #> 243    i    j #> 244    j    j #> 245    k    j #> 246    l    j #> 247    m    j #> 248    n    j #> 249    o    j #> 250    p    j #> 251    q    j #> 252    r    j #> 253    s    j #> 254    t    j #> 255    u    j #> 256    v    j #> 257    w    j #> 258    x    j #> 259    y    j #> 260    z    j #> 261    a    k #> 262    b    k #> 263    c    k #> 264    d    k #> 265    e    k #> 266    f    k #> 267    g    k #> 268    h    k #> 269    i    k #> 270    j    k #> 271    k    k #> 272    l    k #> 273    m    k #> 274    n    k #> 275    o    k #> 276    p    k #> 277    q    k #> 278    r    k #> 279    s    k #> 280    t    k #> 281    u    k #> 282    v    k #> 283    w    k #> 284    x    k #> 285    y    k #> 286    z    k #> 287    a    l #> 288    b    l #> 289    c    l #> 290    d    l #> 291    e    l #> 292    f    l #> 293    g    l #> 294    h    l #> 295    i    l #> 296    j    l #> 297    k    l #> 298    l    l #> 299    m    l #> 300    n    l #> 301    o    l #> 302    p    l #> 303    q    l #> 304    r    l #> 305    s    l #> 306    t    l #> 307    u    l #> 308    v    l #> 309    w    l #> 310    x    l #> 311    y    l #> 312    z    l #> 313    a    m #> 314    b    m #> 315    c    m #> 316    d    m #> 317    e    m #> 318    f    m #> 319    g    m #> 320    h    m #> 321    i    m #> 322    j    m #> 323    k    m #> 324    l    m #> 325    m    m #> 326    n    m #> 327    o    m #> 328    p    m #> 329    q    m #> 330    r    m #> 331    s    m #> 332    t    m #> 333    u    m #> 334    v    m #> 335    w    m #> 336    x    m #> 337    y    m #> 338    z    m #> 339    a    n #> 340    b    n #> 341    c    n #> 342    d    n #> 343    e    n #> 344    f    n #> 345    g    n #> 346    h    n #> 347    i    n #> 348    j    n #> 349    k    n #> 350    l    n #> 351    m    n #> 352    n    n #> 353    o    n #> 354    p    n #> 355    q    n #> 356    r    n #> 357    s    n #> 358    t    n #> 359    u    n #> 360    v    n #> 361    w    n #> 362    x    n #> 363    y    n #> 364    z    n #> 365    a    o #> 366    b    o #> 367    c    o #> 368    d    o #> 369    e    o #> 370    f    o #> 371    g    o #> 372    h    o #> 373    i    o #> 374    j    o #> 375    k    o #> 376    l    o #> 377    m    o #> 378    n    o #> 379    o    o #> 380    p    o #> 381    q    o #> 382    r    o #> 383    s    o #> 384    t    o #> 385    u    o #> 386    v    o #> 387    w    o #> 388    x    o #> 389    y    o #> 390    z    o #> 391    a    p #> 392    b    p #> 393    c    p #> 394    d    p #> 395    e    p #> 396    f    p #> 397    g    p #> 398    h    p #> 399    i    p #> 400    j    p #> 401    k    p #> 402    l    p #> 403    m    p #> 404    n    p #> 405    o    p #> 406    p    p #> 407    q    p #> 408    r    p #> 409    s    p #> 410    t    p #> 411    u    p #> 412    v    p #> 413    w    p #> 414    x    p #> 415    y    p #> 416    z    p #> 417    a    q #> 418    b    q #> 419    c    q #> 420    d    q #> 421    e    q #> 422    f    q #> 423    g    q #> 424    h    q #> 425    i    q #> 426    j    q #> 427    k    q #> 428    l    q #> 429    m    q #> 430    n    q #> 431    o    q #> 432    p    q #> 433    q    q #> 434    r    q #> 435    s    q #> 436    t    q #> 437    u    q #> 438    v    q #> 439    w    q #> 440    x    q #> 441    y    q #> 442    z    q #> 443    a    r #> 444    b    r #> 445    c    r #> 446    d    r #> 447    e    r #> 448    f    r #> 449    g    r #> 450    h    r #> 451    i    r #> 452    j    r #> 453    k    r #> 454    l    r #> 455    m    r #> 456    n    r #> 457    o    r #> 458    p    r #> 459    q    r #> 460    r    r #> 461    s    r #> 462    t    r #> 463    u    r #> 464    v    r #> 465    w    r #> 466    x    r #> 467    y    r #> 468    z    r #> 469    a    s #> 470    b    s #> 471    c    s #> 472    d    s #> 473    e    s #> 474    f    s #> 475    g    s #> 476    h    s #> 477    i    s #> 478    j    s #> 479    k    s #> 480    l    s #> 481    m    s #> 482    n    s #> 483    o    s #> 484    p    s #> 485    q    s #> 486    r    s #> 487    s    s #> 488    t    s #> 489    u    s #> 490    v    s #> 491    w    s #> 492    x    s #> 493    y    s #> 494    z    s #> 495    a    t #> 496    b    t #> 497    c    t #> 498    d    t #> 499    e    t #> 500    f    t #> 501    g    t #> 502    h    t #> 503    i    t #> 504    j    t #> 505    k    t #> 506    l    t #> 507    m    t #> 508    n    t #> 509    o    t #> 510    p    t #> 511    q    t #> 512    r    t #> 513    s    t #> 514    t    t #> 515    u    t #> 516    v    t #> 517    w    t #> 518    x    t #> 519    y    t #> 520    z    t #> 521    a    u #> 522    b    u #> 523    c    u #> 524    d    u #> 525    e    u #> 526    f    u #> 527    g    u #> 528    h    u #> 529    i    u #> 530    j    u #> 531    k    u #> 532    l    u #> 533    m    u #> 534    n    u #> 535    o    u #> 536    p    u #> 537    q    u #> 538    r    u #> 539    s    u #> 540    t    u #> 541    u    u #> 542    v    u #> 543    w    u #> 544    x    u #> 545    y    u #> 546    z    u #> 547    a    v #> 548    b    v #> 549    c    v #> 550    d    v #> 551    e    v #> 552    f    v #> 553    g    v #> 554    h    v #> 555    i    v #> 556    j    v #> 557    k    v #> 558    l    v #> 559    m    v #> 560    n    v #> 561    o    v #> 562    p    v #> 563    q    v #> 564    r    v #> 565    s    v #> 566    t    v #> 567    u    v #> 568    v    v #> 569    w    v #> 570    x    v #> 571    y    v #> 572    z    v #> 573    a    w #> 574    b    w #> 575    c    w #> 576    d    w #> 577    e    w #> 578    f    w #> 579    g    w #> 580    h    w #> 581    i    w #> 582    j    w #> 583    k    w #> 584    l    w #> 585    m    w #> 586    n    w #> 587    o    w #> 588    p    w #> 589    q    w #> 590    r    w #> 591    s    w #> 592    t    w #> 593    u    w #> 594    v    w #> 595    w    w #> 596    x    w #> 597    y    w #> 598    z    w #> 599    a    x #> 600    b    x #> 601    c    x #> 602    d    x #> 603    e    x #> 604    f    x #> 605    g    x #> 606    h    x #> 607    i    x #> 608    j    x #> 609    k    x #> 610    l    x #> 611    m    x #> 612    n    x #> 613    o    x #> 614    p    x #> 615    q    x #> 616    r    x #> 617    s    x #> 618    t    x #> 619    u    x #> 620    v    x #> 621    w    x #> 622    x    x #> 623    y    x #> 624    z    x #> 625    a    y #> 626    b    y #> 627    c    y #> 628    d    y #> 629    e    y #> 630    f    y #> 631    g    y #> 632    h    y #> 633    i    y #> 634    j    y #> 635    k    y #> 636    l    y #> 637    m    y #> 638    n    y #> 639    o    y #> 640    p    y #> 641    q    y #> 642    r    y #> 643    s    y #> 644    t    y #> 645    u    y #> 646    v    y #> 647    w    y #> 648    x    y #> 649    y    y #> 650    z    y #> 651    a    z #> 652    b    z #> 653    c    z #> 654    d    z #> 655    e    z #> 656    f    z #> 657    g    z #> 658    h    z #> 659    i    z #> 660    j    z #> 661    k    z #> 662    l    z #> 663    m    z #> 664    n    z #> 665    o    z #> 666    p    z #> 667    q    z #> 668    r    z #> 669    s    z #> 670    t    z #> 671    u    z #> 672    v    z #> 673    w    z #> 674    x    z #> 675    y    z #> 676    z    z  ## Another fast method; See ## https://stackoverflow.com/questions/10405637/use-outer-instead-of-expand-grid  expand.grid.jc <- function(seq1, seq2) {   cbind(     Var1 = rep.int(seq1, length(seq2)),     Var2 = rep.int(seq2, rep.int(length(seq1), length(seq2)))   ) } # \\donttest{ microbenchmark::microbenchmark(rxExpandGrid(letters, letters), expand.grid.jc(letters, letters)) #> Unit: microseconds #>                              expr   min     lq      mean median     uq      max #>    rxExpandGrid(letters, letters) 10.61 11.121  30.09082 11.412 11.772 1713.883 #>  expand.grid.jc(letters, letters) 16.37 17.312 210.03084 17.994 18.740 5885.897 #>  neval #>    100 #>    100 # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand grid internal function — rxExpandGrid_","title":"Expand grid internal function — rxExpandGrid_","text":"Expand grid internal function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand grid internal function — rxExpandGrid_","text":"","code":"rxExpandGrid_(c1, c2, type)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand grid internal function — rxExpandGrid_","text":"c1 character vector items expanded c2 second character vector items expanded type 0 typical data frame, 1 symengine sensitivity expansion","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandGrid_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand grid internal function — rxExpandGrid_","text":"data frame (type = 0) symengine string (type=1)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"Expand /else clauses multiple different types lines.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"","code":"rxExpandIfElse(model, removeInis = TRUE, removePrint = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"model Model can character, rxode2 model.  needs normalized syntax, (...){} line.  else statement must line closing bracket statement previous line.  else statment must also contain opening bracket, like code else {} removeInis boolean indicating parameter initializations removed model. removePrint boolean indicating printing statements removed model.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"named character vector. names vector logical conditions, values lines satisfy logical conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandIfElse.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Expand if/else clauses into multiple different types of lines. — rxExpandIfElse","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand second order sensitivity — rxExpandSens2_","title":"Expand second order sensitivity — rxExpandSens2_","text":"Expand second order sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand second order sensitivity — rxExpandSens2_","text":"","code":"rxExpandSens2_(state, s1, s2)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand second order sensitivity — rxExpandSens2_","text":"state state expand s1 variables calculate sensitivity s2 variables calculate second order sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens2_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand second order sensitivity — rxExpandSens2_","text":"string symengine second order sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand sensitivity — rxExpandSens_","title":"Expand sensitivity — rxExpandSens_","text":"Expand sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand sensitivity — rxExpandSens_","text":"","code":"rxExpandSens_(state, calcSens)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand sensitivity — rxExpandSens_","text":"state state expand calcSens variables calculate sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxExpandSens_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand sensitivity — rxExpandSens_","text":"symengine string expanded sensitivity","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":null,"dir":"Reference","previous_headings":"","what":"Apply the fixed population estimated parameters — rxFixPop","title":"Apply the fixed population estimated parameters — rxFixPop","text":"Apply fixed population estimated parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Apply the fixed population estimated parameters — rxFixPop","text":"","code":"rxFixPop(ui, returnNull = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Apply the fixed population estimated parameters — rxFixPop","text":"ui rxode2 ui function returnNull boolean unchanged values return original ui (FALSE) null (TRUE)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Apply the fixed population estimated parameters — rxFixPop","text":"returnNull TRUE, NULL nothing changed, changed model ui.  returnNull FALSE, return ui matter changed .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Apply the fixed population estimated parameters — rxFixPop","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFixPop.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Apply the fixed population estimated parameters — rxFixPop","text":"","code":"One.comp.transit.allo <- function() {  ini({    # Where initial conditions/variables are specified    lktr <- log(1.15)  #log k transit (/h)    lcl  <- log(0.15)  #log Cl (L/hr)    lv   <- log(7)     #log V (L)    ALLC <- fix(0.75)  #allometric exponent cl    ALLV <- fix(1.00)  #allometric exponent v    prop.err <- 0.15   #proportional error (SD/mean)    add.err <- 0.6     #additive error (mg/L)    eta.ktr ~ 0.5    eta.cl ~ 0.1    eta.v ~ 0.1  })  model({    #Allometric scaling on weight    cl <- exp(lcl + eta.cl + ALLC * logWT70)    v  <- exp(lv + eta.v + ALLV * logWT70)    ktr <- exp(lktr + eta.ktr)    # RxODE-style differential equations are supported    d/dt(depot)   = -ktr * depot    d/dt(central) =  ktr * trans - (cl/v) * central    d/dt(trans)   =  ktr * depot - ktr * trans    ## Concentration is calculated    cp = central/v    # And is assumed to follow proportional and additive error    cp ~ prop(prop.err) + add(add.err)  }) }  m <- rxFixPop(One.comp.transit.allo) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' m #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       lktr        lcl         lv   prop.err    add.err  #>  0.1397619 -1.8971200  1.9459101  0.1500000  0.6000000  #>  #> Omega ($omega):  #>         eta.ktr eta.cl eta.v #> eta.ktr     0.5    0.0   0.0 #> eta.cl      0.0    0.1   0.0 #> eta.v       0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2          central #> 3                  3            trans #>  ── μ-referencing ($muRefTable): ──   #>   theta     eta level #> 1   lcl  eta.cl    id #> 2    lv   eta.v    id #> 3  lktr eta.ktr    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         lktr <- 0.139761942375159 #>         label(\"log k transit (/h)\") #>         lcl <- -1.89711998488588 #>         label(\"log Cl (L/hr)\") #>         lv <- 1.94591014905531 #>         label(\"log V (L)\") #>         prop.err <- c(0, 0.15) #>         label(\"proportional error (SD/mean)\") #>         add.err <- c(0, 0.6) #>         label(\"additive error (mg/L)\") #>         eta.ktr ~ 0.5 #>         eta.cl ~ 0.1 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         cl <- exp(lcl + eta.cl + 0.75 * logWT70) #>         v <- exp(lv + eta.v + 1 * logWT70) #>         ktr <- exp(lktr + eta.ktr) #>         d/dt(depot) = -ktr * depot #>         d/dt(central) = ktr * trans - (cl/v) * central #>         d/dt(trans) = ktr * depot - ktr * trans #>         cp = central/v #>         cp ~ prop(prop.err) + add(add.err) #>     }) #> }  # now everything is already fixed, so calling again will do nothing  rxFixPop(m) #>  ── rxode2-based free-form 3-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       lktr        lcl         lv   prop.err    add.err  #>  0.1397619 -1.8971200  1.9459101  0.1500000  0.6000000  #>  #> Omega ($omega):  #>         eta.ktr eta.cl eta.v #> eta.ktr     0.5    0.0   0.0 #> eta.cl      0.0    0.1   0.0 #> eta.v       0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2          central #> 3                  3            trans #>  ── μ-referencing ($muRefTable): ──   #>   theta     eta level #> 1   lcl  eta.cl    id #> 2    lv   eta.v    id #> 3  lktr eta.ktr    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         lktr <- 0.139761942375159 #>         label(\"log k transit (/h)\") #>         lcl <- -1.89711998488588 #>         label(\"log Cl (L/hr)\") #>         lv <- 1.94591014905531 #>         label(\"log V (L)\") #>         prop.err <- c(0, 0.15) #>         label(\"proportional error (SD/mean)\") #>         add.err <- c(0, 0.6) #>         label(\"additive error (mg/L)\") #>         eta.ktr ~ 0.5 #>         eta.cl ~ 0.1 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         cl <- exp(lcl + eta.cl + 0.75 * logWT70) #>         v <- exp(lv + eta.v + 1 * logWT70) #>         ktr <- exp(lktr + eta.ktr) #>         d/dt(depot) = -ktr * depot #>         d/dt(central) = ktr * trans - (cl/v) * central #>         d/dt(trans) = ktr * depot - ktr * trans #>         cp = central/v #>         cp ~ prop(prop.err) + add(add.err) #>     }) #> }  # if you call it with returnNull=TRUE when no changes have been # performed, the function will return NULL  rxFixPop(m, returnNull=TRUE) #> NULL"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":null,"dir":"Reference","previous_headings":"","what":"Clear memoise cache for rxode2 — rxForget","title":"Clear memoise cache for rxode2 — rxForget","text":"Clear memoise cache rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Clear memoise cache for rxode2 — rxForget","text":"","code":"rxForget()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Clear memoise cache for rxode2 — rxForget","text":"nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxForget.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Clear memoise cache for rxode2 — rxForget","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":null,"dir":"Reference","previous_headings":"","what":"Add/Create C functions for use in rxode2 — rxFun","title":"Add/Create C functions for use in rxode2 — rxFun","text":"Add/Create C functions use rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add/Create C functions for use in rxode2 — rxFun","text":"","code":"rxFun(name, args, cCode)  rxRmFun(name)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add/Create C functions for use in rxode2 — rxFun","text":"name can either give name user function simple R function wish convert C.  rxode2 convert R function C, name function match function name provided number arguments match R function provided.  Hence, providing R function conversion C, rest arguments implied. args gives arguments user function cCode C-code new function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxFun.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add/Create C functions for use in rxode2 — rxFun","text":"","code":"# \\donttest{ ## Right now rxode2 is not aware of the function fun ## Therefore it cannot translate it to symengine or ## Compile a model with it.  try(rxode2(\"a=fun(a,b,c)\")) #>   #>   #> Error : syntax errors (see above)  ## Note for this approach to work, it cannot interfere with C ## function names or reserved rxode2 special terms.  Therefore ## f(x) would not work since f is an alias for bioavailability.  fun <- \" double fun(double a, double b, double c) {   return a*a+b*a+c; } \" ## C-code for function  rxFun(\"fun\", c(\"a\", \"b\", \"c\"), fun) ## Added function  ## Now rxode2 knows how to translate this function to symengine  rxToSE(\"fun(a,b,c)\") #> [1] \"fun(a,b,c)\"  ## And will take a central difference when calculating derivatives  rxFromSE(\"Derivative(fun(a,b,c),a)\") #> [1] \"(fun((a)+6.05545445239334e-06,b,c)-fun(a,b,c))/6.05545445239334e-06\"  ## Of course, you could specify the derivative table manually rxD(\"fun\", list(   function(a, b, c) {     paste0(\"2*\", a, \"+\", b)   },   function(a, b, c) {     return(a)   },   function(a, b, c) {     return(\"0.0\")   } ))  rxFromSE(\"Derivative(fun(a,b,c),a)\") #> [1] \"2*a+b\"  # You can also remove the functions by `rxRmFun`  rxRmFun(\"fun\")  # you can also use R functions directly in rxode2   gg <- function(x, y) {   x + y }  f <- rxode2({  z = gg(x, y) }) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’   e <- et(1:10) |> as.data.frame()  e$x <- 1:10 e$y <- 21:30  rxSolve(f, e) #> Warning: not thread safe method, using 1 core #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time     z     x     y #>   <dbl> <dbl> <dbl> <dbl> #> 1     1    22     1    21 #> 2     2    24     2    22 #> 3     3    26     3    23 #> 4     4    28     4    24 #> 5     5    30     5    25 #> 6     6    32     6    26 #> # ℹ 4 more rows  # Note that since it touches R, it can only run single-threaded. # There are also requirements for the function: # # 1. It accepts one value per argument (numeric) # # 2. It returns one numeric value  # If it is a simple function (like gg) you can also convert it to C # using rxFun and load it into rxode2  rxFun(gg) #> → finding duplicate expressions in d(gg)/d(x)... #> → finding duplicate expressions in d(gg)/d(y)... #> converted R function 'gg' to C (will now use in rxode2) #> converted R function 'rx_gg_d_x' to C (will now use in rxode2) #> converted R function 'rx_gg_d_y' to C (will now use in rxode2) #> Added derivative table for 'gg'  rxSolve(f, e) #> compiled with R user function 'gg'; now there is a clashing C user function #> triggered a recompile to use the C user function (they are always preferred) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time     z     x     y #>   <dbl> <dbl> <dbl> <dbl> #> 1     1    22     1    21 #> 2     2    24     2    22 #> 3     3    26     3    23 #> 4     4    28     4    24 #> 5     5    30     5    25 #> 6     6    32     6    26 #> # ℹ 4 more rows  # to stop the recompile simply reassign the function f <- rxode2(f) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  rxSolve(f, e) #> ── Solved rxode2 object ── #> ── Parameters ($params): ── #> # A tibble: 1 × 0 #> ── Initial Conditions ($inits): ── #> named numeric(0) #> ── First part of data (object): ── #> # A tibble: 10 × 4 #>    time     z     x     y #>   <dbl> <dbl> <dbl> <dbl> #> 1     1    22     1    21 #> 2     2    24     2    22 #> 3     3    26     3    23 #> 4     4    28     4    24 #> 5     5    30     5    25 #> 6     6    32     6    26 #> # ℹ 4 more rows  rxRmFun(\"gg\") rm(gg) rm(f)   # You can also automatically convert a R function to R code (and # calculate first derivatives)  fun <- function(a, b, c) {   a^2+b*a+c }  rxFun(fun) #> → finding duplicate expressions in d(fun)/d(a)... #> → finding duplicate expressions in d(fun)/d(b)... #> → finding duplicate expressions in d(fun)/d(c)... #> converted R function 'fun' to C (will now use in rxode2) #> converted R function 'rx_fun_d_a' to C (will now use in rxode2) #> converted R function 'rx_fun_d_b' to C (will now use in rxode2) #> converted R function 'rx_fun_d_c' to C (will now use in rxode2) #> Added derivative table for 'fun'  # You can see the R code if you want with rxC  message(rxC(\"fun\")) #> double fun(double  a, double  b, double  c) { #>   double _lastValue=NA_REAL; #>   _lastValue = R_pow_di(a,2)+b*a+c; #>   return _lastValue; #> }  # you can also remove both the function and the # derivatives with rxRmFun(\"fun\")  rxRmFun(\"fun\")   # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":null,"dir":"Reference","previous_headings":"","what":"rxGetControl option from ui — rxGetControl","title":"rxGetControl option from ui — rxGetControl","text":"rxGetControl option ui","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxGetControl option from ui — rxGetControl","text":"","code":"rxGetControl(ui, option, default)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxGetControl option from ui — rxGetControl","text":"ui rxode2 ui object option Option get default Default value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxGetControl option from ui — rxGetControl","text":"Option (present) default value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetControl.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxGetControl option from ui — rxGetControl","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":null,"dir":"Reference","previous_headings":"","what":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"S3 method getting distribution lines rxode2 simulation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"","code":"rxGetDistributionSimulationLines(line)  # S3 method for norm rxGetDistributionSimulationLines(line)  # S3 method for dnorm rxGetDistributionSimulationLines(line)  # S3 method for t rxGetDistributionSimulationLines(line)  # S3 method for cauchy rxGetDistributionSimulationLines(line)  # S3 method for ordinal rxGetDistributionSimulationLines(line)  # S3 method for default rxGetDistributionSimulationLines(line)  # S3 method for rxUi rxGetDistributionSimulationLines(line)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"line Parsed rxode2 model environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"Lines simulation ipred dv. based idea focei parameters defined","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetDistributionSimulationLines.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This is a S3 method for getting the distribution lines for a rxode2 simulation — rxGetDistributionSimulationLines","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the linear compartment model true function — rxGetLin","title":"Get the linear compartment model true function — rxGetLin","text":"Get linear compartment model true function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the linear compartment model true function — rxGetLin","text":"","code":"rxGetLin(   model,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   verbose = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the linear compartment model true function — rxGetLin","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . linCmtSens method calculate linCmt() solutions verbose TRUE verbose linear compartmental model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the linear compartment model true function — rxGetLin","text":"model linCmt() replaced linCmtA()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetLin.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the linear compartment model true function — rxGetLin","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Get model properties without compiling it. — rxGetModel","title":"Get model properties without compiling it. — rxGetModel","text":"Get model properties without compiling .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get model properties without compiling it. — rxGetModel","text":"","code":"rxGetModel(   model,   calcSens = NULL,   calcJac = NULL,   collapseModel = NULL,   indLin = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get model properties without compiling it. — rxGetModel","text":"model rxode2 specification calcSens boolean indicating rxode2 calculate sensitivities according specified ODEs. calcJac boolean indicating rxode2 calculate Jacobain according specified ODEs. collapseModel boolean indicating rxode2 remove LHS variables calculating sensitivities. indLin Calculate inductive linearization matrices compile inductive linearization support.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get model properties without compiling it. — rxGetModel","text":"rxode2 trans list","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get model properties without compiling it. — rxGetModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Get rxode2 model from object — rxGetrxode2","title":"Get rxode2 model from object — rxGetrxode2","text":"Get rxode2 model object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get rxode2 model from object — rxGetrxode2","text":"","code":"rxGetrxode2(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get rxode2 model from object — rxGetrxode2","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxGetrxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get rxode2 model from object — rxGetrxode2","text":"rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":null,"dir":"Reference","previous_headings":"","what":"Format rxSolve and related objects as html. — rxHtml","title":"Format rxSolve and related objects as html. — rxHtml","text":"Format rxSolve related objects html.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Format rxSolve and related objects as html. — rxHtml","text":"","code":"rxHtml(x, ...)  # S3 method for rxSolve rxHtml(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Format rxSolve and related objects as html. — rxHtml","text":"x rxode2 object ... Extra arguments sent kable","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Format rxSolve and related objects as html. — rxHtml","text":"html code rxSolve object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxHtml.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Format rxSolve and related objects as html. — rxHtml","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":null,"dir":"Reference","previous_headings":"","what":"Set the preferred factoring by state — rxIndLinState","title":"Set the preferred factoring by state — rxIndLinState","text":"Set preferred factoring state","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set the preferred factoring by state — rxIndLinState","text":"","code":"rxIndLinState(preferred = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set the preferred factoring by state — rxIndLinState","text":"preferred list state's preferred factorization","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set the preferred factoring by state — rxIndLinState","text":"Nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinState.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set the preferred factoring by state — rxIndLinState","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":null,"dir":"Reference","previous_headings":"","what":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"one state ODE separated specifies incorporated matrix exponential.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"","code":"rxIndLinStrategy(strategy = c(\"curState\", \"split\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"strategy strategy inductive linearization matrix building curState Prefer parameterizing terms current state, followed first state observed term. split Split parameterization states term dividing number states term adding matrix term state.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"Nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLinStrategy.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This sets the inductive linearization strategy for matrix building — rxIndLinStrategy","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLin_.html","id":null,"dir":"Reference","previous_headings":"","what":"Inductive linearization solver — rxIndLin_","title":"Inductive linearization solver — rxIndLin_","text":"Inductive linearization solver","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLin_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Inductive linearization solver — rxIndLin_","text":"cSub = Current subject number op rxode2 solving options tp Prior time point/time zero yp Prior state;  vector size = neq; Final state updated tf Final Time InfusionRate = Rates compartment;  vector size = neq Indicator compartment \"\" cache 0 = Cache doIndLin == 0, cache > 0 = nInf-1 rxode2 matrix exponential function IndF rxode2 Inductive Linearization function F","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIndLin_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Inductive linearization solver — rxIndLin_","text":"Returns status solving 1 = Successful solve -1 = Maximum number iterations reached inductive linearization","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":null,"dir":"Reference","previous_headings":"","what":"Initial Values and State values for a rxode2 object — rxInits","title":"Initial Values and State values for a rxode2 object — rxInits","text":"Returns initial values rxDll object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Initial Values and State values for a rxode2 object — rxInits","text":"","code":"rxInits(   obj,   vec = NULL,   req = NULL,   defaultValue = 0,   noerror = FALSE,   noini = FALSE,   rxLines = FALSE )  rxInit(   obj,   vec = NULL,   req = NULL,   defaultValue = 0,   noerror = FALSE,   noini = FALSE,   rxLines = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Initial Values and State values for a rxode2 object — rxInits","text":"obj rxDll, rxode2, named vector representing default initial arguments vec supplied, named vector model. req Required names, required order ODE solver defaultValue number NA representing default value parameters missing vec, required req. noerror boolean specifying error thrown missing parameter values default = NA","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Initial Values and State values for a rxode2 object — rxInits","text":"Initial values rxDll object","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInits.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Initial Values and State values for a rxode2 object — rxInits","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":null,"dir":"Reference","previous_headings":"","what":"Invert matrix using RcppArmadillo. — rxInv","title":"Invert matrix using RcppArmadillo. — rxInv","text":"Invert matrix using RcppArmadillo.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Invert matrix using RcppArmadillo. — rxInv","text":"","code":"rxInv(matrix)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Invert matrix using RcppArmadillo. — rxInv","text":"matrix matrix inverted.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxInv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Invert matrix using RcppArmadillo. — rxInv","text":"inverse pseudo inverse matrix.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":null,"dir":"Reference","previous_headings":"","what":"Check the type of an object using Rcpp — rxIs","title":"Check the type of an object using Rcpp — rxIs","text":"Check type object using Rcpp","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check the type of an object using Rcpp — rxIs","text":"","code":"rxIs(obj, cls)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check the type of an object using Rcpp — rxIs","text":"obj Object check cls Type class.  s3 classes lists/environments primitive classes checked. matrix types distinguished numeric.matrix, integer.matrix, logical.matrix, character.matrix well traditional matrix class. Additionally checks event.data.frame data.frame object time,  evid amt. (UPPER, lower Title cases accepted)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check the type of an object using Rcpp — rxIs","text":"boolean indicating object member class.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Check the type of an object using Rcpp — rxIs","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":null,"dir":"Reference","previous_headings":"","what":"Checks if the rxode2 object was built with the current build — rxIsCurrent","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"Checks rxode2 object built current build","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"","code":"rxIsCurrent(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsCurrent.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Checks if the rxode2 object was built with the current build — rxIsCurrent","text":"boolean indicating built current rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":null,"dir":"Reference","previous_headings":"","what":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"Determine DLL associated rxode2 object loaded","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"","code":"rxIsLoaded(obj)  rxDllLoaded(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"Boolean returning rxode2 library loaded.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxIsLoaded.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Determine if the DLL associated with the rxode2 object is loaded — rxIsLoaded","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":null,"dir":"Reference","previous_headings":"","what":"Left handed Variables — rxLhs","title":"Left handed Variables — rxLhs","text":"returns model calculated variables","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Left handed Variables — rxLhs","text":"","code":"rxLhs(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Left handed Variables — rxLhs","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Left handed Variables — rxLhs","text":"character vector listing calculated parameters","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLhs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Left handed Variables — rxLhs","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":null,"dir":"Reference","previous_headings":"","what":"Lock/unlocking of rxode2 dll file — rxLock","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"Lock/unlocking rxode2 dll file","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"","code":"rxLock(obj)  rxUnlock(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxLock.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Lock/unlocking of rxode2 dll file — rxLock","text":"nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the md5 of an rxode2 object or file — rxMd5","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"md5 based model possibly extra c code supplied model.  addition md5 based syntax options, compiled rxode2 library md5, rxode2 version/repository.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"","code":"rxMd5(model, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . ... ignored arguments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"rxode2 object, return named list: file_md5 model's file's md5 parsed_md5  parsed model's file's md5. Otherwise return md5 based arguments provided","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxMd5.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Return the md5 of an rxode2 object or file — rxMd5","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":null,"dir":"Reference","previous_headings":"","what":"All model variables for a rxode2 object — rxModelVars","title":"All model variables for a rxode2 object — rxModelVars","text":"Return known model variables specified rxode2 object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"All model variables for a rxode2 object — rxModelVars","text":"","code":"rxModelVars(obj)  rxModelVarsS3(obj)  # S3 method for rxUi rxModelVarsS3(obj)  # S3 method for default rxModelVarsS3(obj)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"All model variables for a rxode2 object — rxModelVars","text":"obj rxode2 family objects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"All model variables for a rxode2 object — rxModelVars","text":"list rxode2 model properties including: params  character vector names model parameters lhs character vector names model calculated parameters state character vector compartments rxode2 object trans named vector translated model properties including type jacobian specified, C function prefixes, well C functions names called compiled model. md5 named vector gives digest model (file_md5) parsed model (parsed_md5) model  named vector giving input model (model), normalized model (comments standard syntax parsing, normModel), interim code used generate final C file parseModel","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"All model variables for a rxode2 object — rxModelVars","text":"items calculated compilation; built-rxode2 compiled DLL. allow extension, s3 hook added function rxModelVarsS3.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModelVars.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"All model variables for a rxode2 object — rxModelVars","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the rxModels information — rxModels_","title":"Get the rxModels information — rxModels_","text":"Get rxModels  information","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the rxModels information — rxModels_","text":"","code":"rxModels_(env = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the rxModels information — rxModels_","text":"env boolean returns environment models stored (TRUE), currently assigned rxode2 model variables (FALSE).","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxModels_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the rxModels information — rxModels_","text":"internal rxModels information environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the normalized model — rxNorm","title":"Get the normalized model — rxNorm","text":"get syntax preferred model processing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the normalized model — rxNorm","text":"","code":"rxNorm(obj, condition = NULL, removeInis, removeJac, removeSens)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the normalized model — rxNorm","text":"obj rxode2 family objects condition Character string logical condition use subsetting normalized model.  missing, condition set via rxCondition, return whole code conditional settings intact.  condition set rxCondition, use condition. removeInis boolean indicating parameter initialization removed model removeJac boolean indicating Jacobians removed. removeSens boolean indicating sensitivities removed.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the normalized model — rxNorm","text":"Normalized Normal syntax (comments)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxNorm.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get the normalized model — rxNorm","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":null,"dir":"Reference","previous_headings":"","what":"Optimize rxode2 for computer evaluation — rxOptExpr","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"optimizes rxode2 code computer evaluation calculating redundant expressions .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"","code":"rxOptExpr(x, msg = \"model\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"x rxode2 model can accessed rxNorm msg name type object rxode2 optimizing message optimizing.  example \"model\" produce following message optimizing model: finding duplicate expressions model...","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"Optimized rxode2 model text.  order type lhs state variables maintained evaluation sped . parameters names maintained, order may modified.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxOptExpr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Optimize rxode2 for computer evaluation — rxOptExpr","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":null,"dir":"Reference","previous_headings":"","what":"Parameters specified by the model — rxParams","title":"Parameters specified by the model — rxParams","text":"returns model's parameters required solve ODE system, can used pipe parameters rxode2 solve","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Parameters specified by the model — rxParams","text":"","code":"rxParams(obj, ...)  # S3 method for rxode2 rxParams(   obj,   constants = TRUE,   ...,   params = NULL,   inits = NULL,   iCov = NULL,   keep = NULL,   thetaMat = NULL,   omega = NULL,   dfSub = NULL,   sigma = NULL,   dfObs = NULL,   nSub = NULL,   nStud = NULL )  # S3 method for rxSolve rxParams(   obj,   constants = TRUE,   ...,   params = NULL,   inits = NULL,   iCov = NULL,   keep = NULL,   thetaMat = NULL,   omega = NULL,   dfSub = NULL,   sigma = NULL,   dfObs = NULL,   nSub = NULL,   nStud = NULL )  # S3 method for rxEt rxParams(   obj,   ...,   params = NULL,   inits = NULL,   iCov = NULL,   keep = NULL,   thetaMat = NULL,   omega = NULL,   dfSub = NULL,   sigma = NULL,   dfObs = NULL,   nSub = NULL,   nStud = NULL )  rxParam(obj, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Parameters specified by the model — rxParams","text":"obj rxode2 family objects ... arguments including scaling factors compartment.  includes S# = numeric scale compartment # dividing compartment amount scale factor, like NONMEM. constants boolean indicting constants included list parameters. Currently rxode2 parses constants variables case wish change without recompiling rxode2 model. params numeric named vector values every parameter ODE system; names must correspond parameter identifiers used ODE specification; inits vector initial values state variables (e.g., amounts compartment), order vector must state variables (e.g., PK/PD compartments); iCov data frame individual non-time varying covariates combine events dataset merge. keep Columns keep either input dataset iCov dataset.  iCov dataset, column kept per line.  input dataset, records added data LOCF (Last Observation Carried forward) imputation performed. thetaMat Named theta matrix. omega Estimate Covariance matrix. omega list, assume block matrix convert full matrix simulations.  omega NA using rxode2 ui model, subject variability described omega matrix set zero. dfSub Degrees freedom sample subject variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. sigma Named sigma covariance Cholesky decomposition covariance matrix.  names columns indicate parameters simulated.  simulated every observation solved system. sigma NA using rxode2 ui model, unexplained variability described sigma matrix set zero. dfObs Degrees freedom sample unexplained variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. nSub Number subject variabilities (ETAs) simulated every realization parameters. nStud Number virtual studies characterize uncertainty estimated parameters.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Parameters specified by the model — rxParams","text":"extracting parameters rxode2 model, character vector listing parameters model.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParams.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Parameters specified by the model — rxParams","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":null,"dir":"Reference","previous_headings":"","what":"Prepare Error function for inclusion in rxode2 — rxParseErr","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"Prepare Error function inclusion rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"","code":"rxParseErr(   x,   baseTheta,   ret = \"rx_r_\",   init = NULL,   addProp = c(\"combined2\", \"combined1\") )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"x error function baseTheta Base theta start numbering add(.) prop(.) . ret Internal return type.  changed user... init Initialization vector","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"rxode2 transformed text","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParseErr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Prepare Error function for inclusion in rxode2 — rxParseErr","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":null,"dir":"Reference","previous_headings":"","what":"Parse PK function for inclusion in rxode2 — rxParsePk","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"Parse PK function inclusion rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"","code":"rxParsePk(x, init = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"x PK function init Initialization vector","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"rxode2 transformed text.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePk.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Parse PK function for inclusion in rxode2 — rxParsePk","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":null,"dir":"Reference","previous_headings":"","what":"Prepare Pred function for inclusion in rxode2 — rxParsePred","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"Prepare Pred function inclusion rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"","code":"rxParsePred(x, init = NULL, err = NULL, addProp = c(\"combined2\", \"combined1\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"x pred function init Initialization vector","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"rxode2 transformed text.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxParsePred.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Prepare Pred function for inclusion in rxode2 — rxParsePred","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates a package from compiled rxode2 models — rxPkg","title":"Creates a package from compiled rxode2 models — rxPkg","text":"Creates package compiled rxode2 models","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates a package from compiled rxode2 models — rxPkg","text":"","code":"rxPkg(   ...,   package,   wd = getwd(),   action = c(\"install\", \"build\", \"binary\", \"create\"),   license = c(\"gpl3\", \"lgpl\", \"mit\", \"agpl3\"),   name = \"Firstname Lastname\",   fields = list() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates a package from compiled rxode2 models — rxPkg","text":"... Models build package package String package name create wd character string working directory create subdirectory according modName. specified, subdirectory named “modName.d” created populated C file, dynamic loading library, plus various working files. missing, files created (removed) temporary directory, rxode2 DLL model created current directory named rx_????_platform, example rx_129f8f97fb94a87ca49ca8dafe691e1e_i386.dll action Type action take package created license type license package. name Full name author fields named list fields add DESCRIPTION, potentially overriding default values. See use_description() can set personalized defaults using package options.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates a package from compiled rxode2 models — rxPkg","text":"function returns nothing used side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPkg.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Creates a package from compiled rxode2 models — rxPkg","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate a from a Poisson process — rxPp","title":"Simulate a from a Poisson process — rxPp","text":"Simulate Poisson process","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate a from a Poisson process — rxPp","text":"","code":"rxPp(   n,   lambda,   gamma = 1,   prob = NULL,   t0 = 0,   tmax = Inf,   randomOrder = FALSE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate a from a Poisson process — rxPp","text":"n Number time points simulate Poisson process lambda Rate Poisson process gamma Asymmetry rate Poisson process.  gamma=1.0, simulates homogenous Poisson process.  gamma<1.0, Poisson process events early, gamma > 1.0, Poisson process events late process. gamma non-zero, tmax infinite indicate end Poisson process simulated.  pharamcometric cases, end study. Internally uses rate : l(t) = lambdagamma(t/tmax)^(gamma-1) prob specified, probability function one argument, time, gives probability Poisson time t accepted rejection time. t0 starting time Poisson process tmax maximum time Poisson process randomOrder TRUE randomize order Poisson events.  default (FALSE) returns Poisson process order events occurred.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate a from a Poisson process — rxPp","text":"returns vector Poisson process times; dropout >= tmax, rest times = tmax indicate dropout equal tmax.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate a from a Poisson process — rxPp","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPp.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate a from a Poisson process — rxPp","text":"","code":"## Sample homogenous Poisson process of rate 1/10 rxPp(10, 1 / 10) #>  [1] 14.16159 22.43681 41.13405 48.71743 51.37508 54.60656 75.32184 84.62128 #>  [9] 85.80475 96.63193  ## Sample inhomogenous Poisson rate of 1/10  rxPp(10, 1 / 10, gamma = 2, tmax = 100) #>  [1] 37.13008 37.33395 38.19765 41.94718 50.45308 51.16211 65.45638 75.56937 #>  [9] 76.00046 83.59277  ## Typically the Poisson process times are in a sequential order, ## using randomOrder gives the Poisson process in random order  rxPp(10, 1 / 10, gamma = 2, tmax = 10, randomOrder = TRUE) #>  [1] 10.000000 10.000000 10.000000 10.000000  6.028535 10.000000 10.000000 #>  [8] 10.000000 10.000000 10.000000  ## This uses an arbitrary function to sample a non-homogenous Poisson process  rxPp(10, 1 / 10, prob = function(x) {   1/(1+abs(x)) }) #>  [1]  14.74011  38.01194  53.91062 235.31848 381.74524 397.80743 504.64689 #>  [8] 721.10404 756.19022 889.86201"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":null,"dir":"Reference","previous_headings":"","what":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"determined internal preferred condition name list .errIdenticalDists","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"","code":"rxPreferredDistributionName(dist)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"dist input distribution","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"Preferred distribution term","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPreferredDistributionName.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Change distribution name to the preferred distribution name term — rxPreferredDistributionName","text":"","code":"rxPreferredDistributionName(\"dt\") #> [1] \"t\"  rxPreferredDistributionName(\"add\") #> [1] \"add\"  # can be vectorized  rxPreferredDistributionName(c(\"add\",\"dt\")) #>   add    dt  #> \"add\"   \"t\""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":null,"dir":"Reference","previous_headings":"","what":"rxode2 progress bar functions — rxProgress","title":"rxode2 progress bar functions — rxProgress","text":"rxProgress sets progress bar","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxode2 progress bar functions — rxProgress","text":"","code":"rxProgress(num, core = 0L)  rxTick()  rxProgressStop(clear = TRUE)  rxProgressAbort(error = \"Aborted calculation\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxode2 progress bar functions — rxProgress","text":"num Tot number operations track core Number cores show.  1, show number cores clear Boolean telling clear progress bar completion (displayed).  default TRUE error rxProgressAbort error displayed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxode2 progress bar functions — rxProgress","text":"return NULL invisibly.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"rxode2 progress bar functions — rxProgress","text":"rxTick progress bar tick rxProgressStop stop progress bar rxProgressAbort shows abort rxProgressStop called.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxode2 progress bar functions — rxProgress","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxProgress.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"rxode2 progress bar functions — rxProgress","text":"","code":"f <- function() {   on.exit({     rxProgressAbort()   })   rxProgress(100)   for (i in 1:100) {     rxTick()     Sys.sleep(1 / 100)   }   rxProgressStop() } # \\donttest{ f() # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":null,"dir":"Reference","previous_headings":"","what":"Prune branches (ie if/else) from rxode2 — rxPrune","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"prunes branches (ie /else) rxode2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"","code":"rxPrune(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"x rxode2 model can accessed rxNorm","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"Pruned rxode2 model text.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxPrune.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Prune branches (ie if/else) from rxode2 — rxPrune","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":null,"dir":"Reference","previous_headings":"","what":"Reload rxode2 DLL — rxReload","title":"Reload rxode2 DLL — rxReload","text":"Can useful debugging","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reload rxode2 DLL — rxReload","text":"","code":"rxReload()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Reload rxode2 DLL — rxReload","text":"boolean object reloaded","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReload.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Reload rxode2 DLL — rxReload","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":null,"dir":"Reference","previous_headings":"","what":"rxRemoveControl options for UI object — rxRemoveControl","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"rxRemoveControl options UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"","code":"rxRemoveControl(ui)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"ui rxode2 ui object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRemoveControl.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxRemoveControl options for UI object — rxRemoveControl","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":null,"dir":"Reference","previous_headings":"","what":"Rename items inside of a rxode2 ui model — rxRename","title":"Rename items inside of a rxode2 ui model — rxRename","text":"rxRename() changes names individual variables, lhs, ode states using new_name = old_name syntax","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Rename items inside of a rxode2 ui model — rxRename","text":"","code":"rxRename(.data, ..., envir = parent.frame())  .rxRename(.data, ..., envir = parent.frame())  rename.rxUi(.data, ...)  rename.function(.data, ...)  # S3 method for rxUi rxRename(.data, ...)  # S3 method for `function` rxRename(.data, ...)  # S3 method for default rxRename(.data, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Rename items inside of a rxode2 ui model — rxRename","text":".data rxode2 ui function, named data consistent dplyr::rename() ... rename items envir Environment evaluation","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Rename items inside of a rxode2 ui model — rxRename","text":"New model items renamed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Rename items inside of a rxode2 ui model — rxRename","text":"similar dplyr's rename() function.  dplyr loaded, s3 methods work ui objects. Note .rxRename() internal function called renaming likely need call unless writing extension function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Rename items inside of a rxode2 ui model — rxRename","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRename.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Rename items inside of a rxode2 ui model — rxRename","text":"","code":"ocmt <- function() {   ini({     tka <- exp(0.45) # Ka     tcl <- exp(1) # Cl     ## This works with interactive models     ## You may also label the preceding line with label(\"label text\")     tv <- exp(3.45) # log V     ## the label(\"Label name\") works with all models     add.sd <- 0.7   })   model({     ka <- tka     cl <- tcl     v <- tv     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  ocmt %>% rxRename(cpParent=cp) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #>  1.568312  2.718282 31.500392  0.700000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 1.56831218549017 #>         label(\"Ka\") #>         tcl <- 2.71828182845905 #>         label(\"Cl\") #>         tv <- 31.5003923087479 #>         label(\"log V\") #>         add.sd <- c(0, 0.7) #>     }) #>     model({ #>         ka <- tka #>         cl <- tcl #>         v <- tv #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cpParent = center/v #>         cpParent ~ add(add.sd) #>     }) #> }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":null,"dir":"Reference","previous_headings":"","what":"Rep R0 for foce — rxRepR0_","title":"Rep R0 for foce — rxRepR0_","text":"Rep R0 foce","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Rep R0 for foce — rxRepR0_","text":"","code":"rxRepR0_(neta)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Rep R0 for foce — rxRepR0_","text":"neta ETA substitute","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxRepR0_.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Rep R0 for foce — rxRepR0_","text":"Returns string R code substitute rx_r expression symengine environment .s","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReservedKeywords.html","id":null,"dir":"Reference","previous_headings":"","what":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","title":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","text":"list description rxode2 supported reserved keywords","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReservedKeywords.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","text":"","code":"rxReservedKeywords"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxReservedKeywords.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"A list and description of rxode2 supported reserved keywords — rxReservedKeywords","text":"data frame 3 columns 31 rows Reserved Name Reserved Keyword Name Meaning Reserved Keyword Meaning Alias Keyword Alias","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxResidualError.html","id":null,"dir":"Reference","previous_headings":"","what":"A description of Rode2 supported residual errors — rxResidualError","title":"A description of Rode2 supported residual errors — rxResidualError","text":"description Rode2 supported residual errors","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxResidualError.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A description of Rode2 supported residual errors — rxResidualError","text":"","code":"rxResidualError"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxResidualError.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"A description of Rode2 supported residual errors — rxResidualError","text":"data frame 6 columns 181 rows Error model description type residual error Functional Form additive proportional functional form used Transformation type transformation done DV prediction code Example code residual error type addProp type add+prop residual error default equivalent lhs left handed side specification represents, either response variable, compartment specification","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":null,"dir":"Reference","previous_headings":"","what":"Load a model into a symengine environment — rxS","title":"Load a model into a symengine environment — rxS","text":"Load model symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load a model into a symengine environment — rxS","text":"","code":"rxS(x, doConst = TRUE, promoteLinSens = FALSE, envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load a model into a symengine environment — rxS","text":"x rxode2 object doConst Load constants environment well. promoteLinSens Promote solved linear compartment systems sensitivity-based solutions. envir default NULL; Environment put symengine variables .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load a model into a symengine environment — rxS","text":"rxode2/symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxS.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Load a model into a symengine environment — rxS","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":null,"dir":"Reference","previous_headings":"","what":"rxSetControl options for UI object — rxSetControl","title":"rxSetControl options for UI object — rxSetControl","text":"rxSetControl options UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxSetControl options for UI object — rxSetControl","text":"","code":"rxSetControl(ui, control)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxSetControl options for UI object — rxSetControl","text":"ui rxode2 ui object control Default value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxSetControl options for UI object — rxSetControl","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetControl.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxSetControl options for UI object — rxSetControl","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":null,"dir":"Reference","previous_headings":"","what":"Assign covariates for piping — rxSetCovariateNamesForPiping","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"Assign covariates piping","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"","code":"rxSetCovariateNamesForPiping(covariates = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"covariates NULL (covariates), list covariates. nlmixr uses function set covariates pipe nlmixr fit.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetCovariateNamesForPiping.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Assign covariates for piping — rxSetCovariateNamesForPiping","text":"","code":"# First set the name of known covariates # Note this is case sensitive  rxSetCovariateNamesForPiping(c(\"WT\",\"HT\", \"TC\"))  one.compartment <- function() {  ini({    tka <- 0.45 ; label(\"Log Ka\")    tcl <- 1 ; label(\"Log Cl\")    tv <- 3.45 ; label(\"Log V\")    eta.ka ~ 0.6    eta.cl ~ 0.3    eta.v ~ 0.1    add.err <- 0.7  })  model({    ka <- exp(tka + eta.ka)    cl <- exp(tcl + eta.cl)    v <- exp(tv + eta.v)    d / dt(depot) <- -ka * depot    d/dt(depot) <- -ka * depot    d / dt(center) <- ka * depot - cl / v * center    cp <- center / v    cp ~ add(add.err)  }) }  # now TC is detected as a covariate instead of a population parameter  one.compartment %>%   model({ka <- exp(tka + eta.ka + TC * cov_C)}) #>   #>   #> ℹ add covariate `TC` (known covariate) #> ℹ add population parameter `cov_C` and set estimate to 1 #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>     tka     tcl      tv add.err   cov_C  #>    0.45    1.00    3.45    0.70    1.00  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level covariates #> 1   tka eta.ka    id   TC*cov_C #> 2   tcl eta.cl    id            #> 3    tv  eta.v    id            #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Log Ka\") #>         tcl <- 1 #>         label(\"Log Cl\") #>         tv <- 3.45 #>         label(\"Log V\") #>         add.err <- c(0, 0.7) #>         cov_C <- 1 #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka + TC * cov_C) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.err) #>     }) #> }  # You can turn it off by simply adding it back  rxSetCovariateNamesForPiping()  one.compartment %>%   model({ka <- exp(tka + eta.ka + TC * cov_C)}) #>   #>   #> ℹ add population parameter `TC` and set estimate to 1 #> ℹ add population parameter `cov_C` and set estimate to 1 #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>     tka     tcl      tv add.err      TC   cov_C  #>    0.45    1.00    3.45    0.70    1.00    1.00  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka    0.6    0.0   0.0 #> eta.cl    0.0    0.3   0.0 #> eta.v     0.0    0.0   0.1 #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.45 #>         label(\"Log Ka\") #>         tcl <- 1 #>         label(\"Log Cl\") #>         tv <- 3.45 #>         label(\"Log V\") #>         add.err <- c(0, 0.7) #>         TC <- 1 #>         cov_C <- 1 #>         eta.ka ~ 0.6 #>         eta.cl ~ 0.3 #>         eta.v ~ 0.1 #>     }) #>     model({ #>         ka <- exp(tka + eta.ka + TC * cov_C) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) <- -ka * depot #>         d/dt(depot) <- -ka * depot #>         d/dt(center) <- ka * depot - cl/v * center #>         cp <- center/v #>         cp ~ add(add.err) #>     }) #> }  # The covariates you set with `rxSetCovariateNamesForPiping()` # are turned off every time you solve (or fit in nlmixr)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":null,"dir":"Reference","previous_headings":"","what":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"Set variables model piping automatic covarite selection","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"","code":"rxSetPipingAuto(   thetamodelVars = rex::rex(or(\"tv\", \"t\", \"pop\", \"POP\", \"Pop\", \"TV\", \"T\", \"cov\", \"err\",     \"eff\")),   covariateExceptions = rex::rex(start, or(\"wt\", \"sex\", \"crcl\", \"kout\"), end),   etaParts = c(\"eta\", \"ETA\", \"Eta\", \"ppv\", \"PPV\", \"Ppv\", \"iiv\", \"Iiv\", \"bsv\", \"Bsv\",     \"BSV\", \"bpv\", \"Bpv\", \"BPV\", \"psv\", \"PSV\", \"Psv\") )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"thetamodelVars prefixes theta model variables regular expression covariateExceptions regular expression covariates always covariates etaParts list eta prefixes/post-fixes identify variable subject variability","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"Nothing, called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"called startup set defaults, though can change wish piping can work differently individual setup","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetPipingAuto.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set the variables for the model piping automatic covarite selection — rxSetPipingAuto","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":null,"dir":"Reference","previous_headings":"","what":"Defunct setting of product — rxSetProd","title":"Defunct setting of product — rxSetProd","text":"Defunct setting product","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Defunct setting of product — rxSetProd","text":"","code":"rxSetProd(type = c(\"long double\", \"double\", \"logify\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Defunct setting of product — rxSetProd","text":"type used type product","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProd.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Defunct setting of product — rxSetProd","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":null,"dir":"Reference","previous_headings":"","what":"Set timing for progress bar — rxSetProgressBar","title":"Set timing for progress bar — rxSetProgressBar","text":"Set timing progress bar","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set timing for progress bar — rxSetProgressBar","text":"","code":"rxSetProgressBar(seconds = 1)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set timing for progress bar — rxSetProgressBar","text":"seconds sets number seconds need elapse drawing next segment progress bar.  zero turns progress bar.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set timing for progress bar — rxSetProgressBar","text":"nothing, used side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetProgressBar.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set timing for progress bar — rxSetProgressBar","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":null,"dir":"Reference","previous_headings":"","what":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"Silence rxode2's C/C++ messages","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"","code":"rxSetSilentErr(silent)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"silent can 0L \"noisy\"  1L \"silent\"","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSilentErr.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Silence some of rxode2's C/C++ messages — rxSetSilentErr","text":"TRUE; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":null,"dir":"Reference","previous_headings":"","what":"Defunct setting of sum — rxSetSum","title":"Defunct setting of sum — rxSetSum","text":"Defunct setting sum","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Defunct setting of sum — rxSetSum","text":"","code":"rxSetSum(type = c(\"pairwise\", \"fsum\", \"kahan\", \"neumaier\", \"c\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Defunct setting of sum — rxSetSum","text":"type used type product","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetSum.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Defunct setting of sum — rxSetSum","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":null,"dir":"Reference","previous_headings":"","what":"Setup the initial conditions. — rxSetupIni","title":"Setup the initial conditions. — rxSetupIni","text":"Setup initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Setup the initial conditions. — rxSetupIni","text":"","code":"rxSetupIni(obj, inits = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Setup the initial conditions. — rxSetupIni","text":"obj rxode2 object inits numeric vector initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Setup the initial conditions. — rxSetupIni","text":"initial conditions setup","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupIni.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Setup the initial conditions. — rxSetupIni","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":null,"dir":"Reference","previous_headings":"","what":"Setup the initial conditions. — rxSetupScale","title":"Setup the initial conditions. — rxSetupScale","text":"Setup initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Setup the initial conditions. — rxSetupScale","text":"","code":"rxSetupScale(obj, scale = NULL, extraArgs = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Setup the initial conditions. — rxSetupScale","text":"obj rxode2 object scale numeric vector scales extraArgs list extra args parse initial conditions.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Setup the initial conditions. — rxSetupScale","text":"setup scale changing compartment values","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSetupScale.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Setup the initial conditions. — rxSetupScale","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":null,"dir":"Reference","previous_headings":"","what":"Use Shiny to help develop an rxode2 model — rxShiny","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"Use Shiny help develop rxode2 model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"","code":"rxShiny(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   data = data.frame() )  # S3 method for rxSolve rxShiny(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   data = data.frame() )  # S3 method for default rxShiny(   object = NULL,   params = NULL,   events = NULL,   inits = NULL,   ...,   data = data.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"object rxode2 family objects. supplied 2-compartment indirect effect model used.  supplied, use model associated rxode2 object model exploration. params Initial parameters model events Event information (currently ignored) inits Initial estimates model ... arguments passed rxShiny.  Currently anything. data data like plot.  data time variable well compartment calculated variable matches rxode2 model, data added plot specific compartment calculated variable.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"Nothing; Starts shiny server","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxShiny.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Use Shiny to help develop an rxode2 model — rxShiny","text":"Zufar Mulyukov Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"Simulate Parameters Theta/Omega specification","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"","code":"rxSimThetaOmega(   params = NULL,   omega = NULL,   omegaDf = NULL,   omegaLower = as.numeric(c(R_NegInf)),   omegaUpper = as.numeric(c(R_PosInf)),   omegaIsChol = FALSE,   omegaSeparation = \"auto\",   omegaXform = 1L,   nSub = 1L,   thetaMat = NULL,   thetaLower = as.numeric(c(R_NegInf)),   thetaUpper = as.numeric(c(R_PosInf)),   thetaDf = NULL,   thetaIsChol = FALSE,   nStud = 1L,   sigma = NULL,   sigmaLower = as.numeric(c(R_NegInf)),   sigmaUpper = as.numeric(c(R_PosInf)),   sigmaDf = NULL,   sigmaIsChol = FALSE,   sigmaSeparation = \"auto\",   sigmaXform = 1L,   nCoresRV = 1L,   nObs = 1L,   dfSub = 0,   dfObs = 0,   simSubjects = TRUE,   simVariability = as.logical(c(NA_LOGICAL)) )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"params Named Vector rxode2 model parameters omega Estimate Covariance matrix. omega list, assume block matrix convert full matrix simulations.  omega NA using rxode2 ui model, subject variability described omega matrix set zero. omegaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. omegaLower Lower bounds simulated ETAs (default -Inf) omegaUpper Upper bounds simulated ETAs (default Inf) omegaIsChol Indicates omega supplied Cholesky decomposed matrix instead traditional symmetric matrix. omegaSeparation Omega separation strategy Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. omegaXform taking omega values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. nSub Number subject variabilities (ETAs) simulated every realization parameters. thetaMat Named theta matrix. thetaLower Lower bounds simulated population parameter variability (default -Inf) thetaUpper Upper bounds simulated population unexplained variability (default Inf) thetaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. thetaIsChol Indicates theta supplied Cholesky decomposed matrix instead traditional symmetric matrix. nStud Number virtual studies characterize uncertainty estimated parameters. sigma Named sigma covariance Cholesky decomposition covariance matrix.  names columns indicate parameters simulated.  simulated every observation solved system. sigma NA using rxode2 ui model, unexplained variability described sigma matrix set zero. sigmaLower Lower bounds simulated unexplained variability (default -Inf) sigmaUpper Upper bounds simulated unexplained variability (default Inf) sigmaDf Degrees freedom sigma t-distribution.  default equivalent Inf, normal distribution. sigmaIsChol Boolean indicating sigma Cholesky decomposition instead symmetric covariance sigmaSeparation separation strategy sigma; Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. sigmaXform taking sigma values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. nCoresRV Number cores used simulation sigma variables.  default 1. reproduce results need run platform number cores. reason set one, regardless number cores used threaded ODE solving. nObs Number observations simulate (sigma matrix) dfSub Degrees freedom sample subject variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. dfObs Degrees freedom sample unexplained variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. simSubjects boolean indicated rxode2 simulate subjects studies (TRUE, default) studies (FALSE) simVariability determines variability simulated. NA (default) determined solver.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"data frame simulated subjects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSimThetaOmega.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate Parameters from a Theta/Omega specification — rxSimThetaOmega","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":null,"dir":"Reference","previous_headings":"","what":"Options, Solving & Simulation of an ODE/solved system — rxSolve","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"uses rxode2 family objects, file, model specification solve ODE system.  many options solved rxode2 model, first required object, events -times optional params inits.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"","code":"rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   scale = NULL,   method = c(\"liblsoda\", \"lsoda\", \"dop853\", \"indLin\"),   sigdig = NULL,   atol = 1e-08,   rtol = 1e-06,   maxsteps = 70000L,   hmin = 0,   hmax = NA_real_,   hmaxSd = 0,   hini = 0,   maxordn = 12L,   maxords = 5L,   ...,   cores,   covsInterpolation = c(\"locf\", \"linear\", \"nocb\", \"midpoint\"),   addCov = TRUE,   sigma = NULL,   sigmaDf = NULL,   sigmaLower = -Inf,   sigmaUpper = Inf,   nCoresRV = 1L,   sigmaIsChol = FALSE,   sigmaSeparation = c(\"auto\", \"lkj\", \"separation\"),   sigmaXform = c(\"identity\", \"variance\", \"log\", \"nlmixrSqrt\", \"nlmixrLog\",     \"nlmixrIdentity\"),   nDisplayProgress = 10000L,   amountUnits = NA_character_,   timeUnits = \"hours\",   theta = NULL,   thetaLower = -Inf,   thetaUpper = Inf,   eta = NULL,   addDosing = FALSE,   stateTrim = Inf,   updateObject = FALSE,   omega = NULL,   omegaDf = NULL,   omegaIsChol = FALSE,   omegaSeparation = c(\"auto\", \"lkj\", \"separation\"),   omegaXform = c(\"variance\", \"identity\", \"log\", \"nlmixrSqrt\", \"nlmixrLog\",     \"nlmixrIdentity\"),   omegaLower = -Inf,   omegaUpper = Inf,   nSub = 1L,   thetaMat = NULL,   thetaDf = NULL,   thetaIsChol = FALSE,   nStud = 1L,   dfSub = 0,   dfObs = 0,   returnType = c(\"rxSolve\", \"matrix\", \"data.frame\", \"data.frame.TBS\", \"data.table\",     \"tbl\", \"tibble\"),   seed = NULL,   nsim = NULL,   minSS = 10L,   maxSS = 1000L,   infSSstep = 12,   strictSS = TRUE,   istateReset = TRUE,   subsetNonmem = TRUE,   maxAtolRtolFactor = 0.1,   from = NULL,   to = NULL,   by = NULL,   length.out = NULL,   iCov = NULL,   keep = NULL,   indLinPhiTol = 1e-07,   indLinPhiM = 0L,   indLinMatExpType = c(\"expokit\", \"Al-Mohy\", \"arma\"),   indLinMatExpOrder = 6L,   drop = NULL,   idFactor = TRUE,   mxhnil = 0,   hmxi = 0,   warnIdSort = TRUE,   warnDrop = TRUE,   ssAtol = 1e-08,   ssRtol = 1e-06,   safeZero = TRUE,   sumType = c(\"pairwise\", \"fsum\", \"kahan\", \"neumaier\", \"c\"),   prodType = c(\"long double\", \"double\", \"logify\"),   sensType = c(\"advan\", \"autodiff\", \"forward\", \"central\"),   linDiff = c(tlag = 1.5e-05, f = 1.5e-05, rate = 1.5e-05, dur = 1.5e-05, tlag2 =     1.5e-05, f2 = 1.5e-05, rate2 = 1.5e-05, dur2 = 1.5e-05),   linDiffCentral = c(tlag = TRUE, f = TRUE, rate = TRUE, dur = TRUE, tlag2 = TRUE, f2 =     TRUE, rate2 = TRUE, dur2 = TRUE),   resample = NULL,   resampleID = TRUE,   maxwhile = 1e+05,   atolSens = 1e-08,   rtolSens = 1e-06,   ssAtolSens = 1e-08,   ssRtolSens = 1e-06,   simVariability = NA,   nLlikAlloc = NULL,   useStdPow = FALSE,   naTimeHandle = c(\"ignore\", \"warn\", \"error\"),   addlKeepsCov = FALSE,   addlDropSs = TRUE,   ssAtDoseTime = TRUE,   ss2cancelAllPending = FALSE,   envir = parent.frame() )  # S3 method for `function` rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for rxUi rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for rxode2tos rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for nlmixr2FitData rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for nlmixr2FitCore rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for default rxSolve(   object,   params = NULL,   events = NULL,   inits = NULL,   ...,   theta = NULL,   eta = NULL,   envir = parent.frame() )  # S3 method for rxSolve update(object, ...)  # S3 method for rxode2 predict(object, ...)  # S3 method for `function` predict(object, ...)  # S3 method for rxUi predict(object, ...)  # S3 method for rxSolve predict(object, ...)  # S3 method for rxEt predict(object, ...)  # S3 method for rxParams predict(object, ...)  # S3 method for rxode2 simulate(object, nsim = 1L, seed = NULL, ...)  # S3 method for rxSolve simulate(object, nsim = 1L, seed = NULL, ...)  # S3 method for rxParams simulate(object, nsim = 1L, seed = NULL, ...)  # S3 method for rxSolve solve(a, b, ...)  # S3 method for rxUi solve(a, b, ...)  # S3 method for `function` solve(a, b, ...)  # S3 method for rxode2 solve(a, b, ...)  # S3 method for rxParams solve(a, b, ...)  # S3 method for rxEt solve(a, b, ...)  rxControl(   ...,   params = NULL,   events = NULL,   inits = NULL,   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"object either rxode2 family objects, file-name rxode2 model specification, string rxode2 model specification. params numeric named vector values every parameter ODE system; names must correspond parameter identifiers used ODE specification; events eventTable object describing input (e.g., doses) dynamic system observation sampling time points (see eventTable()); inits vector initial values state variables (e.g., amounts compartment), order vector must state variables (e.g., PK/PD compartments); scale numeric named vector scaling ode parameters system.  names must correspond parameter identifiers ODE specification. ODE variables divided scaling factor.  example scale=c(center=2) divide center ODE variable 2. method method solving ODEs.  Currently supports: \"liblsoda\" thread safe lsoda.  supports parallel thread-based solving, ignores user Jacobian specification. \"lsoda\" -- LSODA solver.  support parallel thread-based solving, allows user Jacobian specification. \"dop853\" -- DOP853 solver.  support parallel thread-based solving user Jacobian specification \"indLin\" -- Solving inductive linearization.  rxode2 dll must setup specially use solving routine. sigdig Specifies \"significant digits\" ode solving requests.  specified controls relative absolute tolerances ODE solvers.  default tolerance 0.5*10^(-sigdig-2) regular ODEs. sensitivity equations default 0.5*10\\^(-sigdig-1.5) (sensitivity changes applicable liblsoda).  also controls atol/rtol steady state solutions. ssAtol/ssRtol 0.5*10\\^(-sigdig) sensitivities 0.5*10\\^(-sigdig+0.625).  default unspecified (NULL) uses standard atol/rtol. atol numeric absolute tolerance (1e-8 default) used ODE solver determine good solution achieved;  also used solved linear model check prior doses add anything solution. rtol numeric relative tolerance (1e-6 default) used ODE solver determine good solution achieved. also used solved linear model check prior doses add anything solution. maxsteps maximum number (internally defined) steps allowed one call solver. (5000 default) hmin minimum absolute step size allowed. default value 0. hmax maximum absolute step size allowed.  hmax=NA (default), uses average difference + hmaxSd*sd times sampling events. hmaxSd user specified parameter defaults zero.  hmax=NULL rxode2 uses maximum difference times sampling events.  value 0 equivalent infinite maximum absolute step size. hmaxSd number standard deviations time difference add hmax. default 0 hini step size attempted first step. default value determined solver (hini = 0) maxordn maximum order allowed nonstiff (Adams) method.  default 12.  can 1 12. maxords maximum order allowed stiff (BDF) method.  default value 5.  can 1 5. ... arguments including scaling factors compartment.  includes S# = numeric scale compartment # dividing compartment amount scale factor, like NONMEM. cores Number cores used parallel ODE solving.  equivalent calling setRxThreads() covsInterpolation specifies interpolation method time-varying covariates. solving ODEs often samples times outside sampling time specified events. happens, time varying covariates interpolated.  Currently can : \"linear\" interpolation, interpolates covariate solving line observed covariates extrapolating new covariate value. \"constant\" -- Last observation carried forward (default). \"NOCB\" -- Next Observation Carried Backward.  method NONMEM uses. \"midpoint\" Last observation carried forward midpoint; Next observation carried backward midpoint. addCov boolean indicating covariates added output matrix data frame. default disabled. sigma Named sigma covariance Cholesky decomposition covariance matrix.  names columns indicate parameters simulated.  simulated every observation solved system. sigma NA using rxode2 ui model, unexplained variability described sigma matrix set zero. sigmaDf Degrees freedom sigma t-distribution.  default equivalent Inf, normal distribution. sigmaLower Lower bounds simulated unexplained variability (default -Inf) sigmaUpper Upper bounds simulated unexplained variability (default Inf) nCoresRV Number cores used simulation sigma variables.  default 1. reproduce results need run platform number cores. reason set one, regardless number cores used threaded ODE solving. sigmaIsChol Boolean indicating sigma Cholesky decomposition instead symmetric covariance sigmaSeparation separation strategy sigma; Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. sigmaXform taking sigma values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. nDisplayProgress integer indicating minimum number c-based solves progress bar shown.  default 10,000. amountUnits supplies dose units data frame supplied instead event table.  importing data rxode2 event table. timeUnits supplies time units data frame supplied instead event table.  importing data rxode2 event table. theta vector parameters named THETA\\[#\\] added parameters thetaLower Lower bounds simulated population parameter variability (default -Inf) thetaUpper Upper bounds simulated population unexplained variability (default Inf) eta vector parameters named ETA\\[#\\] added parameters addDosing Boolean indicating solve add rxode2 EVID related columns.  also include dosing information estimates doses.  default, rxode2 includes estimates observations. (default FALSE). addDosing NULL, include EVID=0 solve exclude model-times EVID=2. addDosing NA classic rxode2 EVID events returned. addDosing TRUE add event information NONMEM-style format; subsetNonmem=FALSE rxode2 also include extra event types (EVID) ending infusion modeled times: EVID=-1 modeled rate infusions turned (matches rate=-1) EVID=-2 modeled duration infusions turned (matches rate=-2) EVID=-10 specified rate infusions turned (matches rate>0) EVID=-20 specified dur infusions turned (matches dur>0) EVID=101,102,103,... Modeled time 101 first model time, 102 second etc. stateTrim amounts/concentrations one states value, trim value. default Inf.  Also trims -stateTrim large negative amounts/concentrations.  want trim range say c(0, 2000000) may specify 2 values lower upper range make sure state values reasonable range. updateObject internally used flag update rxode2 solved object (supplying rxode2 solved object) well returning new object.  probably modify FALSE default unless willing unexpected results. omega Estimate Covariance matrix. omega list, assume block matrix convert full matrix simulations.  omega NA using rxode2 ui model, subject variability described omega matrix set zero. omegaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. omegaIsChol Indicates omega supplied Cholesky decomposed matrix instead traditional symmetric matrix. omegaSeparation Omega separation strategy Tells type separation strategy simulating covariance parameter uncertainty standard deviations modeled thetaMat matrix. \"lkj\" simulates correlation matrix rLKJ1 matrix distribution parameter eta equal degrees freedom nu (nu-1)/2 \"separation\" simulates identity inverse Wishart covariance matrix nu degrees freedom.  converted covariance matrix augmented modeled standard deviations.  computationally complex \"lkj\" prior, performs better covariance matrix size greater equal 10 \"auto\" chooses \"lkj\" dimension matrix less 10 \"separation\" greater equal 10. omegaXform taking omega values thetaMat simulations (using separation strategy covariance simulation), thetaMat values turned int standard deviation values: identity standard deviation values directly modeled params thetaMat matrix variance params thetaMat simulates variance directly modeled thetaMat matrix log params thetaMat simulates log(sd) nlmixrSqrt params thetaMat simulates inverse cholesky decomposed matrix x\\^2 modeled along diagonal.  works diagonal matrix. nlmixrLog params thetaMat simulates inverse cholesky decomposed matrix exp(x\\^2) along diagonal.  works diagonal matrix. nlmixrIdentity params thetaMat simulates inverse cholesky decomposed matrix. works diagonal matrix. omegaLower Lower bounds simulated ETAs (default -Inf) omegaUpper Upper bounds simulated ETAs (default Inf) nSub Number subject variabilities (ETAs) simulated every realization parameters. thetaMat Named theta matrix. thetaDf degrees freedom t-distribution simulation.  default NULL equivalent Inf degrees, simulate normal distribution instead t-distribution. thetaIsChol Indicates theta supplied Cholesky decomposed matrix instead traditional symmetric matrix. nStud Number virtual studies characterize uncertainty estimated parameters. dfSub Degrees freedom sample subject variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. dfObs Degrees freedom sample unexplained variability matrix inverse Wishart distribution (scaled) scaled inverse chi squared distribution. returnType tells type object returned.  currently supported types : \"rxSolve\" (default) return reactive data frame can change easily change different pieces solve update data frame.  currently standard solving method rxode2,  used rxSolve(object, ...), solve(object,...), \"data.frame\" -- returns plain, non-reactive data frame; Currently slightly faster returnType=\"matrix\" \"matrix\" -- returns plain matrix column names attached solved object.  used object$run well object$solve \"data.table\" -- returns data.table; data.table created reference (ie setDt()), fast. \"tbl\" \"tibble\" returns tibble format. seed object specifying random number generator initialized nsim represents number simulations.  rxode2, supply single subject event tables (created [eventTable()]) minSS Minimum number iterations steady-state dose maxSS Maximum number iterations steady-state dose infSSstep Step size determining constant infusion reached steady state.  default large value, 12. strictSS Boolean indicating strict steady-state required. strict steady-state (TRUE) required least minSS doses administered total number steady states doses continue maxSS reached, atol rtol every compartment reached.  However, ODE solving problems occur minSS reached whole subject considered invalid solve. strictSS FALSE long minSS reached last good solve ODE solving problems occur considered steady state, even though either atol, rtol maxSS achieved. istateReset TRUE, reset ISTATE variable 1 lsoda liblsoda doses, like deSolve; FALSE, reset ISTATE variable doses. subsetNonmem subset NONMEM compatible EVIDs .  default TRUE. maxAtolRtolFactor maximum atol/rtol FOCEi routines may adjust .  default 0.1 observations event table, start observations value. default zero. observations event table, end observations value. default 24 + maximum dose time. observations event table, amount increment observations . length.number observations create observations event table. default 200. iCov data frame individual non-time varying covariates combine events dataset merge. keep Columns keep either input dataset iCov dataset.  iCov dataset, column kept per line.  input dataset, records added data LOCF (Last Observation Carried forward) imputation performed. indLinPhiTol requested accuracy tolerance exponential matrix. indLinPhiM maximum size Krylov basis indLinMatExpType matrix exponential type use rxode2.  Currently following supported: Al-Mohy Uses exponential matrix method Al-Mohy Higham (2009) arma Use exponential matrix RcppArmadillo expokit Use exponential matrix Roger B. Sidje (1998) indLinMatExpOrder integer, order approximation used, Al-Mohy expokit values. best value depends machine precision (slightly matrix). use 6 default. drop Columns drop output idFactor boolean indicates original ID values maintained. changes default sequentially ordered ID factor original ID values original dataset.  default enabled. mxhnil maximum number messages printed (per problem) warning T + H = T step (H = step size).  must positive result non-default value.  default value 0 (infinite). hmxi inverse maximum absolute value H used. hmxi = 0.0 allowed corresponds infinite hmax1 (default).  hminandhmximay changed time, take effect next change ofHis considered. option considered withmethod=\"liblsoda\"`. warnIdSort Warn ID present rxode2 assumes order parameters/iCov order parameters input dataset. warnDrop Warn column(s) supposed dropped, present. ssAtol Steady state atol convergence factor.  Can vector based state. ssRtol Steady state rtol convergence factor.  Can vector based state. safeZero Use safe zero divide log routines.  default turned may turn wish. sumType Sum type use sum() rxode2 code blocks. pairwise uses pairwise sum (fast, default) fsum uses PreciseSum package's fsum function (accurate) kahan uses Kahan correction neumaier uses Neumaier correction c uses correction: default/native summing prodType Product use prod() rxode2 blocks long double converts long double, performs multiplication converts back. double uses standard double scale multiplication. sensType Sensitivity type linCmt() model: advan Use direct advan solutions autodiff Use autodiff advan solutions forward Use forward difference solutions central Use central differences linDiff gives linear difference amount types linear compartment model parameters sensitivities calculated. named components numeric vector : \"lag\" Central compartment lag \"f\" Central compartment bioavailability \"rate\" Central compartment modeled rate \"dur\" Central compartment modeled duration \"lag2\" Depot compartment lag \"f2\" Depot compartment bioavailability \"rate2\" Depot compartment modeled rate \"dur2\" Depot compartment modeled duration linDiffCentral gives parameters use central differences linear compartment model parameters.  components linDiff resample character vector model variables resample input dataset; sampling done replacement. NULL FALSE resampling done.  TRUE resampling done covariates input dataset resampleID boolean representing resampling done individual basis TRUE (ie. whole patient selected) covariate resampled independent subject identifier FALSE.  resampleID=TRUE correlations parameters retained, resampleID=FALSE ignores patient covariate correaltions. Hence default resampleID=TRUE. maxwhile represents maximum times loop evaluated exiting.  default 100000 atolSens Sensitivity atol, can different atol liblsoda.  allows less accurate solve gradients (desired) rtolSens Sensitivity rtol, can different rtol liblsoda.  allows less accurate solve gradients (desired) ssAtolSens Sensitivity absolute tolerance (atol) calculating steady state achieved sensitivity compartments. ssRtolSens Sensitivity relative tolerance (rtol) calculating steady state achieved sensitivity compartments. simVariability determines variability simulated. NA (default) determined solver. nLlikAlloc number log likelihood endpoints used model.  allows independent log likelihood per endpoint focei nlmixr2.  likely set, though hurt anything (just may take memory larger allocations). useStdPow uses C's pow exponentiation instead R's R_pow R_pow_di.  default FALSE naTimeHandle Determines time handling happens time becomes NA: current options : ignore ignores NA time input passes . warn (default) produce warning end solve, continues solving passing NA time error stop solve parallel solved ODE (otherwise stopping can crash R) addlKeepsCov determines additional dosing items repeats dose (FALSE) keeps covariates record dose (TRUE) addlDropSs steady state doses addl specification steady state flag dropped repeated doses (TRUE) retained (FALSE) ssAtDoseTime Boolean TRUE back calculates steady concentration actual time dose, otherwise FALSE doses shifted ss2cancelAllPending TRUE SS=2 event type cancels pending doses like SS=1.  FALSE pending doses canceled SS=2 (infusions started SS=2 occurred canceled, though). envir environment look R user functions (defaults parent environment) using solve(), equivalent object argument.  specify object later argument list overwrites parameter. b using solve(), equivalent params argument.  specify params named argument, overwrites output","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"“rxSolve” solve object stores solved value special data.frame type determined returnType. default many rows sampled time points many columns system variables (defined ODEs additional assignments rxode2 model code).  also stores information call allow dynamic updating solved object. operations object similar data-frame, expand $ [[\"\"]] access operators assignment operators resolve based different parameter values, initial conditions, solver parameters, events (updating time variable). can call eventTable() methods solved object update event table resolve system equations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"rest document focus different ODE solving methods, followed core solving method's options, rxode2 event handling options, rxode2's numerical stability options, rxode2's output options, finally internal rxode2 options compatibility options.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"\"New Scaling Squaring Algorithm Matrix Exponential\", Awad H. Al-Mohy Nicholas J. Higham, August 2009 Roger B. Sidje (1998).  EXPOKIT: Software package computing matrix exponentials.  ACM - Transactions Mathematical Software 24(1), 130-156. Hindmarsh, . C. ODEPACK, Systematized Collection ODE Solvers. Scientific Computing, R. S. Stepleman et al. (Eds.), North-Holland, Amsterdam, 1983, pp. 55-64. Petzold, L. R. Automatic Selection Methods Solving Stiff Nonstiff Systems Ordinary Differential Equations. Siam J. Sci. Stat. Comput. 4 (1983), pp. 136-148. Hairer, E., Norsett, S. P., Wanner, G. Solving ordinary differential equations , nonstiff problems. 2nd edition, Springer Series Computational Mathematics, Springer-Verlag (1993).","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolve.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Options, Solving & Simulation of an ODE/solved system — rxSolve","text":"Matthew Fidler, Melissa Hallow  Wenping Wang","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveFree.html","id":null,"dir":"Reference","previous_headings":"","what":"Free the C solving/parsing information. — rxSolveFree","title":"Free the C solving/parsing information. — rxSolveFree","text":"Take ODE C system free .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveFree.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Free the C solving/parsing information. — rxSolveFree","text":"","code":"rxSolveFree()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveFree.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Free the C solving/parsing information. — rxSolveFree","text":"logical indicating memory successfully freed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":null,"dir":"Reference","previous_headings":"","what":"See if the memory is installed for a solve — rxSolveSetup","title":"See if the memory is installed for a solve — rxSolveSetup","text":"See memory installed solve","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"See if the memory is installed for a solve — rxSolveSetup","text":"","code":"rxSolveSetup()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"See if the memory is installed for a solve — rxSolveSetup","text":"boolean saying memnory currently free rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSolveSetup.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"See if the memory is installed for a solve — rxSolveSetup","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":null,"dir":"Reference","previous_headings":"","what":"This function splits a function based on + or - terms — rxSplitPlusQ","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"uses parser disturb terms within functions.  example:","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"","code":"rxSplitPlusQ(x, level = 0, mult = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"x Quoted R expression splitting level Internal level parsing mult boolean split based * / expressions instead. default turned .","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"character vector split expressions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"aexp(b+c)+dlog(e-f)-g*f return c(\"* exp(b + c)\", \"d * log(e - f)\", \"- g * f\")","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSplitPlusQ.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"This function splits a function based on + or - terms — rxSplitPlusQ","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":null,"dir":"Reference","previous_headings":"","what":"State variables — rxState","title":"State variables — rxState","text":"returns model's compartments states.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"State variables — rxState","text":"","code":"rxState(obj = NULL, state = NULL)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"State variables — rxState","text":"obj rxode2 family objects state string indicating state compartment like lookup.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"State variables — rxState","text":"state missing, return character vector states. state string, return compartment number named state.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxState.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"State variables — rxState","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":null,"dir":"Reference","previous_headings":"","what":"Recast model in terms of sum/prod — rxSumProdModel","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"Recast model terms sum/prod","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"","code":"rxSumProdModel(model, expand = FALSE, sum = TRUE, prod = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"model rxode2 model expand Boolean indicating expression expanded. sum Use sum(...) prod Use prod(...)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"model string prod(.) sum(.) operations.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSumProdModel.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Recast model in terms of sum/prod — rxSumProdModel","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":null,"dir":"Reference","previous_headings":"","what":"Get list of supported functions — rxSupportedFuns","title":"Get list of supported functions — rxSupportedFuns","text":"Get list supported functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get list of supported functions — rxSupportedFuns","text":"","code":"rxSupportedFuns()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get list of supported functions — rxSupportedFuns","text":"list supported functions rxode2","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSupportedFuns.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get list of supported functions — rxSupportedFuns","text":"","code":"rxSupportedFuns() #>   [1] \"gammafn\"              \"lgammafn\"             \"lgamma\"               #>   [4] \"loggamma\"             \"digamma\"              \"trigamma\"             #>   [7] \"tetragamma\"           \"pentagamma\"           \"cospi\"                #>  [10] \"sinpi\"                \"tanpi\"                \"log1p\"                #>  [13] \"expm1\"                \"factorial\"            \"lfactorial\"           #>  [16] \"lgamma1p\"             \"log10\"                \"log2\"                 #>  [19] \"log1pexp\"             \"phi\"                  \"pnorm\"                #>  [22] \"normcdf\"              \"qnorm\"                \"fabs\"                 #>  [25] \"pow\"                  \"R_pow\"                \"R_pow_di\"             #>  [28] \"Rx_pow_di\"            \"Rx_pow\"               \"lbeta\"                #>  [31] \"abs\"                  \"acos\"                 \"acosh\"                #>  [34] \"asin\"                 \"asinh\"                \"atan\"                 #>  [37] \"atan2\"                \"atanh\"                \"beta\"                 #>  [40] \"cos\"                  \"cosh\"                 \"erf\"                  #>  [43] \"erfc\"                 \"exp\"                  \"gamma\"                #>  [46] \"linCmtA\"              \"linCmtC\"              \"linCmtB\"              #>  [49] \"log\"                  \"polygamma\"            \"rxTBS\"                #>  [52] \"rxTBSi\"               \"rxTBSd\"               \"rxTBSd2\"              #>  [55] \"sin\"                  \"sinh\"                 \"sqrt\"                 #>  [58] \"tan\"                  \"tanh\"                 \"gammap\"               #>  [61] \"floor\"                \"round\"                \"ceil\"                 #>  [64] \"trunc\"                \"bessel_i\"             \"bessel_j\"             #>  [67] \"bessel_k\"             \"bessel_y\"             \"logspace_add\"         #>  [70] \"logspace_sub\"         \"fmax2\"                \"fmin2\"                #>  [73] \"sign\"                 \"fsign\"                \"fprec\"                #>  [76] \"fround\"               \"ftrunc\"               \"transit\"              #>  [79] \"gammaq\"               \"gammapDer\"            \"gammapInv\"            #>  [82] \"gammapInva\"           \"gammaqInv\"            \"gammaqInva\"           #>  [85] \"lowergamma\"           \"uppergamma\"           \"max\"                  #>  [88] \"min\"                  \"logit\"                \"expit\"                #>  [91] \"probit\"               \"probitInv\"            \"tlast\"                #>  [94] \"tfirst\"               \"lag\"                  \"lead\"                 #>  [97] \"dose\"                 \"podo\"                 \"dabs\"                 #> [100] \"dabs2\"                \"abs1\"                 \"dabs1\"                #> [103] \"erfinv\"               \"abs0\"                 \"dosenum\"              #> [106] \"first\"                \"last\"                 \"diff\"                 #> [109] \"is.nan\"               \"is.na\"                \"is.finite\"            #> [112] \"is.infinite\"          \"llikPois\"             \"llikPoisDlambda\"      #> [115] \"llikBinom\"            \"llikBinomDprob\"       \"llikNbinom\"           #> [118] \"llikNbinomDprob\"      \"llikNbinomMu\"         \"llikNbinomMuDmu\"      #> [121] \"llikBeta\"             \"llikBetaDshape1\"      \"llikBetaDshape2\"      #> [124] \"llikT\"                \"llikTDdf\"             \"llikTDmean\"           #> [127] \"llikTDsd\"             \"llikChisq\"            \"llikChisqDdf\"         #> [130] \"llikExp\"              \"llikExpDrate\"         \"llikF\"                #> [133] \"llikFDdf1\"            \"llikFDdf2\"            \"llikGeom\"             #> [136] \"llikGeomDprob\"        \"llikUnif\"             \"llikUnifDalpha\"       #> [139] \"llikUnifDbeta\"        \"llikWeibull\"          \"llikWeibullDshape\"    #> [142] \"llikWeibullDscale\"    \"llikGamma\"            \"llikGammaDshape\"      #> [145] \"llikGammaDrate\"       \"llikCauchy\"           \"llikCauchyDlocation\"  #> [148] \"llikCauchyDscale\"     \"llikNorm\"             \"llikNormDmean\"        #> [151] \"llikNormDsd\"          \"llikXPois\"            \"llikXPoisDlambda\"     #> [154] \"llikXBinom\"           \"llikXBinomDprob\"      \"llikXNbinomMu\"        #> [157] \"llikXNbinomMuDmu\"     \"llikXNbinom\"          \"llikXNbinomDprob\"     #> [160] \"llikXBeta\"            \"llikXBetaDshape1\"     \"llikXBetaDshape2\"     #> [163] \"llikXT\"               \"llikXTDdf\"            \"llikXTDmean\"          #> [166] \"llikXTDsd\"            \"llikXChisq\"           \"llikXChisqDdf\"        #> [169] \"llikXExp\"             \"llikXExpDrate\"        \"llikXF\"               #> [172] \"llikXFDdf1\"           \"llikXFDdf2\"           \"llikXGeom\"            #> [175] \"llikXGeomDprob\"       \"llikXUnif\"            \"llikXUnifDalpha\"      #> [178] \"llikXUnifDbeta\"       \"llikXWeibull\"         \"llikXWeibullDshape\"   #> [181] \"llikXWeibullDscale\"   \"llikXGamma\"           \"llikXGammaDshape\"     #> [184] \"llikXGammaDrate\"      \"llikXCauchy\"          \"llikXCauchyDlocation\" #> [187] \"llikXCauchyDscale\"    \"llikXNorm\"            \"llikXNormDmean\"       #> [190] \"llikXNormDsd\"         \"linCmt\"               \"rnorm\"                #> [193] \"rxnorm\"               \"rxbinom\"              \"rbinom\"               #> [196] \"rxcauchy\"             \"rcauchy\"              \"rchisq\"               #> [199] \"rxchisq\"              \"rexp\"                 \"rxexp\"                #> [202] \"rbeta\"                \"rxbeta\"               \"rgeom\"                #> [205] \"rxgeom\"               \"rxpois\"               \"rpois\"                #> [208] \"rxt\"                  \"rt\""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":null,"dir":"Reference","previous_headings":"","what":"Respect suppress messages — rxSuppressMsg","title":"Respect suppress messages — rxSuppressMsg","text":"turns silent REprintf C suppressMessages() turned . makes REprintf act like messages R, can suppressed suppressMessages()","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Respect suppress messages — rxSuppressMsg","text":"","code":"rxSuppressMsg()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Respect suppress messages — rxSuppressMsg","text":"Nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Respect suppress messages — rxSuppressMsg","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSuppressMsg.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Respect suppress messages — rxSuppressMsg","text":"","code":"# rxSupressMsg() is called with rxode2()  # Note the errors are output to the console  try(rxode2(\"d/dt(matt)=/3\"), silent = TRUE) #>   #>    # When using suppressMessages, the output is suppressed  suppressMessages(try(rxode2(\"d/dt(matt)=/3\"), silent = TRUE))  # In rxode2, we use REprintf so that interrupted threads do not crash R # if there is a user interrupt. This isn't captured by R's messages, but # This interface allows the `suppressMessages()` to suppress the C printing # as well  # If you  want to suppress messages from rxode2 in other packages, you can use # this function"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":null,"dir":"Reference","previous_headings":"","what":"Get Omega^-1 and derivatives — rxSymInvChol","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"Get Omega^-1 derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"","code":"rxSymInvChol(   invObjOrMatrix,   theta = NULL,   type = \"cholOmegaInv\",   thetaNumber = 0L )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"invObjOrMatrix Object inverse-type calculations.  matrix, setup object inversion rxSymInvCholCreate() default arguments return reactive s3 object.  Otherwise, use inversion object calculate requested derivative/inverse. theta Thetas used calculation.  missing (NULL), special s3 class created returned access Omega^1 objects needed cache based theta used. type type object.  Currently following types supported: cholOmegaInv gives Cholesky decomposition Omega Inverse matrix. omegaInv gives Omega Inverse matrix. d(omegaInv) gives d(Omega^-1) withe respect theta parameter specified thetaNumber. d(D) gives d(diagonal(Omega^-1)) respect theta parameter specified thetaNumber parameter thetaNumber types d(omegaInv) d(D), theta number derivative taken .  must positive 1 number thetas defining Omega matrix.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"Matrix based parameters environment matrixes calculated variables omega, omegaInv, dOmega, dOmegaInv.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvChol.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Get Omega^-1 and derivatives — rxSymInvChol","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"Creates object calculating Omega/Omega^-1 derivatives","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"","code":"rxSymInvCholCreate(   mat,   diag.xform = c(\"sqrt\", \"log\", \"identity\"),   create.env = TRUE,   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"mat Initial Omega matrix diag.xform transformation diagonal elements OMEGA. chol(Omega^-1) create.env -- Create environment calculate inverses. (default TRUE) envir -- Environment evaluate function, bu default parent frame.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"rxSymInv object rxSymInv environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholCreate.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Creates an object for calculating Omega/Omega^-1 and derivatives — rxSymInvCholCreate","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholN.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","title":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","text":"Return dimension built-derivatives/inverses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholN.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","text":"","code":"rxSymInvCholN()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSymInvCholN.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the dimension of the built-in derivatives/inverses — rxSymInvCholN","text":"dimension built-derivatives/inverses","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":null,"dir":"Reference","previous_headings":"","what":"Sync options with rxode2 variables — rxSyncOptions","title":"Sync options with rxode2 variables — rxSyncOptions","text":"Accessing rxode2 options via getOption slows solving.  allows options synced variables.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Sync options with rxode2 variables — rxSyncOptions","text":"","code":"rxSyncOptions(setDefaults = c(\"none\", \"permissive\", \"strict\"))"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Sync options with rxode2 variables — rxSyncOptions","text":"setDefaults setup rxode2's default solving options following options: \"none\" leave options alone \"permissive\" permissive option set similar R language specifications. \"strict\" strict option set similar original rxode2(). requires semicolons end lines equals assignment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Sync options with rxode2 variables — rxSyncOptions","text":"nothing; called side effects","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyncOptions.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Sync options with rxode2 variables — rxSyncOptions","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyntaxFunctions.html","id":null,"dir":"Reference","previous_headings":"","what":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","title":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","text":"list description Rode supported syntax functions","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyntaxFunctions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","text":"","code":"rxSyntaxFunctions"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxSyntaxFunctions.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"A list and description of Rode supported syntax functions — rxSyntaxFunctions","text":"data frame 3 columns 98 rows Function Reserved function Name Description Description function Aliases Function Aliases","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTempDir.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the rxode2 temporary directory — rxTempDir","title":"Get the rxode2 temporary directory — rxTempDir","text":"Get rxode2 temporary directory","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTempDir.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the rxode2 temporary directory — rxTempDir","text":"","code":"rxTempDir()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTempDir.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the rxode2 temporary directory — rxTempDir","text":"rxode2 temporary directory.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":null,"dir":"Reference","previous_headings":"","what":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"rxTheme ggplot2 theme rxode2 plots","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"","code":"rxTheme(   base_size = 11,   base_family = \"\",   base_line_size = base_size/22,   base_rect_size = base_size/22,   grid = TRUE )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"base_size base font size, given pts. base_family base font family base_line_size base size line elements base_rect_size base size rect elements grid Boolean indicating grid (TRUE) (FALSE). also character indicating x y.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTheme.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxTheme is the ggplot2 theme for rxode2 plots — rxTheme","text":"ggplot2 theme used rxode2","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":null,"dir":"Reference","previous_headings":"","what":"rxode2 to symengine environment — rxToSE","title":"rxode2 to symengine environment — rxToSE","text":"rxode2 symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"rxode2 to symengine environment — rxToSE","text":"","code":"rxToSE(   x,   envir = NULL,   progress = FALSE,   promoteLinSens = TRUE,   parent = parent.frame() )  .rxToSE(x, envir = NULL, progress = FALSE)  rxFromSE(   x,   unknownDerivatives = c(\"forward\", \"central\", \"error\"),   parent = parent.frame() )  .rxFromSE(x)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"rxode2 to symengine environment — rxToSE","text":"x expression envir default NULL; Environment put symengine variables . progress shows progress bar true. promoteLinSens Promote solved linear compartment systems sensitivity-based solutions. parent parent environment look R-based user functions unknownDerivatives handling derivatives unknown functions, translator translate different types numeric derivatives.  currently supported methods :","code":"- `forward` for forward differences - `central` for central differences - `error` for throwing an error for unknown derivatives"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"rxode2 to symengine environment — rxToSE","text":"rxode2 symengine environment","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxToSE.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"rxode2 to symengine environment — rxToSE","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":null,"dir":"Reference","previous_headings":"","what":"Translate the model to C code if needed — rxTrans","title":"Translate the model to C code if needed — rxTrans","text":"function translates model C code, needed","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Translate the model to C code if needed — rxTrans","text":"","code":"rxTrans(   model,   modelPrefix = \"\",   md5 = \"\",   modName = NULL,   modVars = FALSE,   ... )  # S3 method for default rxTrans(   model,   modelPrefix = \"\",   md5 = \"\",   modName = NULL,   modVars = FALSE,   ... )  # S3 method for character rxTrans(   model,   modelPrefix = \"\",   md5 = \"\",   modName = NULL,   modVars = FALSE,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Translate the model to C code if needed — rxTrans","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . modelPrefix Prefix model functions compiled make sure multiple rxode2 objects can coexist R session. md5 md5 model parsing, used embed md5 DLL, provide functions like rxModelVars(). modName string used model name. string used naming various aspects computations, including generating C symbol names, dynamic libraries, etc. Therefore, necessary modName consists simple ASCII alphanumeric characters starting letter. modVars returns model variables instead named vector translated properties. ... Ignored parameters.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Translate the model to C code if needed — rxTrans","text":"named vector translated model properties including type jacobian specified, C function prefixes, well C functions names called compiled model.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxTrans.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Translate the model to C code if needed — rxTrans","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":null,"dir":"Reference","previous_headings":"","what":"Compress/Decompress rxode2 ui — rxUiDecompress","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"Compress/Decompress rxode2 ui","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"","code":"rxUiDecompress(ui)  rxUiCompress(ui)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"ui rxode2 ui object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"compressed decompressed rxui object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiDecompress.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compress/Decompress rxode2 ui — rxUiDecompress","text":"","code":"one.cmt <- function() {   ini({     ## You may label each parameter with a comment     tka <- 0.45 # Log Ka     tcl <- log(c(0, 2.7, 100)) # Log Cl     ## This works with interactive models     ## You may also label the preceding line with label(\"label text\")     tv <- 3.45; label(\"log V\")     ## the label(\"Label name\") works with all models     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     linCmt() ~ add(add.sd) | tmp   }) }  f <- rxode2(one.cmt) #>   #>   #> ℹ parameter labels from comments will be replaced by 'label()' print(class(f)) #> [1] \"rxUi\" \"raw\"  print(is.environment(f)) #> [1] FALSE  f  <- rxUiDecompress(f) print(class(f)) #> [1] \"rxUi\" print(is.environment(f)) #> [1] TRUE  f  <- rxUiCompress(f) print(class(f)) #> [1] \"rxUi\" \"raw\"  print(is.environment(f)) #> [1] FALSE"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":null,"dir":"Reference","previous_headings":"","what":"S3 for getting information from UI model — rxUiGet.cmtLines","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"S3 getting information UI model","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"","code":"# S3 method for cmtLines rxUiGet(x, ...)  # S3 method for dvidLine rxUiGet(x, ...)  # S3 method for paramsLine rxUiGet(x, ...)  # S3 method for simulationSigma rxUiGet(x, ...)  # S3 method for simulationModel rxUiGet(x, ...)  # S3 method for symengineModelNoPrune rxUiGet(x, ...)  # S3 method for symengineModelPrune rxUiGet(x, ...)  # S3 method for simulationIniModel rxUiGet(x, ...)  rxUiGet(x, ...)  # S3 method for params rxUiGet(x, ...)  # S3 method for theta rxUiGet(x, ...)  # S3 method for lstChr rxUiGet(x, ...)  # S3 method for omega rxUiGet(x, ...)  # S3 method for funTxt rxUiGet(x, ...)  # S3 method for allCovs rxUiGet(x, ...)  # S3 method for muRefTable rxUiGet(x, ...)  # S3 method for multipleEndpoint rxUiGet(x, ...)  # S3 method for funPrint rxUiGet(x, ...)  # S3 method for fun rxUiGet(x, ...)  # S3 method for md5 rxUiGet(x, ...)  # S3 method for ini rxUiGet(x, ...)  # S3 method for iniFun rxUiGet(x, ...)  # S3 method for modelFun rxUiGet(x, ...)  # S3 method for model rxUiGet(x, ...)  # S3 method for modelDesc rxUiGet(x, ...)  # S3 method for thetaLower rxUiGet(x, ...)  # S3 method for thetaUpper rxUiGet(x, ...)  # S3 method for lhsVar rxUiGet(x, ...)  # S3 method for varLhs rxUiGet(x, ...)  # S3 method for lhsEta rxUiGet(x, ...)  # S3 method for lhsTheta rxUiGet(x, ...)  # S3 method for lhsCov rxUiGet(x, ...)  # S3 method for etaLhs rxUiGet(x, ...)  # S3 method for thetaLhs rxUiGet(x, ...)  # S3 method for covLhs rxUiGet(x, ...)  # S3 method for default rxUiGet(x, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"x list (UIenvironment, exact).  UI environment parsed function rxode2.  exact boolean says exact match required. ... arguments","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"value requested UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUiGet.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"S3 for getting information from UI model — rxUiGet.cmtLines","text":"Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":null,"dir":"Reference","previous_headings":"","what":"Unloads all rxode2 compiled DLLs — rxUnloadAll","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"Unloads rxode2 compiled DLLs","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"","code":"rxUnloadAll()"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"List rxode2 dlls still loaded boolean rxode2 dlls unloaded","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUnloadAll.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Unloads all rxode2 compiled DLLs — rxUnloadAll","text":"","code":"print(rxUnloadAll()) #> NULL"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":null,"dir":"Reference","previous_headings":"","what":"Use model object in your package — rxUse","title":"Use model object in your package — rxUse","text":"Use model object package","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Use model object in your package — rxUse","text":"","code":"rxUse(obj, overwrite = TRUE, compress = \"bzip2\", internal = FALSE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Use model object in your package — rxUse","text":"obj model save. overwrite default, use_data() overwrite existing files. really want , set TRUE. compress Choose type compression used save(). one \"gzip\", \"bzip2\", \"xz\". internal run internally.  default FALSE","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxUse.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Use model object in your package — rxUse","text":"Nothing; used side effects called user","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":null,"dir":"Reference","previous_headings":"","what":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"Validate rxode2 allows easy validation/qualification nlmixr running testing suite system.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"","code":"rxValidate(type = NULL, skipOnCran = TRUE)  rxTest(type = NULL, skipOnCran = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"type Type test filter test type, expression, evaluate contents, respecting skipOnCran skipOnCran TRUE skip test CRAN.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"nothing","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxValidate.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Validate rxode2 This allows easy validation/qualification of nlmixr by running the testing suite on your system. — rxValidate","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":null,"dir":"Reference","previous_headings":"","what":"Version and repository for this dparser package. — rxVersion","title":"Version and repository for this dparser package. — rxVersion","text":"Version repository dparser package.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Version and repository for this dparser package. — rxVersion","text":"","code":"rxVersion(   extra = \"\",   echo = FALSE,   version = sessionInfo()$otherPkgs$rxode2$Version )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Version and repository for this dparser package. — rxVersion","text":"extra extra text display logo echo Boolean echo text logo, default FALSE version Version display/return","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Version and repository for this dparser package. — rxVersion","text":"character vector version repository.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxVersion.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Version and repository for this dparser package. — rxVersion","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate beta variable from threefry generator — rxbeta","title":"Simulate beta variable from threefry generator — rxbeta","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate beta variable from threefry generator — rxbeta","text":"","code":"rxbeta(shape1, shape2, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate beta variable from threefry generator — rxbeta","text":"shape1, shape2 non-negative parameters Beta distribution. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate beta variable from threefry generator — rxbeta","text":"beta random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate beta variable from threefry generator — rxbeta","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbeta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate beta variable from threefry generator — rxbeta","text":"","code":"# \\donttest{  ## Use threefry engine  rxbeta(0.5, 0.5, n = 10) # with rxbeta you have to explicitly state n #>  [1] 0.477852330 0.303485792 0.039549351 0.172127808 0.029142179 0.228528849 #>  [7] 0.002045024 0.123955210 0.122974648 0.003892215 rxbeta(5, 1, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 0.9837921 0.9619446 0.9528713 0.8186334 0.8958941 0.9547040 0.9377216 #>  [8] 0.8771444 0.9122368 0.9036474  rxbeta(1, 3) #> [1] 0.4918356   ## This example uses `rxbeta` directly in the model  rx <- function() {   model({     a <- rxbeta(2, 2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Binomial variable from threefry generator — rxbinom","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"","code":"rxbinom(size, prob, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"size number trials (zero ). prob probability success trial. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"binomial random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxbinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Binomial variable from threefry generator — rxbinom","text":"","code":"# \\donttest{ ## Use threefry engine  rxbinom(10, 0.9, n = 10) # with rxbinom you have to explicitly state n #>  [1] 10  9  9 10  9  9  9 10 10 10 rxbinom(3, 0.5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 0 3 0 2 2 1 2 1 1 3  rxbinom(4, 0.7) #> [1] 2   ## This example uses `rxbinom` directly in the model  rx <- function() {   model({     a <- rxbinom(1, 0.5)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Cauchy variable from threefry generator — rxcauchy","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"","code":"rxcauchy(location = 0, scale = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"location, scale location scale parameters. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"Cauchy random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxcauchy.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Cauchy variable from threefry generator — rxcauchy","text":"","code":"# \\donttest{  ## Use threefry engine  rxcauchy(0, 1, n = 10) # with rxcauchy you have to explicitly state n #>  [1] -4.058532e-01 -3.735548e-01  2.292820e+00  1.715716e+01 -3.247341e+04 #>  [6] -1.381140e+00  6.254309e-02  1.008986e+02  1.192805e+01  5.395418e-01 rxcauchy(0.5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  1.35507665 -0.09529856  0.15545628  0.52832597  2.91300334  0.67517336 #>  [7]  2.74087058  0.92316592  1.35864663 -0.36322597  rxcauchy(3) #> [1] 6.705508   ## This example uses `rxcauchy` directly in the model  rx <- function() {   model({     a <- rxcauchy(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate chi-squared variable from threefry generator — rxchisq","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"","code":"rxchisq(df, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"df degrees freedom (non-negative, can non-integer). n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"chi squared random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxchisq.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate chi-squared variable from threefry generator — rxchisq","text":"","code":"# \\donttest{  ## Use threefry engine  rxchisq(0.5, n = 10) # with rxchisq you have to explicitly state n #>  [1] 8.310255e-01 5.056546e-04 1.283680e-02 1.156829e-01 1.791516e-01 #>  [6] 1.262242e+00 7.069398e-01 7.941145e-01 2.610348e-07 7.152674e+00 rxchisq(5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 2.285712 4.831273 3.575691 8.457173 6.136148 4.831445 3.387999 1.390385 #>  [9] 4.401743 4.226414  rxchisq(1) #> [1] 0.4391989   ## This example uses `rxchisq` directly in the model  rx <- function() {   model({     a <- rxchisq(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate exponential variable from threefry generator — rxexp","title":"Simulate exponential variable from threefry generator — rxexp","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate exponential variable from threefry generator — rxexp","text":"","code":"rxexp(rate, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate exponential variable from threefry generator — rxexp","text":"rate vector rates. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate exponential variable from threefry generator — rxexp","text":"exponential random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate exponential variable from threefry generator — rxexp","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxexp.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate exponential variable from threefry generator — rxexp","text":"","code":"# \\donttest{  ## Use threefry engine  rxexp(0.5, n = 10) # with rxexp you have to explicitly state n #>  [1] 2.3809833 1.2842193 2.9579776 0.2992783 2.2362850 1.1166611 0.1232554 #>  [8] 0.3223169 3.1258548 0.1552611 rxexp(5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 0.18422563 0.04552514 0.03749666 0.07330912 0.13716012 0.77328162 #>  [7] 0.40892297 0.07167627 0.24915418 0.04749041  rxexp(1) #> [1] 1.054455   ## This example uses `rxexp` directly in the model  rx <- function() {   model({     a <- rxexp(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate F variable from threefry generator — rxf","title":"Simulate F variable from threefry generator — rxf","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate F variable from threefry generator — rxf","text":"","code":"rxf(df1, df2, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate F variable from threefry generator — rxf","text":"df1, df2 degrees freedom.  Inf allowed. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate F variable from threefry generator — rxf","text":"f random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate F variable from threefry generator — rxf","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxf.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate F variable from threefry generator — rxf","text":"","code":"# \\donttest{  ## Use threefry engine  rxf(0.5, 0.5, n = 10) # with rxf you have to explicitly state n #>  [1] 8.504451e+00 3.630793e+00 8.678236e-03 3.880547e+00 2.243042e+06 #>  [6] 2.801723e-01 4.225989e-01 1.000520e-01 3.304096e-01 1.149793e+00 rxf(5, 1, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  2.2338907 11.2163069  0.1898719  4.8782950  2.2084744  7.4676924 #>  [7]  1.2271939  7.7742898  0.1681263  4.3235439  rxf(1, 3) #> [1] 0.01338054   ## This example uses `rxf` directly in the model  rx <- function() {   model({     a <- rxf(2, 2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate gamma variable from threefry generator — rxgamma","title":"Simulate gamma variable from threefry generator — rxgamma","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate gamma variable from threefry generator — rxgamma","text":"","code":"rxgamma(shape, rate = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate gamma variable from threefry generator — rxgamma","text":"shape shape gamma random variable rate alternative way specify scale. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate gamma variable from threefry generator — rxgamma","text":"gamma random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate gamma variable from threefry generator — rxgamma","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate gamma variable from threefry generator — rxgamma","text":"","code":"# \\donttest{  ## Use threefry engine  rxgamma(0.5, n = 10) # with rxgamma you have to explicitly state n #>  [1] 1.355794e-06 1.184154e-01 9.402169e-03 2.871131e-01 5.795578e-01 #>  [6] 1.362098e+00 2.274397e-01 2.389341e-01 2.283278e-02 2.453927e-01 rxgamma(5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 6.081028 4.553699 4.216456 3.053786 2.837575 3.992559 2.974116 4.653716 #>  [9] 5.765096 3.788091  rxgamma(1) #> [1] 0.4450636   ## This example uses `rxbeta` directly in the model  rx <- function() {   model({     a <- rxgamma(2)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate geometric variable from threefry generator — rxgeom","title":"Simulate geometric variable from threefry generator — rxgeom","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate geometric variable from threefry generator — rxgeom","text":"","code":"rxgeom(prob, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate geometric variable from threefry generator — rxgeom","text":"prob probability success trial. 0 < prob <= 1. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate geometric variable from threefry generator — rxgeom","text":"geometric random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate geometric variable from threefry generator — rxgeom","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxgeom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate geometric variable from threefry generator — rxgeom","text":"","code":"# \\donttest{  ## Use threefry engine  rxgeom(0.5, n = 10) # with rxgeom you have to explicitly state n #>  [1] 1 1 0 0 0 1 3 2 0 0 rxgeom(0.25, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  0  8  0 13  1  5  6  1 11  6  rxgeom(0.75) #> [1] 0   ## This example uses `rxgeom` directly in the model  rx <- function() {   model({     a <- rxgeom(0.24)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Binomial variable from threefry generator — rxnbinom","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"","code":"rxnbinom(size, prob, n = 1L, ncores = 1L)  rxnbinomMu(size, mu, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"size target number successful trials, dispersion     parameter (shape parameter gamma mixing distribution).     Must strictly positive, need integer. prob probability success trial. 0 < prob <= 1. n number observations.  length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks mu alternative parametrization via mean: see ‘Details’.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"negative binomial random deviates. Note rxbinom2 uses mu parameterization rxbinom uses prob parameterization (mu=size/(prob+size))","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnbinom.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Binomial variable from threefry generator — rxnbinom","text":"","code":"# \\donttest{ ## Use threefry engine  rxnbinom(10, 0.9, n = 10) # with rxbinom you have to explicitly state n #>  [1]  9 10  9  8 10 10  8  9  9 10 rxnbinom(3, 0.5, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 3 2 1 3 3 1 3 3 2 2  rxnbinom(4, 0.7) #> [1] 4  # use mu parameter rxnbinomMu(40, 40, n=10) #>  [1] 20 24 20 21 21 22 20 21 20 12  ## This example uses `rxbinom` directly in the model  rx <- function() {   model({     a <- rxnbinom(10, 0.5)   }) }  et <- et(1, id = 1:100)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  rx <- function() {   model({     a <- rxnbinomMu(10, 40)   }) }  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate random normal variable from threefry generator — rxnormV","title":"Simulate random normal variable from threefry generator — rxnormV","text":"Simulate random normal variable threefry generator","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate random normal variable from threefry generator — rxnormV","text":"","code":"rxnormV(mean = 0, sd = 1, n = 1L, ncores = 1L)  rxnorm(mean = 0, sd = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate random normal variable from threefry generator — rxnormV","text":"mean vector means. sd vector standard deviations. n number observations ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate random normal variable from threefry generator — rxnormV","text":"normal random number deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxnormV.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate random normal variable from threefry generator — rxnormV","text":"","code":"# \\donttest{ ## Use threefry engine  rxnorm(n = 10) # with rxnorm you have to explicitly state n #>  [1]  0.37076053 -0.96034099  0.02749010 -0.82762453  0.58984915 -1.80511477 #>  [7]  0.50379739 -0.67643890  1.45510439  0.04165521 rxnorm(n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] -0.51162593 -1.13782819  0.06566640 -0.12404287  1.79306612  0.83099897 #>  [7] -0.44485847  0.06223717  1.88310497  0.22993201  rxnorm(2, 3) ## The first 2 arguments are the mean and standard deviation #> [1] -0.9022625   ## This example uses `rxnorm` directly in the model  rx <- function() {   model({     a <- rxnorm()   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":null,"dir":"Reference","previous_headings":"","what":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"Set function body rxUi object retaining object information (like data)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"","code":"rxode2(x, envir = environment(x)) <- value  # S3 method for `function` rxode2(x, envir = environment(x)) <- value  # S3 method for default rxode2(x, envir = environment(x)) <- value  rxode(x, envir = environment(x)) <- value  RxODE(x, envir = environment(x)) <- value"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"x rxUi object envir environment assignment ocurs value value assigned","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"rxode2 ui/function","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2-set.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Set the function body of an rxUi object while retaining other object information (like data) — rxode2<-","text":"","code":"one.compartment <- function() {   ini({     tka <- log(1.57); label(\"Ka\")     tcl <- log(2.72); label(\"Cl\")     tv <- log(31.5); label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  two.compartment <- function() {   ini({     lka <- 0.45 ; label(\"Absorption rate (Ka)\")     lcl <- 1 ; label(\"Clearance (CL)\")     lvc  <- 3 ; label(\"Central volume of distribution (V)\")     lvp  <- 5 ; label(\"Peripheral volume of distribution (Vp)\")     lq  <- 0.1 ; label(\"Intercompartmental clearance (Q)\")     propSd <- 0.5 ; label(\"Proportional residual error (fraction)\")   })   model({     ka <- exp(lka)     cl <- exp(lcl)     vc <- exp(lvc)     vp <- exp(lvp)     q  <- exp(lq)     kel <- cl/vc     k12 <- q/vc     k21 <- q/vp     d/dt(depot) <- -ka*depot     d/dt(central) <-  ka*depot - kel*central - k12*central + k21*peripheral1     d/dt(peripheral1) <- k12*central - k21*peripheral1     cp <- central / vc     cp ~ prop(propSd)   }) }  ui <- rxode2(one.compartment) #>   #>    rxode2(ui) <- two.compartment"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Create an ODE-based model specification — rxode2","title":"Create an ODE-based model specification — rxode2","text":"Create dynamic ODE-based model object suitably translation fast C code","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Create an ODE-based model specification — rxode2","text":"","code":"rxode2(   model,   modName = basename(wd),   wd = getwd(),   filename = NULL,   extraC = NULL,   debug = FALSE,   calcJac = NULL,   calcSens = NULL,   collapseModel = FALSE,   package = NULL,   ...,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   indLin = FALSE,   verbose = FALSE,   fullPrint = getOption(\"rxode2.fullPrint\", FALSE),   envir = parent.frame() )  RxODE(   model,   modName = basename(wd),   wd = getwd(),   filename = NULL,   extraC = NULL,   debug = FALSE,   calcJac = NULL,   calcSens = NULL,   collapseModel = FALSE,   package = NULL,   ...,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   indLin = FALSE,   verbose = FALSE,   fullPrint = getOption(\"rxode2.fullPrint\", FALSE),   envir = parent.frame() )  rxode(   model,   modName = basename(wd),   wd = getwd(),   filename = NULL,   extraC = NULL,   debug = FALSE,   calcJac = NULL,   calcSens = NULL,   collapseModel = FALSE,   package = NULL,   ...,   linCmtSens = c(\"linCmtA\", \"linCmtB\", \"linCmtC\"),   indLin = FALSE,   verbose = FALSE,   fullPrint = getOption(\"rxode2.fullPrint\", FALSE),   envir = parent.frame() )"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Create an ODE-based model specification — rxode2","text":"model ODE model specification.  can : string containing set ordinary differential equations (ODE) expressions defining changes dynamic system. file name ODE system equation contained ODE expression enclosed \\{\\} (see also filename argument). details, see sections “Details” rxode2 Syntax . modName string used model name. string used naming various aspects computations, including generating C symbol names, dynamic libraries, etc. Therefore, necessary modName consists simple ASCII alphanumeric characters starting letter. wd character string working directory create subdirectory according modName. specified, subdirectory named “modName.d” created populated C file, dynamic loading library, plus various working files. missing, files created (removed) temporary directory, rxode2 DLL model created current directory named rx_????_platform, example rx_129f8f97fb94a87ca49ca8dafe691e1e_i386.dll filename file name connection object ODE-based model specification resides. one model filename may specified. extraC Extra c code include model.  can useful specify functions model.  C functions usually take double precision arguments, return double precision values. debug boolean indicating executable compiled verbose debugging information turned . calcJac boolean indicating rxode2 calculate Jacobain according specified ODEs. calcSens boolean indicating rxode2 calculate sensitivities according specified ODEs. collapseModel boolean indicating rxode2 remove LHS variables calculating sensitivities. package Package name pre-compiled binaries. ... ignored arguments. linCmtSens method calculate linCmt() solutions indLin Calculate inductive linearization matrices compile inductive linearization support. verbose TRUE verbose linear compartmental model fullPrint using printf within model, TRUE print every step (except /indLin), otherwise FALSE print calculating d/dt envir environment look R user functions (defaults parent environment)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Create an ODE-based model specification — rxode2","text":"object (environment) class rxode2 (see Chambers Temple Lang (2001)) consisting following list strings functions:","code":"* `model` a character string holding the source model specification. * `get.modelVars`a function that returns a list with 3 character     vectors, `params`, `state`, and `lhs` of variable names used in the model     specification. These will be output when the model is computed (i.e., the ODE solved by integration).    * `solve`{this function solves (integrates) the ODE. This       is done by passing the code to [rxSolve()].       This is as if you called `rxSolve(rxode2object, ...)`,       but returns a matrix instead of a rxSolve object.        `params`: a numeric named vector with values for every parameter       in the ODE system; the names must correspond to the parameter       identifiers used in the ODE specification;        `events`: an `eventTable` object describing the       input (e.g., doses) to the dynamic system and observation       sampling time points (see  [eventTable()]);        `inits`: a vector of initial values of the state variables       (e.g., amounts in each compartment), and the order in this vector       must be the same as the state variables (e.g., PK/PD compartments);         `stiff`: a logical (`TRUE` by default) indicating whether       the ODE system is stiff or not.        For stiff ODE systems (`stiff = TRUE`), `rxode2` uses       the LSODA (Livermore Solver for Ordinary Differential Equations)       Fortran package, which implements an automatic method switching       for stiff and non-stiff problems along the integration interval,       authored by Hindmarsh and Petzold (2003).        For non-stiff systems (`stiff = FALSE`), `rxode2` uses `DOP853`,       an explicit Runge-Kutta method of order 8(5, 3) of Dormand and Prince       as implemented in C by Hairer and Wanner (1993).        `trans_abs`: a logical (`FALSE` by default) indicating       whether to fit a transit absorption term       (TODO: need further documentation and example);        `atol`: a numeric absolute tolerance (1e-08 by default);        `rtol`: a numeric relative tolerance (1e-06 by default).        The output of \\dQuote{solve} is a matrix with as many rows as there       are sampled time points and as many columns as system variables       (as defined by the ODEs and additional assignments in the rxode2 model           code).}    * `isValid` a function that (naively) checks for model validity,       namely that the C object code reflects the latest model       specification.   * `version` a string with the version of the `rxode2`       object (not the package).   * `dynLoad` a function with one `force = FALSE` argument       that dynamically loads the object code if needed.   * `dynUnload` a function with no argument that unloads       the model object code.   * `delete` removes all created model files, including C and DLL files.       The model object is no longer valid and should be removed, e.g.,       `rm(m1)`.   * `run` deprecated, use `solve`.   * `get.index` deprecated.   * `getObj` internal (not user callable) function."},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Create an ODE-based model specification — rxode2","text":"Rx name rxode2 meant suggest abbreviation Rx medical prescription, thus suggest package emphasis pharmacometrics modeling, including pharmacokinetics (PK), pharmacodynamics (PD), disease progression, drug-disease modeling, etc. ODE-based model specification may coded inside four places: Inside rxode2({}) block statements:      Inside rxode2(\"\") string statement:      file name loaded rxode2:    model function can parsed rxode2:    block statement, character string text file internal rxode2 compilation manager translates ODE system C, compiles loads R session. call rxode2 produces object class rxode2 consists list-like structure (environment) various member functions. last type model (model function), call rxode2 creates parsed rxode2 ui can translated rxode2 compilation model.         type function required nlmixr2 estimation can extended modified model piping. reason focused documentation. basic model specification consists one statements optionally terminated semi-colons ; optional comments (comments delimited # end--line). block statements set statements delimited curly braces, { ... }. Statements can either assignments, conditional /else /else, loops (can exited break), special statements, printing statements (debugging/testing). Assignment statements can : simple assignments, left hand identifier (.e., variable) special time-derivative assignments, left hand specifies change amount corresponding state variable (compartment) respect time e.g., d/dt(depot): special initial-condition assignments left hand specifies compartment initial condition specified, e.g. depot(0) = 0 special model event changes including bioavailability (f(depot)=1), lag time (alag(depot)=0), modeled rate (rate(depot)=2) modeled duration (dur(depot)=2). example model features event specification modeled infusions rxode2 data specification found rxode2 events vignette. special change point syntax, model times. model times specified mtime(var)=time special Jacobian-derivative assignments, left hand specifies change compartment ode respect variable. example, d/dt(y) = dy, Jacobian compartment can specified df(y)/dy(dy) = 1. may advantage obtaining solution specifying Jacobian stiff ODE systems. However, stiff systems tried LSODA, actually slightly slowed solving. Note assignment can done =, <- ~. assigning ~ operator, simple assignments time-derivative assignments output. Note rxode2 model functions assignment ~ can also overloaded residual distribution specification. Special statements can : Compartment declaration statements, can change default dosing compartment assumed compartment number(s) well add extra compartment names end (useful multiple-endpoint nlmixr models); specified cmt(compartmentName) Parameter declaration statements, can make sure input parameters certain order instead ordering parameters order parsed. useful keeping parameter order using 2 different ODE models. specified param(par1, par2,...) example model shown :   Expressions assignment statements can numeric logical. Numeric expressions can include following numeric operators +, -, *, /, ^ mathematical functions defined C R math libraries (e.g., fabs, exp, log, sin, abs). may also access R’s functions R math libraries, like lgammafn log gamma function. rxode2 syntax case-sensitive, .e., ABC different abc, Abc, ABc, etc.","code":"library(rxode2) mod <- rxode2({   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; }) ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_36fd6c4312c660ddf5390102e91a4bb3_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^ mod <- rxode2(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \") ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_4fdd80e5dd5fd33d1951b83e06f7ad84_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^ writeLines(\"   # simple assignment   C2 <- centr/V2    # time-derivative assignment   d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3; \", \"modelFile.rxode2\") mod <- rxode2(filename='modelFile.rxode2') unlink(\"modelFile.rxode2\") mod <- function() {   model({     # simple assignment     C2 <- centr/V2      # time-derivative assignment     d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;   }) }  mod <- rxode2(mod) # or simply mod() if the model is at the end of the function  # These model functions often have residual components and initial # (`ini({})`) conditions attached as well. For example the # theophylline model can be written as:  one.compartment <- function() {   ini({     tka <- 0.45 # Log Ka     tcl <- 1 # Log Cl     tv <- 3.45    # Log V     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) }  # after parsing the model mod <- one.compartment() mod$simulationModel ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_bbee652505e8bddf28c7e688baef12d9_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^  ## rxode2 2.1.2.9000 model named rx_bbee652505e8bddf28c7e688baef12d9 model (ready).  ## x$state: depot, center ## x$stateExtra: cp ## x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp ## x$lhs: ka, cl, v, cp, ipredSim, sim # or  mod$simulationIniModel ## using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ ## In file included from /usr/share/R/include/R.h:71, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2parse.h:33, ##                  from /home/matt/src/rxode2/inst/include/rxode2.h:9, ##                  from /home/matt/R/x86_64-pc-linux-gnu-library/4.3/rxode2parse/include/rxode2_model_shared.h:3, ##                  from rx_fe01e26500a95dd1ba4d8972bc71bc7a_.c:117: ## /usr/share/R/include/R_ext/Complex.h:80:6: warning: ISO C99 doesn’t support unnamed structs/unions [-Wpedantic] ##    80 |     }; ##       |      ^  ## rxode2 2.1.2.9000 model named rx_fe01e26500a95dd1ba4d8972bc71bc7a model (ready).  ## x$state: depot, center ## x$stateExtra: cp ## x$params: tka, tcl, tv, add.sd, eta.ka, eta.cl, eta.v, rxerr.cp ## x$lhs: ka, cl, v, cp, ipredSim, sim # simple assignment    C2 <- centr/V2     # time-derivative assignment    d/dt(centr) <- F*KA*depot - CL*C2 - Q*C2 + Q*C3;"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"identifiers","dir":"Reference","previous_headings":"","what":"Identifiers","title":"Create an ODE-based model specification — rxode2","text":"Like R, Identifiers (variable names) may consist one alphanumeric, underscore _ period . characters, first character digit underscore _. Identifiers model specification can refer : State variables dynamic system (e.g., compartments pharmacokinetics model). Implied input variable, t (time), tlast (last time point), podo (oral dose, undocumented case absorption transit models). Special constants like pi R’s predefined constants. Model parameters (e.g., ka rate absorption, CL clearance, etc.) Others, created assignments part model specification; referred LHS (left-hand side) variable. Currently, rxode2 modeling language recognizes system state variables “parameters”, thus, values need passed R ODE model (e.g., age) either passed params argument integrator function rxSolve() supplied event data-set. certain variable names rxode2 event tables. avoid confusion, following event table-related items assigned, used state can accessed rxode2 code: cmt dvid addl ss rate id However following variables used model specification: evid ii Sometimes rxode2 generates variables fed back rxode2. Similarly, nlmixr2 generates variables used nlmixr estimation simulation. variables start either rx nlmixr prefixes. avoid problems, suggested use variables starting either rx nlmixr prefixes.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"logical-operators","dir":"Reference","previous_headings":"","what":"Logical Operators","title":"Create an ODE-based model specification — rxode2","text":"Logical operators support standard R operators ==, != >= <= > <. Like R can () () statements, ifelse() expressions. Additionally can standard assignment. instance, following valid:   Notice can also use character expressions comparisons. convenience comes cost since character comparisons slower numeric expressions. Unlike R, .numeric .integer logical statements needed, cause syntax error try use function.","code":"cov1 = covm*(sexf == \"female\") + covm*(sexf != \"female\")"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"supported-functions","dir":"Reference","previous_headings":"","what":"Supported functions","title":"Create an ODE-based model specification — rxode2","text":"supported functions rxode2 can seen rxSupportedFuns(). brief description built-functions following table: Note lag(cmt) = equivalent alag(cmt) = = lag(wt)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"reserved-keywords","dir":"Reference","previous_headings":"","what":"Reserved keywords","title":"Create an ODE-based model specification — rxode2","text":"reserved keywords rxode2 model. following table: Note rxFlag always output 11 calc_lhs since final variables calculated, though can tweak test certain parts rxode2 using flag.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"residual-functions-when-using-rxode-functions","dir":"Reference","previous_headings":"","what":"Residual functions when using rxode2 functions","title":"Create an ODE-based model specification — rxode2","text":"addition ~ hiding output certain types output, also used specify residual output endpoint input rxode2 model function (includes residual model({}) block). specifications form:   Indicating variable var variable represents individual central tendencies model also represents compartment specification data-set. can also change compartment name using | syntax, :   case var represents central tendency cmt represents compartment dvid specification.","code":"var ~ add(add.sd) var ~ add(add.sd) | cmt"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"transformations","dir":"Reference","previous_headings":"","what":"Transformations","title":"Create an ODE-based model specification — rxode2","text":"normal related distributions, can apply transformation sides using keywords/functions apply transformations. default likelihood transformations calculated untransformed scale. bounded variables like logit-normal probit-normal low high values defaulted 0 1 missing. models wish proportional model one transformation can replace standard deviation NA allow transformations, lnorm(), probitNorm() logitNorm() can combined variance stabilizing yeoJohnson() transformation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"normal-and-t-related-distributions","dir":"Reference","previous_headings":"","what":"Normal and t-related distributions","title":"Create an ODE-based model specification — rxode2","text":"normal t-related distributions, wanted keep ability use skewed distributions additive proportional t/cauchy-space, distributions specified differently comparison supported distributions within nlmixr2: Note normal t-related distributions nlmixr2 calculate cwres npde normal assumption help assess goodness fit model. Also note +dnorm() mostly testing purposes slow estimation procedure nlmixr2. suggest adding (except explicit testing). multiple endpoint models mix non-normal normal distributions, whole problem shifted log-likelihood method estimation nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"notes-on-additive-proportional-models","dir":"Reference","previous_headings":"","what":"Notes on additive + proportional models","title":"Create an ODE-based model specification — rxode2","text":"two different ways specify additive proportional models, call combined1 combined2, way Monolix calls two distributions (avoid software differences naming). first, combined1, assumes additive proportional differences standard deviation scale, : y=f+(+b* f^c)*err second, combined2, assumes additive proportional differences combined variance scale: y=f+[sqrt(^2+b^2 *f^(2c))]*err default nlmixr2/rxode2 otherwise specified combined2 since mirrors adding 2 normal distributions statistics add variances (standard deviations). However, combined1 can describe data possibly even better combined2 possible options rxode2/nlmixr2.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"distributions-of-known-likelihoods","dir":"Reference","previous_headings":"","what":"Distributions of known likelihoods","title":"Create an ODE-based model specification — rxode2","text":"residuals related normal, t-distribution cauchy, often residual specification form:   compartment specification left handed side specification. generalized likelihood can specify:","code":"cmt ~ dbeta(alpha, beta) ll(cmt) ~ llik specification"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"ordinal-likelihoods","dir":"Reference","previous_headings":"","what":"Ordinal likelihoods","title":"Create an ODE-based model specification — rxode2","text":"Finally, ordinal likelihoods/simulations can specified 2 ways. first :   err represents compartment p0 probability specific category: model ensure sum p values less 1. Additionally can write arbitrary number categories ordinal model described . seems little p0 probability category 1 sometimes scores non-whole numbers. can modeled follows:   numeric categories specified explicitly, probabilities remain :","code":"err ~ c(p0, p1, p2) err ~ c(p0=0, p1=1, p2=2, 3)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"general-table-of-supported-residual-distributions","dir":"Reference","previous_headings":"","what":"General table of supported residual distributions","title":"Create an ODE-based model specification — rxode2","text":"general supported following table (available rxode2::rxResidualError)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"creating-rxode-models","dir":"Reference","previous_headings":"","what":"Creating rxode2 models","title":"Create an ODE-based model specification — rxode2","text":"NA","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Create an ODE-based model specification — rxode2","text":"Chamber, J. M. Temple Lang, D. (2001) Object Oriented Programming R. R News, Vol. 1, . 3, September 2001. https://cran.r-project.org/doc/Rnews/Rnews_2001-3.pdf. Hindmarsh, . C. ODEPACK, Systematized Collection ODE Solvers. Scientific Computing, R. S. Stepleman et al. (Eds.), North-Holland, Amsterdam, 1983, pp. 55-64. Petzold, L. R. Automatic Selection Methods Solving Stiff Nonstiff Systems Ordinary Differential Equations. Siam J. Sci. Stat. Comput. 4 (1983), pp. 136-148. Hairer, E., Norsett, S. P., Wanner, G. Solving ordinary differential equations , nonstiff problems. 2nd edition, Springer Series Computational Mathematics, Springer-Verlag (1993). Plevyak, J. dparser, https://dparser.sourceforge.net/. Web. 12 Oct. 2015.","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Create an ODE-based model specification — rxode2","text":"Melissa Hallow, Wenping Wang Matthew Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxode2.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Create an ODE-based model specification — rxode2","text":"","code":"# \\donttest{  mod <- function() {   ini({     KA   <- .291     CL   <- 18.6     V2   <- 40.2     Q    <- 10.5     V3   <- 297.0     Kin  <- 1.0     Kout <- 1.0     EC50 <- 200.0   })   model({     # A 4-compartment model, 3 PK and a PD (effect) compartment     # (notice state variable names 'depot', 'centr', 'peri', 'eff')     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot;     d/dt(centr) <- KA*depot - CL*C2 - Q*C2 + Q*C3;     d/dt(peri)  <-                    Q*C2 - Q*C3;     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff;     eff(0)      <- 1   }) }  m1 <- rxode2(mod) #>   #>   print(m1) #>  ── rxode2-based free-form 4-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>      KA      CL      V2       Q      V3     Kin    Kout    EC50  #>   0.291  18.600  40.200  10.500 297.000   1.000   1.000 200.000  #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2            centr #> 3                  3             peri #> 4                  4              eff #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         KA <- 0.291 #>         CL <- 18.6 #>         V2 <- 40.2 #>         Q <- 10.5 #>         V3 <- 297 #>         Kin <- 1 #>         Kout <- 1 #>         EC50 <- 200 #>     }) #>     model({ #>         C2 <- centr/V2 #>         C3 <- peri/V3 #>         d/dt(depot) <- -KA * depot #>         d/dt(centr) <- KA * depot - CL * C2 - Q * C2 + Q * C3 #>         d/dt(peri) <- Q * C2 - Q * C3 #>         d/dt(eff) <- Kin - Kout * (1 - C2/(EC50 + C2)) * eff #>         eff(0) <- 1 #>     }) #> }  # Step 2 - Create the model input as an EventTable, # including dosing and observation (sampling) events  # QD (once daily) dosing for 5 days.  qd <- et(amountUnits = \"ug\", timeUnits = \"hours\") %>%   et(amt = 10000, addl = 4, ii = 24)  # Sample the system hourly during the first day, every 8 hours # then after qd <- qd %>% et(0:24) %>%   et(from = 24 + 8, to = 5 * 24, by = 8)  # Step 3 - solve the system  qd.cp <- rxSolve(m1, qd) #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  head(qd.cp) #>    time       C2        C3     depot    centr      peri      eff #> 1 0 [h]  0.00000 0.0000000 10000.000    0.000    0.0000 1.000000 #> 2 1 [h] 43.99334 0.9113641  7475.157 1768.532  270.6751 1.083968 #> 3 2 [h] 54.50866 2.6510696  5587.797 2191.248  787.3677 1.179529 #> 4 3 [h] 51.65163 4.4243597  4176.966 2076.396 1314.0348 1.227523 #> 5 4 [h] 44.37513 5.9432612  3122.347 1783.880 1765.1486 1.233503 #> 6 5 [h] 36.46382 7.1389804  2334.004 1465.845 2120.2772 1.214084  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate ordinal value — rxord","title":"Simulate ordinal value — rxord","text":"Simulate ordinal value","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate ordinal value — rxord","text":"","code":"rxord(...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate ordinal value — rxord","text":"... probabilities simulated.  sum number one.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate ordinal value — rxord","text":"number 1 (number probabilities specified + 1)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate ordinal value — rxord","text":"values entered 'rxord' simulation simulate probability falling group. falls outside specified probabilities, simulate group (number probabilities specified + 1)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate ordinal value — rxord","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxord.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate ordinal value — rxord","text":"","code":"# This will give values 1, and 2 rxord(0.5) #> [1] 2 rxord(0.5) #> [1] 2 rxord(0.5) #> [1] 2 rxord(0.5) #> [1] 2  # This will give values 1, 2 and 3 rxord(0.3, 0.3) #> [1] 3 rxord(0.3, 0.3) #> [1] 3 rxord(0.3, 0.3) #> [1] 1"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate random Poisson variable from threefry generator — rxpois","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"","code":"rxpois(lambda, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"lambda vector (non-negative) means. n number random values return. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"poission random number deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxpois.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate random Poisson variable from threefry generator — rxpois","text":"","code":"# \\donttest{ ## Use threefry engine  rxpois(lambda = 3, n = 10) # with rxpois you have to explicitly state n #>  [1] 3 2 3 3 4 4 2 1 5 3 rxpois(lambda = 3, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 5 5 4 5 3 4 2 4 4 4  rxpois(4) ## The first arguments are the lambda parameter #> [1] 5   ## This example uses `rxpois` directly in the model  rx <- function() {   model({     a <- rxpois(3)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate student t variable from threefry generator — rxt","title":"Simulate student t variable from threefry generator — rxt","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate student t variable from threefry generator — rxt","text":"","code":"rxt(df, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate student t variable from threefry generator — rxt","text":"df degrees freedom (\\(> 0\\), maybe non-integer).  df       = Inf allowed. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate student t variable from threefry generator — rxt","text":"t-distribution random numbers","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate student t variable from threefry generator — rxt","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxt.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate student t variable from threefry generator — rxt","text":"","code":"# \\donttest{  ## Use threefry engine rxt(df = 3, n = 10) # with rxt you have to explicitly state n #>  [1]  0.56020706  0.03692413 -0.14483694 -1.17839233 -0.07091994  0.59518023 #>  [7] -0.86772970 -0.13114717 -0.82988663 -0.44191039 rxt(df = 3, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1]  0.03932487 -0.52306475  0.43368480 -0.34707505  0.95056248 -0.59519017 #>  [7]  0.88227899  2.54392958 -0.24328189 -1.64972949  rxt(4) ## The first argument is the df parameter #> [1] 0.2307802   ## This example uses `rxt` directly in the model  rx <- function() {    model({     a <- rxt(3)    }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate uniform variable from threefry generator — rxunif","title":"Simulate uniform variable from threefry generator — rxunif","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate uniform variable from threefry generator — rxunif","text":"","code":"rxunif(min = 0, max = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate uniform variable from threefry generator — rxunif","text":"min, max lower upper limits distribution.  Must finite. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate uniform variable from threefry generator — rxunif","text":"uniform random numbers","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate uniform variable from threefry generator — rxunif","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxunif.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate uniform variable from threefry generator — rxunif","text":"","code":"# \\donttest{  ## Use threefry engine  rxunif(min = 0, max = 4, n = 10) # with rxunif you have to explicitly state n #>  [1] 2.67123413 1.53680622 0.03693407 3.10817092 1.82090596 0.87096816 #>  [7] 3.00018567 0.74465379 0.44367887 1.04829595 rxunif(min = 0, max = 4, n = 10, ncores = 2) # You can parallelize the simulation using openMP #>  [1] 1.3636154 2.2496600 2.1135472 3.6623704 2.0958273 3.2123092 1.1121945 #>  [8] 3.2064501 0.7037566 3.7979522  rxunif() #> [1] 0.5753584   ## This example uses `rxunif` directly in the model  rx <- function() {   model({     a <- rxunif(0, 3)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate Weibull variable from threefry generator — rxweibull","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"Care taken method encounter birthday problem, described https://www.johndcook.com/blog/2016/01/29/random-number-generator-seed-mistakes/. Since sitmo threefry, currently generates one random deviate uniform distribution seed engine threefry run code.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"","code":"rxweibull(shape, scale = 1, n = 1L, ncores = 1L)"},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"shape, scale shape scale parameters, latter defaulting 1. n number observations. length(n) > 1, length     taken number required. ncores Number cores simulation rxnorm simulates using threefry sitmo generator. rxnormV used simulate vandercorput simulator, since satisfy normal properties changed simple alias rxnorm. longer supported rxode2({}) blocks","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"Weibull random deviates","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"Therefore, simple call random number generated followed second call random number generated may identical seeds.  number random number generator calls increased probability birthday problem increase. key avoid problem either run simulations rxode2 environment (therefore one seed series seeds whole simulation), pre-generate random variables used simulation, seed rxode2 engine rxSetSeed() Internally ID seeded unique number results depend number cores used.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/rxweibull.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate Weibull variable from threefry generator — rxweibull","text":"","code":"# \\donttest{  ## Use threefry engine  # with rxweibull you have to explicitly state n rxweibull(shape = 1, scale = 4, n = 10) #>  [1] 6.44811941 9.31743179 3.71729577 3.93298872 1.93424653 0.50995473 #>  [7] 0.08795812 3.51517749 3.87991413 1.65760268  # You can parallelize the simulation using openMP rxweibull(shape = 1, scale = 4, n = 10, ncores = 2) #>  [1]  8.43975112  0.60066178  3.19996975  8.95933943  0.90401027  1.12725152 #>  [7]  0.02614905  2.35284437  2.74451590 14.14117829  rxweibull(3) #> [1] 1.146769   ## This example uses `rxweibull` directly in the model  rx <- function() {   model({     a <- rxweibull(1, 3)   }) }  et <- et(1, id = 1:2)  s <- rxSolve(rx, et) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’ # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":null,"dir":"Reference","previous_headings":"","what":"Dosing/Amt geom/stat — stat_amt","title":"Dosing/Amt geom/stat — stat_amt","text":"dosing geom shows vertical lines dose occurs","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Dosing/Amt geom/stat — stat_amt","text":"","code":"stat_amt(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   ... )  geom_amt(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Dosing/Amt geom/stat — stat_amt","text":"mapping Set aesthetic mappings created aes(). specified inherit.aes = TRUE (default), combined default mapping top level plot. must supply mapping plot mapping. data data displayed layer. three options: NULL, default, data inherited plot data specified call ggplot(). data.frame, object, override plot data. objects fortified produce data frame. See fortify() variables created. function called single argument, plot data. return value must data.frame, used layer data. function can created formula (e.g. ~ head(.x, 10)). position Position adjustment, either string naming adjustment (e.g. \"jitter\" use position_jitter), result call position adjustment function. Use latter need change settings adjustment. show.legend logical. layer included legends? NA, default, includes aesthetics mapped. FALSE never includes, TRUE always includes. can also named logical vector finely select aesthetics display. inherit.aes FALSE, overrides default aesthetics, rather combining . useful helper functions define data aesthetics inherit behaviour default plot specification, e.g. borders(). ... arguments passed layer(). often aesthetics, used set aesthetic fixed value, like colour = \"red\" size = 3. may also parameters paired geom/stat.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Dosing/Amt geom/stat — stat_amt","text":"returns stat_amt context ggplot2 plot","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Dosing/Amt geom/stat — stat_amt","text":"Requires following aesthetics: x representing x values, usually time amt representing dosing values;  missing zero dose given","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_amt.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Dosing/Amt geom/stat — stat_amt","text":"","code":"# \\donttest{ library(rxode2) library(units) #> udunits database from /usr/share/xml/udunits/udunits2.xml   ## Model from RxODE tutorial mod1 <- function() {   ini({     KA <- 2.94E-01     CL <- 1.86E+01     V2 <- 4.02E+01     Q  <- 1.05E+01     V3 <- 2.97E+02     Kin <- 1     Kout <- 1     EC50 <- 200   })   model({     C2 <- centr/V2     C3 <- peri/V3     d/dt(depot) <- -KA*depot     d/dt(centr) <-  KA*depot - CL*C2 - Q*C2 + Q*C3     d/dt(peri)  <-                    Q*C2 - Q*C3     d/dt(eff)   <- Kin - Kout*(1-C2/(EC50+C2))*eff   }) }  ## These are making the more complex regimens of the rxode2 tutorial  ## bid for 5 days bid <- et(timeUnits=\"hr\") %>%   et(amt=10000,ii=12,until=set_units(5, \"days\"))  ## qd for 5 days qd <- et(timeUnits=\"hr\") %>%       et(amt=20000,ii=24,until=set_units(5, \"days\"))  ## bid for 5 days followed by qd for 5 days  et <- seq(bid,qd) %>% et(seq(0,11*24,length.out=100))  bidQd <- rxSolve(mod1, et, addDosing=TRUE) #>   #>   #>   #>   #> using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’  # by default dotted and under-stated plot(bidQd, C2) + geom_amt(aes(amt=amt))   # of course you can make it a bit more visible  plot(bidQd, C2) + geom_amt(aes(amt=amt), col=\"red\", lty=1, linewidth=1.2)   # }"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":null,"dir":"Reference","previous_headings":"","what":"Censoring geom/stat — stat_cens","title":"Censoring geom/stat — stat_cens","text":"censoring geom shows left right censoring specified nlmixr input data-set fit","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Censoring geom/stat — stat_cens","text":"","code":"stat_cens(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   width = 0.01,   ... )  geom_cens(   mapping = NULL,   data = NULL,   position = \"identity\",   show.legend = NA,   inherit.aes = TRUE,   width = 0.01,   ... )"},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Censoring geom/stat — stat_cens","text":"mapping Set aesthetic mappings created aes(). specified inherit.aes = TRUE (default), combined default mapping top level plot. must supply mapping plot mapping. data data displayed layer. three options: NULL, default, data inherited plot data specified call ggplot(). data.frame, object, override plot data. objects fortified produce data frame. See fortify() variables created. function called single argument, plot data. return value must data.frame, used layer data. function can created formula (e.g. ~ head(.x, 10)). position Position adjustment, either string naming adjustment (e.g. \"jitter\" use position_jitter), result call position adjustment function. Use latter need change settings adjustment. show.legend logical. layer included legends? NA, default, includes aesthetics mapped. FALSE never includes, TRUE always includes. can also named logical vector finely select aesthetics display. inherit.aes FALSE, overrides default aesthetics, rather combining . useful helper functions define data aesthetics inherit behaviour default plot specification, e.g. borders(). width represents width (\\ censoring box ... arguments passed layer(). often aesthetics, used set aesthetic fixed value, like colour = \"red\" size = 3. may also parameters paired geom/stat.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Censoring geom/stat — stat_cens","text":"returns ggplot2 stat","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/stat_cens.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Censoring geom/stat — stat_cens","text":"Requires following aesthetics: x Represents independent variable, often time scale y represents dependent variable CENS censoring information; (-1 right censored, 0 censoring 1 left censoring) LIMIT represents corresponding limit () add boxes representing areas fit censored.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":null,"dir":"Reference","previous_headings":"","what":"Summary of rxDll object — summary.rxDll","title":"Summary of rxDll object — summary.rxDll","text":"gives expanded information rxDll object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Summary of rxDll object — summary.rxDll","text":"","code":"# S3 method for rxDll summary(object, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Summary of rxDll object — summary.rxDll","text":"object RxDll object ... arguments.  Includes noprint, logical telling object print rxDll object first. default FALSE","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Summary of rxDll object — summary.rxDll","text":"object returned","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxDll.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Summary of rxDll object — summary.rxDll","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":null,"dir":"Reference","previous_headings":"","what":"Print expanded information about the rxode2 object. — summary.rxode2","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"prints expanded information rxode2 object.","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"","code":"# S3 method for rxode2 summary(object, ...)"},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"object rxode2 object ... Ignored parameters","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"object returned","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/summary.rxode2.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Print expanded information about the rxode2 object. — summary.rxode2","text":"Matthew L.Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":null,"dir":"Reference","previous_headings":"","what":"Update for rxUi — update.rxUi","title":"Update for rxUi — update.rxUi","text":"Update rxUi","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update for rxUi — update.rxUi","text":"","code":"# S3 method for rxUi update(object, ..., envir = parent.frame())"},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update for rxUi — update.rxUi","text":"object rxode2 UI object ... Lines update envir Environment evaluating ini() style calls","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/update.rxUi.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update for rxUi — update.rxUi","text":"new rxode2 updated UI object","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":null,"dir":"Reference","previous_headings":"","what":"uppergamma: upper incomplete gamma function — uppergamma","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"tgamma boost library","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"","code":"uppergamma(a, z)"},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"numeric '' parameter upper incomplete gamma z numeric 'z' parameter upper incomplete gamma","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"uppergamma results","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"uppergamma function given : \\(uppergamma(, z) = \\int_{z}^{\\infty}t^{-1}\\cdot e^{-t} dt\\)","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"Matthew L. Fidler","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/uppergamma.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"uppergamma: upper incomplete gamma function — uppergamma","text":"","code":"uppergamma(1, 3) #> [1] 0.04978707  uppergamma(1:3, 3) #> [1] 0.04978707 0.19914827 0.84638016  uppergamma(1, 1:3) #> [1] 0.36787944 0.13533528 0.04978707"},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":null,"dir":"Reference","previous_headings":"","what":"Set random effects and residual error to zero — zeroRe","title":"Set random effects and residual error to zero — zeroRe","text":"Set random effects residual error zero","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set random effects and residual error to zero — zeroRe","text":"","code":"zeroRe(object, which = c(\"omega\", \"sigma\"), fix = TRUE)"},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set random effects and residual error to zero — zeroRe","text":"object model modify types parameters set zero fix parameters fixed zero value?","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Set random effects and residual error to zero — zeroRe","text":"object parameters set zero","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Set random effects and residual error to zero — zeroRe","text":"Bill Denney","code":""},{"path":"https://nlmixr2.github.io/rxode2/reference/zeroRe.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Set random effects and residual error to zero — zeroRe","text":"","code":"one.compartment <- function() {   ini({     tka <- log(1.57); label(\"Ka\")     tcl <- log(2.72); label(\"Cl\")     tv <- log(31.5); label(\"V\")     eta.ka ~ 0.6     eta.cl ~ 0.3     eta.v ~ 0.1     add.sd <- 0.7   })   model({     ka <- exp(tka + eta.ka)     cl <- exp(tcl + eta.cl)     v <- exp(tv + eta.v)     d/dt(depot) = -ka * depot     d/dt(center) = ka * depot - cl / v * center     cp = center / v     cp ~ add(add.sd)   }) } zeroRe(one.compartment) #>   #>   #>  ── rxode2-based free-form 2-cmt ODE model ──────────────────────────────────────  #>  ── Initalization: ──   #> Fixed Effects ($theta):  #>       tka       tcl        tv    add.sd  #> 0.4510756 1.0006319 3.4499875 0.0000000  #>  #> Omega ($omega):  #>        eta.ka eta.cl eta.v #> eta.ka      0      0     0 #> eta.cl      0      0     0 #> eta.v       0      0     0 #> attr(,\"lotriFix\") #>        eta.ka eta.cl eta.v #> eta.ka   TRUE  FALSE FALSE #> eta.cl  FALSE   TRUE FALSE #> eta.v   FALSE  FALSE  TRUE #>  #> States ($state or $stateDf):  #>   Compartment Number Compartment Name #> 1                  1            depot #> 2                  2           center #>  ── μ-referencing ($muRefTable): ──   #>   theta    eta level #> 1   tka eta.ka    id #> 2   tcl eta.cl    id #> 3    tv  eta.v    id #>  #>  ── Model (Normalized Syntax): ──  #> function() { #>     ini({ #>         tka <- 0.451075619360217 #>         label(\"Ka\") #>         tcl <- 1.00063188030791 #>         label(\"Cl\") #>         tv <- 3.44998754583159 #>         label(\"V\") #>         add.sd <- fix(0, 0) #>         eta.ka ~ fix(0) #>         eta.cl ~ fix(0) #>         eta.v ~ fix(0) #>     }) #>     model({ #>         ka <- exp(tka + eta.ka) #>         cl <- exp(tcl + eta.cl) #>         v <- exp(tv + eta.v) #>         d/dt(depot) = -ka * depot #>         d/dt(center) = ka * depot - cl/v * center #>         cp = center/v #>         cp ~ add(add.sd) #>     }) #> }"},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-1-3","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.1.3","text":"Make sure object uncompressed rxode2 ui solving rxSolve (See #661) Fix #670 using last simulated observation residual trailing doses.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-1-3","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.1.3","text":"Create function see rxode2 solve loaded memory (rxode2::rxSolveSetup()) Create new function fixes rxode2 population values model (drops initial estimates); rxFixPop()","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"other-changes-2-1-3","dir":"Changelog","previous_headings":"","what":"Other changes","title":"rxode2 2.1.3","text":"Pendantic -remap (requested CRAN)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-212","dir":"Changelog","previous_headings":"","what":"rxode2 2.1.2","title":"rxode2 2.1.2","text":"CRAN release: 2024-01-30","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"other-changes-2-1-2","dir":"Changelog","previous_headings":"","what":"Other changes","title":"rxode2 2.1.2","text":"rxUi compression now defaults fast compression Fixes String literal formatting issues identified CRAN (#643) Removes linear compartment solutions gradients intel c++ compiler (since crash compiler).","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-210","dir":"Changelog","previous_headings":"","what":"rxode2 2.1.0","title":"rxode2 2.1.0","text":"CRAN release: 2023-12-11","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"breaking-changes-2-1-0","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"rxode2 2.1.0","text":"Steady state lag times longer shifted lag time solved steady state default. addition steady state original time dosing also back-calculated. want old behavior can bring back option ssAtDoseTime=FALSE. “dop853” now uses hmax/h0 values rxControl() rxSolve(). may change ODE solving using “dop853” specified (xgxr available), x axis longer assumed hours","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-1-0","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.1.0","text":"User defined functions can now R functions. many R functions can converted C rxFun() (can see C code afterwards rxC(\"funName\")) Parallel solving models require sorting (like modeled lag times, modeled duration etc) now solve parallel instead downgrading single threaded solving Steady state infusions duration infusions greater inter-dose interval now supported. Added $symengineModelNoPrune $symengineModelPrune loading models rxode2 rxS() plotting creating confidence intervals multiple endpoint models simulated rxode2 ui model, can plot/summarize endpoint sim. (ie. confint(model,   \"sim\") plot(model, sim)). want summarize subset endpoints, can focus endpoint pre-pending endpoint sim. example wanted plot/summarize endpoint eff use sim.eff. (ie confint(model, \"sim.eff\") plot(model, sim.eff)) Added model$simulationIniModel prepend initial conditions ini({}) block classic rxode2({}) model. Now model$simulationModel model$simulationIniModel save use initialization values compiled model, solve original ui model. Allow ini(model) <- NULL drop ini block .ini(NULL) gives ini({}) (Issue #523) Add function modelExtract() extract model lines allow modifying changing model piping simply assigning modified lines model(ui) <- newModifiedLines Add Algebraic mu-referencing detection (mu2) allows express mu-referenced covariates : Instead previously required (log.WT.div.70.5 calculated data) mu expressions. ui now information allow transformation data internally transformation old mu-referencing style run optimization. Allow steady state infusions duration infusion greater inter-dose interval solved. Solves now possibly print information issuing “solve system” error function rxSetPipingAuto() now exported change way affect piping individual setup Allow covariates specified model piping, mod   %>% model(=var+3, cov=\"var\") add \"var\" covariate. calculating confidence intervals rxode2 simulated objects can now use stratify simulation summary. example can now stratify gender race : confint(sim,   \"sim\", =c(\"race\", \"gender\")) calculating intervals rxode2 simulated objects can now use ci=FALSE calculates default intervals without bands percentiles; can also choose match secondary bands limits levels use ci=0.99 instance new function introduced meanProbs() calculates mean expected confidence bands either normal t distribution related new function introduced calculates mean confidence bands Bernoulli/Binomial distribution (binomProbs()) calculating intervals rxode2 simulated objects can also use mean=TRUE use mean first level confidence using meanProbs(). confidence interval can override n used confidence interval using n=#. can also change prediction interval instead using pred=TRUE. Also calculating intervals rxode2 simulated object can also use mean=\"binom\" use binomial distributional information (ci) first level confidence using binomProbs(). confidence interval can override n used confidence interval using n=#. can also change prediction interval instead using pred=TRUE. pred=TRUE can override number predicted samples m=# plotting confint derived intervals rxode2 simulation, can now subset based simulated value like plot(ci, Cc) plot variable Cc summarized even also summarized eff (instance). rxode2 ui compressed ui object, can modify ini block $ini <- modify model block $model <-. equivalent ini(model) <- model(model) <-, respectively. Otherwise, object added user defined components function (ie $meta). object uncompressed, simply assigns environment instead (just like ). printing meta information happens lotri compatible matrix, use lotri express instead default R expression. Allow character vectors converted expressions piping (#552) rxAppendModel() now take arbitrary number models append together; also better handling models duplicate parameters models without ini() blocks (#617 / #573 / #575). keep now also keep attributes input data (special handling levels); means broader variety classes kept carrying information (example ordered factors, data frame columns unit information, etc) Piping arguments append ini() model() aligned perform similarly. Therefore ini(append=) now can take expressions instead simply strings model(append=) can also take strings. Also model piping now can specify integer line number modified just like ini() . Also model(append=FALSE) changed model(append=NULL). behavior don’t specify argument, behavior changed align ini() piping. Hence model(append=TRUE) append model(append=FALSE) now pre-pend model. model(append=NULL) modify lines like behavior ini(append=NULL). default model(line) modifying line -place still applies. breaking change, code perform . Labels can now dropped ini(param=label(NULL)). Also parameters can dropped idiom model(param=NULL) ini(param=NULL) changes parameter covariate align idiom dropping parameters rxRename refactored run faster","code":"cl <- exp(tcl + eta.cl + wt_cl * log(WT/70.5)) cl <- exp(tcl + eta.cl + wt_cl * log.WT.div.70.5)"},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"internal-new-features-2-1-0","dir":"Changelog","previous_headings":"","what":"Internal new features","title":"rxode2 2.1.0","text":"Add .model() list expressions, implies model(ui) <-   ui$lstExpr assign model components. also robustly work character vectors Simulated objects rxSolve now can access model variables $rxModelVars Simulation models UI now use rxerr.endpoint instead err.endpoint sigma residual error. align convention internally generated variables start rx nlmixr Sorting uses timsort now, upgraded latest version Morwenn","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-1-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.1.0","text":"Simulating/solving functions/ui now prefers params omega sigma model (#632) Piping add constants initial estimates constants specified model({}) block (like k <- 1), ini block Bug fix geom_amt() aes transformation x Bug fix covariate updates may affect multiple compartment models (like issue #581)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"maintenance-fixes-2-1-0","dir":"Changelog","previous_headings":"","what":"Maintenance fixes","title":"rxode2 2.1.0","text":"Modify plot code work development xgxr","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2014","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.14","title":"rxode2 2.0.14","text":"CRAN release: 2023-10-07 CRAN requested FORTRAN kind changed portable; commented code, simply removed comment. Bug-fix geom_amt(); also now uses linewidth least ggplot2 3.4.0 documentation cleaned rxode2 2.0.13","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2013","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.13","title":"rxode2 2.0.13","text":"CRAN release: 2023-04-22","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-0-13","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.0.13","text":"bug fixed zeroRe() function works correlated omega values. bug fixed rename() function works initial conditions compartments (cmt(0))","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-0-13","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.0.13","text":"new function zeroRe() allows simple setting omega /sigma values zero model (#456) Diagonal zeros omega sigma matrices treated zeros model. corresponding omega sigma matrices drop columns/rows diagonals zero create new omega sigma matrix simulation. idiom NONMEM uses simulation matrices. Add ability pipe model estimates another model parentModel %>% ini(modelWithNewEsts) Add ability append model statements piping using %>%   model(x=3, append=d/dt(depot)), still supports appending append=TRUE pre-pending append=NA (default replace lines append=FALSE) rxSolve’s keep argument now maintain character factor classes input data class (#190) Parameter labels may now modified via ini(param = label(\"text\")) (#351). Parameter order may modified via append argument ini() piping model. example, ini(param = 1, append = 0) ini(param = label(\"text\"), append = \"param2\") (#352).","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"internal-changes-2-0-13","dir":"Changelog","previous_headings":"","what":"Internal changes","title":"rxode2 2.0.13","text":"lower/upper bounds outside required bounds, adjustment displayed. initial values piped break model’s boundary condition reset boundary unbounded message boundary reset. Added .rxUi() function convert following objects rxUi objects: rxode2, rxModelVars, function. Converting nlmixr2 fits rxUi placed s3 method corresponding package. assertRxUi(x) now uses .rxUi() can extended outside rxode2/nlmixr2. rxode2 now supports addl ss doses Moved rxDerived rxode2parse (re-exported ). Added test transit compartment solving absence dosing transit compartment (fixed rxode2parse solving tested ) Using ini() without arguments rxode2 type function return ini() block. Also added method ini(mod) <-   iniBlock modify ini block wish. iniBlock expression. Using model() without arguments rxode2 type function return model() block. Also added new method model(mod) <- modelBlock Added new method rxode2(mod) <- modFunction allows replacing function new function maintaining meta information ui (like information comes nonmem2rx models). modFunction body new function, new function, new rxode2 ui. rxode2 ui objects now $sticky item inside internal (compressed) environment. $sticky tells variables keep “significant” change ui piping sort model change. respected model piping, modifying model ini(mod)<-, model(mod)<-, rxode2(mod)<-. significant change change model block, change number estimates, change value estimates. Estimate bounds, weather estimate fixed estimate label changes considered significant. Added .ini() method convert various formats ini expression. used internally ini(mod)<-. want assign something new can convert ini expression, add method .ini(). Added .model() method convert various formats model expression. used internally model(mod)<-. want assign something new can convert model expression, add method .model().","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2011","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.11","title":"rxode2 2.0.11","text":"CRAN release: 2022-11-01 Give meaningful error ‘rxode2’ ui models error expressions Break ABI requirement roxde2() rxode2parse() new rxode2parse fix sprintf exclusion shown CRAN.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-2010","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.10","title":"rxode2 2.0.10","text":"CRAN release: 2022-10-20 Time invariant covariates can now contain ‘NA’ values. column ‘NA’ entire id, now ‘rxode2’ warns id column instead just id. fix CRAN issues ‘nlmixr2est’, make version dependency explicit.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-209","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.9","title":"rxode2 2.0.9","text":"CRAN release: 2022-10-19 Remove log likelihoods ‘rxode2’ reduce compilation time increase maintainability ‘rxode2’. transferred ‘rxode2ll’ (requested CRAN). Remove parsing ‘rxode2’ solved linear compartment code move ‘rxode2parse’ reduce compilation time (requested CRAN). Remove random number generation ‘rxode2’ move ‘rxode2random’ reduce compilation time (requested CRAN). Remove event table translation generation ‘rxode2’ move ‘rxode2et’ reduce compilation time (requested CRAN). Change rxode2 ui object compressed, serialized object default. reduce C stack size problem occurs many environments R. Warn ignoring items simulations Export method change ‘rxode2’ solve methods internal integers Bug fix time invariant covariates identified time variant covariate individual’s time starts 0.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-208","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.8","title":"rxode2 2.0.8","text":"CRAN release: 2022-09-23","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"breaking-changes-2-0-8","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"rxode2 2.0.8","text":"rxgamma now allows rate input. aligns internal rxode2 version rxgamma clarifies used. also aligned llikGamma function used generalized likelihood estimation. ui cauchy simulations now follow ui normal t distributions, means can combine transformations. cauchy t distribution one degree freedom. ui dnorm() norm() longer equivalent add(). Now allows use loglik llikNorm() instead standard nlmixr2 style focei likelihood. done adding dnorm() end line. also means dnorm() now doesn’t take arguments. Vandercorput normal removed (non-random number generator)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-0-8","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.0.8","text":"Allow models nlmixr2 form without ini({}) block Allow model piping omega matrix f %>% ini(omegaMatrix) Standard models created rxode2() can piped model function Families log-likelihood added rxode2 mixed likelihood nonlinear mixed effects models may specified run. memory footprint rxode2 solving reduced Piping now allow named strings (issue #249)","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-0-8","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.0.8","text":"rxode2’s symengine convert sqrt(2) M_SQRT_2 M_SQRT2. fixed; noticeable nlmixr2 log-likelihood estimation methods rxode2 treats DV non-covariate etTran (last time duplicate model). noticeable nlmixr2 log-likelihood estimation methods.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"new-features-2-0-8-1","dir":"Changelog","previous_headings":"","what":"New features","title":"rxode2 2.0.8","text":"new flag (rxFlag) created tell rxode2 solving process . useful debugging. outputting variable always 11 calculating left handed equations. using conjunction printf() methods, double variable formatted \"%f\". additional option fullPrint added rxode2() allows rprintf() used almost rxode2() steps (inductive linearization matrix exponential exception ) instead just integration ddt step. defaults FALSE.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-207","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.7","title":"rxode2 2.0.7","text":"CRAN release: 2022-05-17 Removed accidental ^S news requested CRAN. Bug fix complicated mu-referencing. Change rxode2 md5 depend C/C++/Fortran code headers R files. way binary compatibility nlmixr2est rxode2, new version nlmixr2est need submitted CRAN.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"rxode2-206","dir":"Changelog","previous_headings":"","what":"rxode2 2.0.6","title":"rxode2 2.0.6","text":"CRAN release: 2022-05-09","code":""},{"path":[]},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"solving-controls-2-0-6","dir":"Changelog","previous_headings":"Breaking changes","what":"Solving controls","title":"rxode2 2.0.6","text":"options rxControl rxSolve strict. camelCase now always used. Old options like add.cov transit_abs longer supported, addCov supported. new option, sigdig added rxControl(), controls common significant figure options like atol, rtol, ssAtol, ssRtol, single option.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"simulations-2-0-6","dir":"Changelog","previous_headings":"Breaking changes","what":"Simulations","title":"rxode2 2.0.6","text":"simulations, $simulationSigma now assumes diagonal matrix. sigma values assumed standard normal, uncorrelated endpoints. Simulation uncertainty still draw identity diagonal matrix Parallel solving now seeds simulation per individual based initial seed plus simulation id. makes simulation reproducible regardless number cores running simulation.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"other-breaking-changes-2-0-6","dir":"Changelog","previous_headings":"Breaking changes","what":"Other breaking changes","title":"rxode2 2.0.6","text":"Solved objects now access underlying rxode model $rxode2 instead $rxode Since change names, rxode2, rxode RxODE perform function. Options changed RxODE.syntax rxode2.syntax. Assigning states rxode2.syntax.assign.state (RxODE.syntax.assign.state) longer supported. Enforcing “pure” assignment syntax = syntax longer supported rxode2.syntax.assign longer supported (RxODE.syntax.assign). Since R supports ** exponentiation operator, pure syntax without ** can longer enabled. Hence rxode2.syntax.star.pow (RxODE.syntax.star.pow) longer effect. “pure” syntax requires semicolon can longer enabled. Therefore rxode2.syntax.require.semicolon (RxODE.syntax.require.semicolon) longer effect. syntax state(0) can longer turned . rxode2.syntax.allow.ini0 (RxODE.syntax.allow.ini0) removed. Variable dots variable state names like state.name works R. Therefore, “pure” syntax excluding . values variables enforced rxode2.syntax.allow.dots (RxODE.syntax.allow.dots). mnemonic et(rate=model) et(dur=model) mnemonics removed. rate needs set -1 -2 manually instead. function rxode2Test() removed favor using testthat directly. Transit compartments need use new evid, evid=7. said, transitAbs option longer supported. ID columns input parameter data frames sorted merged original dataset ; underlying assumption ID order now checked outside rxode2(). Note event data frame still sorted.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"additional-features-2-0-6","dir":"Changelog","previous_headings":"","what":"Additional features","title":"rxode2 2.0.6","text":"UI functions nlmixr ported work rxode2 directly. rxModelVars({}) now supported. may now combine 2 models rxode2 rxAppendModel(). fact, long first value rxode2 evaluated ui model, can use c/rbind bind 2 models together. may now append model lines piping using %>% model(lines,   append=TRUE) can also pre-pend lines %>% model(lines,   append=NA) may now rename model variables, states defined parameters %>% rxRename(new=old) dplyr loaded: %>%   rename(new=old) can fix parameters %>% ini(tcl=fix) %>% ini(fix(tcl)) well unfix parameters %>% ini(tcl=unfix) %>% ini(unfix(tcl))","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"internal-changes-2-0-6","dir":"Changelog","previous_headings":"","what":"Internal changes","title":"rxode2 2.0.6","text":"Strict R headers enforced places Since many changes incompatible, version renamed rxode2 rxode2() printout longer uses rules centered headings make display better larger variety systems.","code":""},{"path":"https://nlmixr2.github.io/rxode2/news/index.html","id":"bug-fixes-2-0-6","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"rxode2 2.0.6","text":"tad() related time features reset start infusion (opposed starting beginning end infusion)","code":""}]