-
Notifications
You must be signed in to change notification settings - Fork 14
/
Copy pathnotebook.tex
1604 lines (1254 loc) · 76.6 KB
/
notebook.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
% Default to the notebook output style
% Inherit from the specified cell style.
\documentclass[11pt]{article}
\usepackage[T1]{fontenc}
% Nicer default font (+ math font) than Computer Modern for most use cases
\usepackage{mathpazo}
% Basic figure setup, for now with no caption control since it's done
% automatically by Pandoc (which extracts  syntax from Markdown).
\usepackage{graphicx}
% We will generate all images so they have a width \maxwidth. This means
% that they will get their normal width if they fit onto the page, but
% are scaled down if they would overflow the margins.
\makeatletter
\def\maxwidth{\ifdim\Gin@nat@width>\linewidth\linewidth
\else\Gin@nat@width\fi}
\makeatother
\let\Oldincludegraphics\includegraphics
% Set max figure width to be 80% of text width, for now hardcoded.
\renewcommand{\includegraphics}[1]{\Oldincludegraphics[width=.8\maxwidth]{#1}}
% Ensure that by default, figures have no caption (until we provide a
% proper Figure object with a Caption API and a way to capture that
% in the conversion process - todo).
\usepackage{caption}
\DeclareCaptionLabelFormat{nolabel}{}
\captionsetup{labelformat=nolabel}
\usepackage{adjustbox} % Used to constrain images to a maximum size
\usepackage{xcolor} % Allow colors to be defined
\usepackage{enumerate} % Needed for markdown enumerations to work
\usepackage{geometry} % Used to adjust the document margins
\usepackage{amsmath} % Equations
\usepackage{amssymb} % Equations
\usepackage{textcomp} % defines textquotesingle
% Hack from http://tex.stackexchange.com/a/47451/13684:
\AtBeginDocument{%
\def\PYZsq{\textquotesingle}% Upright quotes in Pygmentized code
}
\usepackage{upquote} % Upright quotes for verbatim code
\usepackage{eurosym} % defines \euro
\usepackage[mathletters]{ucs} % Extended unicode (utf-8) support
\usepackage[utf8x]{inputenc} % Allow utf-8 characters in the tex document
\usepackage{fancyvrb} % verbatim replacement that allows latex
\usepackage{grffile} % extends the file name processing of package graphics
% to support a larger range
% The hyperref package gives us a pdf with properly built
% internal navigation ('pdf bookmarks' for the table of contents,
% internal cross-reference links, web links for URLs, etc.)
\usepackage{hyperref}
\usepackage{longtable} % longtable support required by pandoc >1.10
\usepackage{booktabs} % table support for pandoc > 1.12.2
\usepackage[inline]{enumitem} % IRkernel/repr support (it uses the enumerate* environment)
\usepackage[normalem]{ulem} % ulem is needed to support strikethroughs (\sout)
% normalem makes italics be italics, not underlines
% Colors for the hyperref package
\definecolor{urlcolor}{rgb}{0,.145,.698}
\definecolor{linkcolor}{rgb}{.71,0.21,0.01}
\definecolor{citecolor}{rgb}{.12,.54,.11}
% ANSI colors
\definecolor{ansi-black}{HTML}{3E424D}
\definecolor{ansi-black-intense}{HTML}{282C36}
\definecolor{ansi-red}{HTML}{E75C58}
\definecolor{ansi-red-intense}{HTML}{B22B31}
\definecolor{ansi-green}{HTML}{00A250}
\definecolor{ansi-green-intense}{HTML}{007427}
\definecolor{ansi-yellow}{HTML}{DDB62B}
\definecolor{ansi-yellow-intense}{HTML}{B27D12}
\definecolor{ansi-blue}{HTML}{208FFB}
\definecolor{ansi-blue-intense}{HTML}{0065CA}
\definecolor{ansi-magenta}{HTML}{D160C4}
\definecolor{ansi-magenta-intense}{HTML}{A03196}
\definecolor{ansi-cyan}{HTML}{60C6C8}
\definecolor{ansi-cyan-intense}{HTML}{258F8F}
\definecolor{ansi-white}{HTML}{C5C1B4}
\definecolor{ansi-white-intense}{HTML}{A1A6B2}
% commands and environments needed by pandoc snippets
% extracted from the output of `pandoc -s`
\providecommand{\tightlist}{%
\setlength{\itemsep}{0pt}\setlength{\parskip}{0pt}}
\DefineVerbatimEnvironment{Highlighting}{Verbatim}{commandchars=\\\{\}}
% Add ',fontsize=\small' for more characters per line
\newenvironment{Shaded}{}{}
\newcommand{\KeywordTok}[1]{\textcolor[rgb]{0.00,0.44,0.13}{\textbf{{#1}}}}
\newcommand{\DataTypeTok}[1]{\textcolor[rgb]{0.56,0.13,0.00}{{#1}}}
\newcommand{\DecValTok}[1]{\textcolor[rgb]{0.25,0.63,0.44}{{#1}}}
\newcommand{\BaseNTok}[1]{\textcolor[rgb]{0.25,0.63,0.44}{{#1}}}
\newcommand{\FloatTok}[1]{\textcolor[rgb]{0.25,0.63,0.44}{{#1}}}
\newcommand{\CharTok}[1]{\textcolor[rgb]{0.25,0.44,0.63}{{#1}}}
\newcommand{\StringTok}[1]{\textcolor[rgb]{0.25,0.44,0.63}{{#1}}}
\newcommand{\CommentTok}[1]{\textcolor[rgb]{0.38,0.63,0.69}{\textit{{#1}}}}
\newcommand{\OtherTok}[1]{\textcolor[rgb]{0.00,0.44,0.13}{{#1}}}
\newcommand{\AlertTok}[1]{\textcolor[rgb]{1.00,0.00,0.00}{\textbf{{#1}}}}
\newcommand{\FunctionTok}[1]{\textcolor[rgb]{0.02,0.16,0.49}{{#1}}}
\newcommand{\RegionMarkerTok}[1]{{#1}}
\newcommand{\ErrorTok}[1]{\textcolor[rgb]{1.00,0.00,0.00}{\textbf{{#1}}}}
\newcommand{\NormalTok}[1]{{#1}}
% Additional commands for more recent versions of Pandoc
\newcommand{\ConstantTok}[1]{\textcolor[rgb]{0.53,0.00,0.00}{{#1}}}
\newcommand{\SpecialCharTok}[1]{\textcolor[rgb]{0.25,0.44,0.63}{{#1}}}
\newcommand{\VerbatimStringTok}[1]{\textcolor[rgb]{0.25,0.44,0.63}{{#1}}}
\newcommand{\SpecialStringTok}[1]{\textcolor[rgb]{0.73,0.40,0.53}{{#1}}}
\newcommand{\ImportTok}[1]{{#1}}
\newcommand{\DocumentationTok}[1]{\textcolor[rgb]{0.73,0.13,0.13}{\textit{{#1}}}}
\newcommand{\AnnotationTok}[1]{\textcolor[rgb]{0.38,0.63,0.69}{\textbf{\textit{{#1}}}}}
\newcommand{\CommentVarTok}[1]{\textcolor[rgb]{0.38,0.63,0.69}{\textbf{\textit{{#1}}}}}
\newcommand{\VariableTok}[1]{\textcolor[rgb]{0.10,0.09,0.49}{{#1}}}
\newcommand{\ControlFlowTok}[1]{\textcolor[rgb]{0.00,0.44,0.13}{\textbf{{#1}}}}
\newcommand{\OperatorTok}[1]{\textcolor[rgb]{0.40,0.40,0.40}{{#1}}}
\newcommand{\BuiltInTok}[1]{{#1}}
\newcommand{\ExtensionTok}[1]{{#1}}
\newcommand{\PreprocessorTok}[1]{\textcolor[rgb]{0.74,0.48,0.00}{{#1}}}
\newcommand{\AttributeTok}[1]{\textcolor[rgb]{0.49,0.56,0.16}{{#1}}}
\newcommand{\InformationTok}[1]{\textcolor[rgb]{0.38,0.63,0.69}{\textbf{\textit{{#1}}}}}
\newcommand{\WarningTok}[1]{\textcolor[rgb]{0.38,0.63,0.69}{\textbf{\textit{{#1}}}}}
% Define a nice break command that doesn't care if a line doesn't already
% exist.
\def\br{\hspace*{\fill} \\* }
% Math Jax compatability definitions
\def\gt{>}
\def\lt{<}
% Document parameters
\title{Predicting Hard Drive Failure - A Juul Labs Case Study}
% Pygments definitions
\makeatletter
\def\PY@reset{\let\PY@it=\relax \let\PY@bf=\relax%
\let\PY@ul=\relax \let\PY@tc=\relax%
\let\PY@bc=\relax \let\PY@ff=\relax}
\def\PY@tok#1{\csname PY@tok@#1\endcsname}
\def\PY@toks#1+{\ifx\relax#1\empty\else%
\PY@tok{#1}\expandafter\PY@toks\fi}
\def\PY@do#1{\PY@bc{\PY@tc{\PY@ul{%
\PY@it{\PY@bf{\PY@ff{#1}}}}}}}
\def\PY#1#2{\PY@reset\PY@toks#1+\relax+\PY@do{#2}}
\expandafter\def\csname PY@tok@w\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.73,0.73}{##1}}}
\expandafter\def\csname PY@tok@c\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.25,0.50,0.50}{##1}}}
\expandafter\def\csname PY@tok@cp\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.74,0.48,0.00}{##1}}}
\expandafter\def\csname PY@tok@k\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@kp\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@kt\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.69,0.00,0.25}{##1}}}
\expandafter\def\csname PY@tok@o\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@ow\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.67,0.13,1.00}{##1}}}
\expandafter\def\csname PY@tok@nb\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@nf\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.00,1.00}{##1}}}
\expandafter\def\csname PY@tok@nc\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.00,1.00}{##1}}}
\expandafter\def\csname PY@tok@nn\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.00,1.00}{##1}}}
\expandafter\def\csname PY@tok@ne\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.82,0.25,0.23}{##1}}}
\expandafter\def\csname PY@tok@nv\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.10,0.09,0.49}{##1}}}
\expandafter\def\csname PY@tok@no\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.53,0.00,0.00}{##1}}}
\expandafter\def\csname PY@tok@nl\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.63,0.63,0.00}{##1}}}
\expandafter\def\csname PY@tok@ni\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.60,0.60,0.60}{##1}}}
\expandafter\def\csname PY@tok@na\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.49,0.56,0.16}{##1}}}
\expandafter\def\csname PY@tok@nt\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@nd\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.67,0.13,1.00}{##1}}}
\expandafter\def\csname PY@tok@s\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@sd\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@si\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.73,0.40,0.53}{##1}}}
\expandafter\def\csname PY@tok@se\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.73,0.40,0.13}{##1}}}
\expandafter\def\csname PY@tok@sr\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.40,0.53}{##1}}}
\expandafter\def\csname PY@tok@ss\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.10,0.09,0.49}{##1}}}
\expandafter\def\csname PY@tok@sx\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@m\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@gh\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.00,0.50}{##1}}}
\expandafter\def\csname PY@tok@gu\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.50,0.00,0.50}{##1}}}
\expandafter\def\csname PY@tok@gd\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.63,0.00,0.00}{##1}}}
\expandafter\def\csname PY@tok@gi\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.63,0.00}{##1}}}
\expandafter\def\csname PY@tok@gr\endcsname{\def\PY@tc##1{\textcolor[rgb]{1.00,0.00,0.00}{##1}}}
\expandafter\def\csname PY@tok@ge\endcsname{\let\PY@it=\textit}
\expandafter\def\csname PY@tok@gs\endcsname{\let\PY@bf=\textbf}
\expandafter\def\csname PY@tok@gp\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.00,0.50}{##1}}}
\expandafter\def\csname PY@tok@go\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.53,0.53,0.53}{##1}}}
\expandafter\def\csname PY@tok@gt\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.27,0.87}{##1}}}
\expandafter\def\csname PY@tok@err\endcsname{\def\PY@bc##1{\setlength{\fboxsep}{0pt}\fcolorbox[rgb]{1.00,0.00,0.00}{1,1,1}{\strut ##1}}}
\expandafter\def\csname PY@tok@kc\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@kd\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@kn\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@kr\endcsname{\let\PY@bf=\textbf\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@bp\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.50,0.00}{##1}}}
\expandafter\def\csname PY@tok@fm\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.00,0.00,1.00}{##1}}}
\expandafter\def\csname PY@tok@vc\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.10,0.09,0.49}{##1}}}
\expandafter\def\csname PY@tok@vg\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.10,0.09,0.49}{##1}}}
\expandafter\def\csname PY@tok@vi\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.10,0.09,0.49}{##1}}}
\expandafter\def\csname PY@tok@vm\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.10,0.09,0.49}{##1}}}
\expandafter\def\csname PY@tok@sa\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@sb\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@sc\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@dl\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@s2\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@sh\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@s1\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.73,0.13,0.13}{##1}}}
\expandafter\def\csname PY@tok@mb\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@mf\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@mh\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@mi\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@il\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@mo\endcsname{\def\PY@tc##1{\textcolor[rgb]{0.40,0.40,0.40}{##1}}}
\expandafter\def\csname PY@tok@ch\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.25,0.50,0.50}{##1}}}
\expandafter\def\csname PY@tok@cm\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.25,0.50,0.50}{##1}}}
\expandafter\def\csname PY@tok@cpf\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.25,0.50,0.50}{##1}}}
\expandafter\def\csname PY@tok@c1\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.25,0.50,0.50}{##1}}}
\expandafter\def\csname PY@tok@cs\endcsname{\let\PY@it=\textit\def\PY@tc##1{\textcolor[rgb]{0.25,0.50,0.50}{##1}}}
\def\PYZbs{\char`\\}
\def\PYZus{\char`\_}
\def\PYZob{\char`\{}
\def\PYZcb{\char`\}}
\def\PYZca{\char`\^}
\def\PYZam{\char`\&}
\def\PYZlt{\char`\<}
\def\PYZgt{\char`\>}
\def\PYZsh{\char`\#}
\def\PYZpc{\char`\%}
\def\PYZdl{\char`\$}
\def\PYZhy{\char`\-}
\def\PYZsq{\char`\'}
\def\PYZdq{\char`\"}
\def\PYZti{\char`\~}
% for compatibility with earlier versions
\def\PYZat{@}
\def\PYZlb{[}
\def\PYZrb{]}
\makeatother
% Exact colors from NB
\definecolor{incolor}{rgb}{0.0, 0.0, 0.5}
\definecolor{outcolor}{rgb}{0.545, 0.0, 0.0}
% Prevent overflowing lines due to hard-to-break entities
\sloppy
% Setup hyperref package
\hypersetup{
breaklinks=true, % so long urls are correctly broken across lines
colorlinks=true,
urlcolor=urlcolor,
linkcolor=linkcolor,
citecolor=citecolor,
}
% Slightly bigger margins than the latex defaults
\geometry{verbose,tmargin=1in,bmargin=1in,lmargin=1in,rmargin=1in}
\begin{document}
\maketitle
\subsection{Predicting Hard Drive Failures Using SMART
Metrics}\label{predicting-hard-drive-failures-using-smart-metrics}
\subsubsection{\texorpdfstring{ - A Juul Labs Case
Study}{ - A Juul Labs Case Study}}\label{a-juul-labs-case-study}
\begin{verbatim}
- by Harshvardhan Pandey | Jan 09 2019
\end{verbatim}
\subsubsection{What are SMART systems ?}\label{what-are-smart-systems}
SMART features or \emph{S.M.A.R.T. (Self-Monitoring, Analysis and
Reporting Technology)} is a software monitoring system for hard drives.
It is a widely used industry practice around data center management and
disk heavy resources. SMART generates a collection different metrics
related to help evaluate the overall health of a Hard Drive. These
metrics can be specific to a certain number of manufacturers or be more
general, sometimes.
A single metrics may not always determine the exact failure prediction
but are commonly accepted to help identify any imminent failure and help
handle the backup and restore, in time.
\subsubsection{About this case study :}\label{about-this-case-study}
This case study relies on a given data stream provided for this purpose.
The goal of this case study is to try and analyze given data and find
out meaningful information that can help determine drives failure trends
and different factors that may idicate if a drive would fail, and
attempt to propose a more data driven answer to future failures based on
SMART metrics.
The study concludes with discussing possible opportunities and
challenges with existing model and features that can help design a
better predictive model for future.
\begin{center}\rule{0.5\linewidth}{\linethickness}\end{center}
Here's a quick look of how this problem has been approached:
\subsubsection{Extraction and Load}\label{extraction-and-load}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
Connect to the postgres server.
\item
Download the dataset offline
\end{enumerate}
\subsubsection{Transform}\label{transform}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\setcounter{enumi}{2}
\tightlist
\item
Wrangle and explore
\item
Change Dimentions, clean and slice and dice
\end{enumerate}
\subsubsection{Analyze}\label{analyze}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\setcounter{enumi}{4}
\tightlist
\item
Analyze dataset, plot most significant trends
\end{enumerate}
\subsubsection{Predict:}\label{predict}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\setcounter{enumi}{5}
\tightlist
\item
Feature Selection
\item
Model and predict
\end{enumerate}
\subsubsection{Conclusion and Improvement
Ideas:}\label{conclusion-and-improvement-ideas}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\setcounter{enumi}{7}
\tightlist
\item
Conclude
\item
Challenges with the current dataset and ways to improve it
\end{enumerate}
\begin{longtable}[]{@{}l@{}}
\toprule
\begin{minipage}[t]{0.15\columnwidth}\raggedright\strut
\#\# Extraction and Load\strut
\end{minipage}\tabularnewline
\begin{minipage}[t]{0.15\columnwidth}\raggedright\strut
1. Connect to the postgres server\strut
\end{minipage}\tabularnewline
\bottomrule
\end{longtable}
\begin{itemize}
\tightlist
\item
I'll begin by importing libraries to connect to postgres and download
the dataset offline.
\item
I will create a few database utility funtion get the table data and
columns
\item
next up, I will use pandas to join columns and dataset and transform
incoming data into a pandas dataframe.
\item
Lastly I will save the data locally in a csv format.
\end{itemize}
Next up, I will beging wrangling and exploring the data to understand
different attributes that will be used later on in analysis.
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}72}]:} \PY{k+kn}{import} \PY{n+nn}{psycopg2}
\PY{k+kn}{import} \PY{n+nn}{matplotlib}\PY{n+nn}{.}\PY{n+nn}{pyplot} \PY{k}{as} \PY{n+nn}{plt}
\PY{o}{\PYZpc{}}\PY{k}{matplotlib} inline
\PY{k+kn}{import} \PY{n+nn}{pandas} \PY{k}{as} \PY{n+nn}{pd}
\PY{k+kn}{from} \PY{n+nn}{sklearn} \PY{k}{import} \PY{n}{ensemble}\PY{p}{,} \PY{n}{metrics}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}10}]:} \PY{c+c1}{\PYZsh{}\PYZsh{}\PYZsh{}\PYZsh{} postgres database utility functions}
\PY{c+c1}{\PYZsh{}\PYZsh{}\PYZsh{}\PYZsh{}}
\PY{c+c1}{\PYZsh{} db connection object generator}
\PY{k}{def} \PY{n+nf}{postgres\PYZus{}db\PYZus{}connection}\PY{p}{(}\PY{p}{)}\PY{p}{:}
\PY{c+c1}{\PYZsh{} postgresql://35.230.114.237\PYZdq{}, \PYZdq{}postgres\PYZdq{}, \PYZdq{}luuj\PYZdq{}}
\PY{n}{conn} \PY{o}{=} \PY{n}{psycopg2}\PY{o}{.}\PY{n}{connect}\PY{p}{(}\PY{n}{host}\PY{o}{=}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{35.230.114.237}\PY{l+s+s2}{\PYZdq{}}\PY{p}{,} \PY{n}{dbname}\PY{o}{=}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{postgres}\PY{l+s+s2}{\PYZdq{}}\PY{p}{,}
\PY{n}{user}\PY{o}{=}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{candidate}\PY{l+s+s2}{\PYZdq{}}\PY{p}{,} \PY{n}{password}\PY{o}{=}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{luuj}\PY{l+s+s2}{\PYZdq{}}\PY{p}{)}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Connecting to postgresql server...}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{n}{cur} \PY{o}{=} \PY{n}{conn}\PY{o}{.}\PY{n}{cursor}\PY{p}{(}\PY{p}{)}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Successfully connected to the host}\PY{l+s+se}{\PYZbs{}n}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{k}{return} \PY{n}{cur}
\PY{k}{def} \PY{n+nf}{get\PYZus{}all\PYZus{}tables}\PY{p}{(}\PY{n}{cur}\PY{p}{)}\PY{p}{:}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Extracting list of tables:}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{n}{cur}\PY{o}{.}\PY{n}{execute}\PY{p}{(}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{SELECT * FROM pg\PYZus{}catalog.pg\PYZus{}tables where schemaname NOT IN (}\PY{l+s+s2}{\PYZsq{}}\PY{l+s+s2}{pg\PYZus{}catalog}\PY{l+s+s2}{\PYZsq{}}\PY{l+s+s2}{, }\PY{l+s+s2}{\PYZsq{}}\PY{l+s+s2}{information\PYZus{}schema}\PY{l+s+s2}{\PYZsq{}}\PY{l+s+s2}{)}\PY{l+s+s2}{\PYZdq{}}\PY{p}{)}
\PY{n}{tables} \PY{o}{=} \PY{n}{cur}\PY{o}{.}\PY{n}{fetchall}\PY{p}{(}\PY{p}{)}
\PY{n}{t} \PY{o}{=} \PY{p}{[}\PY{n}{i}\PY{p}{[}\PY{l+m+mi}{1}\PY{p}{]} \PY{k}{for} \PY{n}{i} \PY{o+ow}{in} \PY{n}{tables}\PY{p}{]}
\PY{k}{return} \PY{n}{t}
\PY{k}{def} \PY{n+nf}{lookup\PYZus{}a\PYZus{}table}\PY{p}{(}\PY{n}{cur}\PY{p}{,} \PY{n}{tablename}\PY{p}{)}\PY{p}{:}
\PY{c+c1}{\PYZsh{} get data from a given table: tablename}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+se}{\PYZbs{}n}\PY{l+s+s2}{Reading table: }\PY{l+s+s2}{\PYZdq{}}\PY{o}{+}\PY{n}{tablename}\PY{o}{+}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{...}\PY{l+s+s2}{\PYZdq{}}\PY{p}{)}
\PY{c+c1}{\PYZsh{} cur.execute(\PYZsq{}SELECT * from \PYZsq{}+tablename+\PYZsq{} limit 10\PYZsq{})}
\PY{c+c1}{\PYZsh{} get table\PYZus{}data}
\PY{n}{cur}\PY{o}{.}\PY{n}{execute}\PY{p}{(}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{SELECT * from }\PY{l+s+s2}{\PYZdq{}}\PY{o}{+}\PY{n}{tablename}\PY{o}{+}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{ limit 10}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{n}{table\PYZus{}data} \PY{o}{=} \PY{n}{cur}\PY{o}{.}\PY{n}{fetchall}\PY{p}{(}\PY{p}{)}
\PY{k}{return} \PY{n}{table\PYZus{}data}
\PY{k}{def} \PY{n+nf}{get\PYZus{}table\PYZus{}columns}\PY{p}{(}\PY{n}{cur}\PY{p}{,} \PY{n}{tablename}\PY{p}{)}\PY{p}{:}
\PY{c+c1}{\PYZsh{} get column\PYZus{}names}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Fetching columns in: }\PY{l+s+s1}{\PYZsq{}}\PY{p}{,} \PY{n}{tablename}\PY{p}{)}
\PY{k}{try}\PY{p}{:}
\PY{n}{cur}\PY{o}{.}\PY{n}{execute}\PY{p}{(}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{SELECT table\PYZus{}name, column\PYZus{}name from information\PYZus{}schema.columns where table\PYZus{}name = }\PY{l+s+s2}{\PYZsq{}}\PY{l+s+s2}{\PYZdq{}}\PY{o}{+}\PY{n}{tablename}\PY{o}{+}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{\PYZsq{}}\PY{l+s+s2}{\PYZdq{}}\PY{p}{)}
\PY{n}{column\PYZus{}names} \PY{o}{=} \PY{n}{cur}\PY{o}{.}\PY{n}{fetchall}\PY{p}{(}\PY{p}{)}
\PY{n}{column\PYZus{}names} \PY{o}{=} \PY{p}{[}\PY{n}{j}\PY{p}{[}\PY{l+m+mi}{1}\PY{p}{]} \PY{k}{for} \PY{n}{j} \PY{o+ow}{in} \PY{n}{column\PYZus{}names}\PY{p}{]}
\PY{k}{except}\PY{p}{:}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Column fetch failed}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{k}{return} \PY{n}{column\PYZus{}names}
\PY{c+c1}{\PYZsh{} transform data in pandas and save table locally for offline analysis}
\PY{k}{def} \PY{n+nf}{clean\PYZus{}response}\PY{p}{(}\PY{n}{table}\PY{p}{,} \PY{n}{data}\PY{p}{,} \PY{n}{column\PYZus{}names}\PY{p}{)}\PY{p}{:}
\PY{c+c1}{\PYZsh{} inp: table data and column\PYZus{}names}
\PY{c+c1}{\PYZsh{} out: pandas dataframe}
\PY{n}{data} \PY{o}{=} \PY{n}{pd}\PY{o}{.}\PY{n}{DataFrame}\PY{p}{(}\PY{n}{data}\PY{p}{)}
\PY{n}{data}\PY{o}{.}\PY{n}{columns} \PY{o}{=} \PY{p}{[}\PY{n}{column\PYZus{}names}\PY{p}{]}
\PY{n}{out\PYZus{}file} \PY{o}{=} \PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{out\PYZus{}data\PYZus{}from\PYZus{}tablename\PYZus{}}\PY{l+s+s1}{\PYZsq{}}\PY{o}{+}\PY{n}{table}\PY{o}{+}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{.csv}\PY{l+s+s1}{\PYZsq{}}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Saving data from table: }\PY{l+s+si}{\PYZob{}\PYZcb{}}\PY{l+s+s1}{, to file: }\PY{l+s+si}{\PYZob{}\PYZcb{}}\PY{l+s+s1}{\PYZsq{}}\PY{o}{.}\PY{n}{format}\PY{p}{(}\PY{n}{table}\PY{p}{,} \PY{n}{out\PYZus{}file}\PY{p}{)}\PY{p}{)}
\PY{c+c1}{\PYZsh{} data.to\PYZus{}csv(out\PYZus{}file, index=False, encoding=\PYZsq{}utf\PYZhy{}8\PYZsq{})}
\PY{c+c1}{\PYZsh{} Etracting all tables at the host in a list and finally,}
\PY{c+c1}{\PYZsh{} extracting the table we want i.e. \PYZsq{}hard\PYZus{}drive\PYZus{}stats}
\PY{n}{db\PYZus{}conn\PYZus{}obj} \PY{o}{=} \PY{n}{postgres\PYZus{}db\PYZus{}connection}\PY{p}{(}\PY{p}{)}
\PY{n}{tables} \PY{o}{=} \PY{n}{get\PYZus{}all\PYZus{}tables}\PY{p}{(}\PY{n}{db\PYZus{}conn\PYZus{}obj}\PY{p}{)}
\PY{n}{table} \PY{o}{=} \PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{hard\PYZus{}drive\PYZus{}stats}\PY{l+s+s1}{\PYZsq{}}
\PY{n}{data} \PY{o}{=} \PY{n}{lookup\PYZus{}a\PYZus{}table}\PY{p}{(}\PY{n}{db\PYZus{}conn\PYZus{}obj}\PY{p}{,} \PY{n}{table}\PY{p}{)}
\PY{n}{table\PYZus{}data} \PY{o}{=} \PY{n}{lookup\PYZus{}a\PYZus{}table}\PY{p}{(}\PY{n}{db\PYZus{}conn\PYZus{}obj}\PY{p}{,} \PY{n}{table}\PY{p}{)}
\PY{n}{table\PYZus{}column\PYZus{}names} \PY{o}{=} \PY{n}{get\PYZus{}table\PYZus{}columns}\PY{p}{(}\PY{n}{db\PYZus{}conn\PYZus{}obj}\PY{p}{,} \PY{n}{table}\PY{p}{)}
\PY{c+c1}{\PYZsh{} transform data in pandas}
\PY{n}{clean\PYZus{}response}\PY{p}{(}\PY{n}{table}\PY{p}{,} \PY{n}{table\PYZus{}data}\PY{p}{,} \PY{n}{table\PYZus{}column\PYZus{}names}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
Connecting to postgresql server{\ldots}
Successfully connected to the host
Extracting list of tables:
Reading table: hard\_drive\_stats{\ldots}
Reading table: hard\_drive\_stats{\ldots}
Fetching columns in: hard\_drive\_stats
Saving data from table: hard\_drive\_stats, to file: out\_data\_from\_tablename\_hard\_drive\_stats.csv
\end{Verbatim}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
At the end of the above \texttt{code} snippet, data is downloaded and
saved locally to current directory.
\item
Name: out\_data\_from\_tablename\_hard\_drive\_stats.csv
\end{enumerate}
\begin{center}\rule{0.5\linewidth}{\linethickness}\end{center}
\subsection{Transform}\label{transform}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
Now, the dataset is downloaded.
\item
Filename is: out\_data\_from\_tablename\_hard\_drive\_stats.csv
\item
We shall be be using this file going forward, in order to avoid
calling the postgres again and again.
\end{enumerate}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}73}]:} \PY{c+c1}{\PYZsh{} loading dataset from local machine}
\PY{n}{df} \PY{o}{=} \PY{n}{pd}\PY{o}{.}\PY{n}{read\PYZus{}csv}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{out\PYZus{}data\PYZus{}from\PYZus{}tablename\PYZus{}hard\PYZus{}drive\PYZus{}stats.csv}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{n}{df}\PY{o}{.}\PY{n}{head}\PY{p}{(}\PY{l+m+mi}{5}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{outcolor}Out[{\color{outcolor}73}]:} row.names date serial\_number model \textbackslash{}
0 1865121 2018-01-20 ZA11RRZY ST8000DM002
1 1865122 2018-01-20 PL1331LAHD3Y7H HGST HMS5C4040BLE640
2 1865123 2018-01-20 ZA174A42 ST8000NM0055
3 1865124 2018-01-20 PL1331LAHGB3VH HGST HMS5C4040ALE640
4 1865125 2018-01-20 ZA14ELXG ST8000NM0055
capacity\_bytes failure read\_error\_rate throughput\_performance \textbackslash{}
0 8001563222016 0 209151808.0 NaN
1 4000787030016 0 0.0 106.0
2 8001563222016 0 28504744.0 NaN
3 4000787030016 0 0.0 104.0
4 8001563222016 0 77116864.0 NaN
spin\_up\_time start\_stop\_count reallocated\_sector seek\_time\_performance \textbackslash{}
0 0.0 3.0 0.0 NaN
1 0.0 4.0 0.0 42.0
2 0.0 4.0 0.0 NaN
3 431.0 5.0 0.0 42.0
4 0.0 3.0 0.0 NaN
power\_on\_hours power\_cycle\_count reported\_uncorrect command\_timeout \textbackslash{}
0 13874.0 3.0 0.0 0.0
1 14041.0 4.0 NaN NaN
2 4726.0 4.0 0.0 0.0
3 8475.0 5.0 NaN NaN
4 7309.0 3.0 0.0 0.0
high\_fly\_writes airflow\_temprature load\_cycle\_count total\_lbas\_written
0 0.0 29.0 3888.0 5.189666e+10
1 NaN NaN 143.0 NaN
2 0.0 35.0 716.0 2.152110e+10
3 NaN NaN 11.0 NaN
4 0.0 34.0 2380.0 3.561304e+10
\end{Verbatim}
\paragraph{Get basic look of the
dataset}\label{get-basic-look-of-the-dataset}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}12}]:} \PY{c+c1}{\PYZsh{} number of rows}
\PY{n}{rows} \PY{o}{=} \PY{n}{df}\PY{o}{.}\PY{n}{shape}\PY{p}{[}\PY{l+m+mi}{0}\PY{p}{]}
\PY{n}{columns} \PY{o}{=} \PY{n}{df}\PY{o}{.}\PY{n}{shape}\PY{p}{[}\PY{l+m+mi}{1}\PY{p}{]}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Number of rows are: }\PY{l+s+si}{\PYZob{}\PYZcb{}}\PY{l+s+s1}{ and number of columns: }\PY{l+s+si}{\PYZob{}\PYZcb{}}\PY{l+s+se}{\PYZbs{}n}\PY{l+s+s1}{\PYZsq{}}\PY{o}{.}\PY{n}{format}\PY{p}{(}\PY{n}{rows}\PY{p}{,} \PY{n}{columns}\PY{p}{)}\PY{p}{)}
\PY{n+nb}{print}\PY{p}{(}\PY{n}{df}\PY{o}{.}\PY{n}{dtypes}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
Number of rows are: 8949492 and number of columns: 20
row.names int64
date object
serial\_number object
model object
capacity\_bytes int64
failure int64
read\_error\_rate float64
throughput\_performance float64
spin\_up\_time float64
start\_stop\_count float64
reallocated\_sector float64
seek\_time\_performance float64
power\_on\_hours float64
power\_cycle\_count float64
reported\_uncorrect float64
command\_timeout float64
high\_fly\_writes float64
airflow\_temprature float64
load\_cycle\_count float64
total\_lbas\_written float64
dtype: object
\end{Verbatim}
\paragraph{More wrangling of the data}\label{more-wrangling-of-the-data}
First up, I get rid of some irrelevant columns and then indentify the
top 10 hard drives.
I will apply some cleaning on the columns, changing dtypes and more.
Next, I discard/drop columns based on: 1. high number of Nan 2.
irrelevance 3. top 10 models
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}13}]:} \PY{c+c1}{\PYZsh{} drop everything where 10 or more rows are Nan}
\PY{n}{df} \PY{o}{=} \PY{n}{df}\PY{o}{.}\PY{n}{dropna}\PY{p}{(}\PY{n}{thresh}\PY{o}{=}\PY{l+m+mi}{10}\PY{p}{,} \PY{n}{axis}\PY{o}{=}\PY{l+m+mi}{0}\PY{p}{)}
\PY{c+c1}{\PYZsh{} drop column row.names}
\PY{n}{df} \PY{o}{=} \PY{n}{df}\PY{o}{.}\PY{n}{drop}\PY{p}{(}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{row.names}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]}\PY{p}{,} \PY{n}{axis}\PY{o}{=}\PY{l+m+mi}{1}\PY{p}{)}
\PY{c+c1}{\PYZsh{} change data type}
\PY{n}{new\PYZus{}date} \PY{o}{=} \PY{n}{pd}\PY{o}{.}\PY{n}{to\PYZus{}datetime}\PY{p}{(}\PY{n}{df}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{date}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]}\PY{p}{)}
\PY{n}{df}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{date}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]} \PY{o}{=} \PY{n}{new\PYZus{}date}
\end{Verbatim}
Data types now look like this:
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}14}]:} \PY{n}{df}\PY{o}{.}\PY{n}{shape}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{outcolor}Out[{\color{outcolor}14}]:} (8949141, 19)
\end{Verbatim}
\subsubsection{Top 10 most common hard
drives}\label{top-10-most-common-hard-drives}
\begin{itemize}
\tightlist
\item
After getting rid of some of the duplicates.
\item
Now based the dataset, I am making the following assumptions:
\end{itemize}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
Hard drives with number of datapoints are the most common hard drives.
\item
Since there are multiple serial\_numbers that belong to the same Hard
Drive model, I am taking a unique count only.
\end{enumerate}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}15}]:} \PY{c+c1}{\PYZsh{} getting a list of the msot common drives}
\PY{n}{most\PYZus{}common\PYZus{}models} \PY{o}{=} \PY{n}{df}\PY{o}{.}\PY{n}{groupby}\PY{p}{(}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{model}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]}\PY{p}{,} \PY{n}{as\PYZus{}index}\PY{o}{=}\PY{k+kc}{True}\PY{p}{)}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{model}\PY{l+s+s1}{\PYZsq{}}\PY{p}{,} \PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{date}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]}\PY{o}{.} \PY{n}{size}\PY{p}{(}\PY{p}{)}
\PY{n}{most\PYZus{}common\PYZus{}models} \PY{o}{=} \PY{n}{most\PYZus{}common\PYZus{}models}\PY{o}{.}\PY{n}{sort\PYZus{}values}\PY{p}{(}\PY{n}{ascending}\PY{o}{=}\PY{k+kc}{False}\PY{p}{)}
\end{Verbatim}
Most Common Models in descending order are:
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}16}]:} \PY{c+c1}{\PYZsh{} number of different models }
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+s2}{There are 53 models: }\PY{l+s+s2}{\PYZdq{}}\PY{p}{,} \PY{n+nb}{len}\PY{p}{(}\PY{n}{most\PYZus{}common\PYZus{}models}\PY{p}{)}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
There are 53 models: 53
\end{Verbatim}
\paragraph{Top 10 Models based on most number of Hard Drives
are}\label{top-10-models-based-on-most-number-of-hard-drives-are}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}17}]:} \PY{n}{top10\PYZus{}models} \PY{o}{=} \PY{n}{most\PYZus{}common\PYZus{}models}\PY{p}{[}\PY{p}{:}\PY{l+m+mi}{10}\PY{p}{]}
\PY{n}{top10\PYZus{}models}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{outcolor}Out[{\color{outcolor}17}]:} model
ST4000DM000 2822270
HGST HMS5C4040BLE640 1363173
ST12000NM0007 1296241
ST8000NM0055 1293502
ST8000DM002 888733
HGST HMS5C4040ALE640 505026
ST6000DX000 169017
Hitachi HDS5C4040ALE630 115984
ST10000NM0086 109738
HGST HUH728080ALE600 94024
dtype: int64
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}61}]:} \PY{n}{top10\PYZus{}models}\PY{o}{.}\PY{n}{plot}\PY{p}{(}\PY{n}{kind}\PY{o}{=}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{bar}\PY{l+s+s1}{\PYZsq{}}\PY{p}{,} \PY{n}{legend}\PY{o}{=} \PY{k+kc}{False}\PY{p}{)}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{Top 10 common models and the number of hard Drives in each:}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
Top 10 common models and the number of hard Drives in each:
\end{Verbatim}
\begin{center}
\adjustimage{max size={0.9\linewidth}{0.9\paperheight}}{output_22_1.png}
\end{center}
{ \hspace*{\fill} \\}
\subsubsection{Filtering - Limiting the dataset by only top 10
models}\label{filtering---limiting-the-dataset-by-only-top-10-models}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}62}]:} \PY{c+c1}{\PYZsh{} top 10 models}
\PY{n}{list\PYZus{}top10\PYZus{}models} \PY{o}{=} \PY{p}{[}\PY{n}{i} \PY{k}{for} \PY{n}{i} \PY{o+ow}{in} \PY{n}{top10\PYZus{}models}\PY{o}{.}\PY{n}{index}\PY{p}{]}
\PY{c+c1}{\PYZsh{} this is the new dataframe based on the top 10 models}
\PY{n}{new\PYZus{}df} \PY{o}{=} \PY{n}{df}\PY{p}{[}\PY{n}{df}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{model}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]}\PY{o}{.}\PY{n}{isin}\PY{p}{(}\PY{n}{list\PYZus{}top10\PYZus{}models}\PY{p}{)}\PY{p}{]}
\PY{n}{new\PYZus{}df}\PY{o}{.}\PY{n}{shape}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{outcolor}Out[{\color{outcolor}62}]:} (8657708, 19)
\end{Verbatim}
\begin{center}\rule{0.5\linewidth}{\linethickness}\end{center}
\paragraph{Using the new dataframe from here
on}\label{using-the-new-dataframe-from-here-on}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}63}]:} \PY{c+c1}{\PYZsh{} Changing datatypes}
\PY{n}{new\PYZus{}df}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{failure}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]} \PY{o}{=} \PY{n}{new\PYZus{}df}\PY{p}{[}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{failure}\PY{l+s+s1}{\PYZsq{}}\PY{p}{]}\PY{o}{.}\PY{n}{astype}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{bool}\PY{l+s+s1}{\PYZsq{}}\PY{p}{)}
\PY{c+c1}{\PYZsh{} saving top10 models data to csv}
\PY{c+c1}{\PYZsh{} new\PYZus{}df.to\PYZus{}csv(\PYZsq{}top10\PYZus{}models.csv\PYZsq{}, index=False)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
/anaconda3/lib/python3.6/site-packages/ipykernel\_launcher.py:2: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row\_indexer,col\_indexer] = value instead
See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html\#indexing-view-versus-copy
\end{Verbatim}
\begin{center}\rule{0.5\linewidth}{\linethickness}\end{center}
\subsection{Analysis}\label{analysis}
Keeping in mind that the resources available \emph{do not} accurately
describe this particular dataset. It is crucial to proceed with caution.
I researched online and read a number of articles. I settled one the
ones I found most relevant. I have used this information to help me
understand the schema and it's various attributes.
\subsubsection{Resources:}\label{resources}
These are some of the resources I found helpful.
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
Understanding differet SMART stats:
https://www.backblaze.com/blog/what-smart-stats-indicate-hard-drive-failures/
\item
SMART schema on WIKI:
https://en.wikipedia.org/wiki/S.M.A.R.T.\#ATA\_S.M.A.R.T.\_attributes
\item
Research Paper:
http://cs229.stanford.edu/proj2017/final-reports/5242080.pdf
\end{enumerate}
\subsubsection{Tools:}\label{tools}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
I have utilized scikit library for prediction.
\item
Partly used pandas and Google Big Query for faster analysis in SQL,
and
\item
matplotlib + Google Data Studio for plotting charts.
\end{enumerate}
\paragraph{Let's check the cardinality of each
columns}\label{lets-check-the-cardinality-of-each-columns}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}21}]:} \PY{c+c1}{\PYZsh{} find unique values per columns}
\PY{k}{for} \PY{n}{cols} \PY{o+ow}{in} \PY{n}{new\PYZus{}df}\PY{o}{.}\PY{n}{columns}\PY{p}{:}
\PY{n+nb}{print}\PY{p}{(}\PY{l+s+s2}{\PYZdq{}}\PY{l+s+si}{\PYZob{}\PYZcb{}}\PY{l+s+s2}{ has: }\PY{l+s+si}{\PYZob{}\PYZcb{}}\PY{l+s+s2}{ unique values}\PY{l+s+s2}{\PYZdq{}}\PY{o}{.}\PY{n}{format}\PY{p}{(}\PY{n}{cols}\PY{p}{,} \PY{n+nb}{len}\PY{p}{(}\PY{n}{new\PYZus{}df}\PY{p}{[}\PY{n}{cols}\PY{p}{]}\PY{o}{.}\PY{n}{unique}\PY{p}{(}\PY{p}{)}\PY{p}{)}\PY{p}{)}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
date has: 90 unique values
serial\_number has: 101310 unique values
model has: 10 unique values
capacity\_bytes has: 5 unique values
failure has: 2 unique values
read\_error\_rate has: 5938125 unique values
throughput\_performance has: 67 unique values
spin\_up\_time has: 288 unique values
start\_stop\_count has: 187 unique values
reallocated\_sector has: 1023 unique values
seek\_time\_performance has: 11 unique values
power\_on\_hours has: 43903 unique values
power\_cycle\_count has: 123 unique values
reported\_uncorrect has: 70 unique values
command\_timeout has: 161 unique values
high\_fly\_writes has: 1175 unique values
airflow\_temprature has: 45 unique values
load\_cycle\_count has: 190691 unique values
total\_lbas\_written has: 6562030 unique values
\end{Verbatim}
\subsection{\#\#\# Plotting graphs to get a visual look and
analyze}\label{plotting-graphs-to-get-a-visual-look-and-analyze}
\paragraph{Using Google Data Studio and Big
Query:}\label{using-google-data-studio-and-big-query}
\begin{itemize}
\tightlist
\item
Google Data Studio provides for a much more robust and interactive
reporting system.
\item
I loaded the dataset into Big Query and used Google Data Studio
because of it's SQL support, interactive platform and robustness with
doing exploratory analysis on a large dataset.
\item
There are some key charts provided below.
\end{itemize}
\paragraph{Click Here for a full report:
https://datastudio.google.com/open/1vzmbcHsLQ-OMZZsfXUnECJbIteK\_kdF7}\label{click-here-for-a-full-report-httpsdatastudio.google.comopen1vzmbchslq-omzzsfxunecjbitek_kdf7}
\paragraph{Looking at the trends displayed below, we can derive the
following (refer to the Studio Report below)
:}\label{looking-at-the-trends-displayed-below-we-can-derive-the-following-refer-to-the-studio-report-below}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\item
Number of positive Hard drives failure trend are going down. This
trend is proportional to the power cycle of the hard drives. This
means that as the hard drives get old over time, they are more likely
to fail. This is also verified by external sources, a typical life of
a hard drive is around 5 years. This can help find out the likelyhood
of a drive failing.
\item
Reported uncorrectable errors tend to go down as the failure count
goes down over time. On the other hand, the reallocated sectors a
going up. Both of these features should ideally be of a lower value
for a healthy hard drive. There are higher chances of failure if both
of these factors go up in the future.
\item
Hard Drives have to reallocated sectors at a much higher rate in the
event of a failure. This happens becase hard drives need to remap the
data to a different sector in order to avoid data loss. Frequent
remapping like this is not a good sign of a healthy hard drive.
\item
High fly rates decrease with decrease in failure. This may indicate
that a lower high fly rate is a potential sign of healthier hard
drives.
\end{enumerate}
There are more exploratory analysis performed on google data studio
report, link is provided below.
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}67}]:} \PY{c+c1}{\PYZsh{} rendering google data studio report}
\PY{k+kn}{from} \PY{n+nn}{IPython}\PY{n+nn}{.}\PY{n+nn}{display} \PY{k}{import} \PY{n}{IFrame}
\PY{n}{IFrame}\PY{p}{(}\PY{l+s+s1}{\PYZsq{}}\PY{l+s+s1}{https://datastudio.google.com/embed/reporting/1vzmbcHsLQ\PYZhy{}OMZZsfXUnECJbIteK\PYZus{}kdF7/page/xJMf}\PY{l+s+s1}{\PYZsq{}}\PY{p}{,} \PY{n}{width}\PY{o}{=}\PY{l+m+mi}{900}\PY{p}{,} \PY{n}{height}\PY{o}{=}\PY{l+m+mi}{675}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{outcolor}Out[{\color{outcolor}67}]:} <IPython.lib.display.IFrame at 0x10bc9af98>
\end{Verbatim}
\subsection{\#\#\#\# Assumptions made and references drawn, in
performing the
analysis:}\label{assumptions-made-and-references-drawn-in-performing-the-analysis}
Since there isn't enough detail about the dataset in this case study,
some external researching is required to get an understanding.
There are exponetial values in some of SMART metrics. The provided data
stream is raw and there isn't much information available online about
different expoential values. I couldn't find a meaningful method to
normalize the raw data to a 100 point scale in order to make a better
correlation.
\subsection{\#\#\#\# Analysis Conclusion}\label{analysis-conclusion}
In conclusion, the metrics in SMART systems are most often high
uncorrelated. It wouldn't be recommended to rely on one of them to make
a decision about a possible drive failure.
\paragraph{(This is Optional)}\label{this-is-optional}
\emph{In case the above embeded code for Data Studio report failed, I am
including local PNG import of some of the charts.}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
Number Hard Drives per model
\begin{enumerate}
\def\labelenumii{\arabic{enumii}.}
\setcounter{enumii}{1}
\tightlist
\item
Number of positive failures by model
\end{enumerate}
\end{enumerate}
\begin{verbatim}
<tr>
<td>3. Failure Trend over time
<img src="graphs/failure-trend-timeseries.png" width="600"></td>
\end{verbatim}
\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\setcounter{enumi}{3}
\tightlist
\item
Daily Failure Trend to determine missing failure data pattern
\end{enumerate}
\begin{center}\rule{0.5\linewidth}{\linethickness}\end{center}
\subsection{Machine learning to Predict Possible Failures
based}\label{machine-learning-to-predict-possible-failures-based}
\subsubsection{Feature selection:}\label{feature-selection}
Based on my findings and research on SMART attributes, I have found the
following variables to be the most significant out of the total
available dataset. The variables are highly non correlated, I made the
selection based on what works as a industry standard for SMART
predictions.
\begin{Verbatim}[commandchars=\\\{\}]
{\color{incolor}In [{\color{incolor}23}]:} \PY{c+c1}{\PYZsh{} get all columns and the number of NaN in them}
\PY{n}{new\PYZus{}df}\PY{o}{.}\PY{n}{isna}\PY{p}{(}\PY{p}{)}\PY{o}{.}\PY{n}{sum}\PY{p}{(}\PY{p}{)}
\end{Verbatim}
\begin{Verbatim}[commandchars=\\\{\}]
{\color{outcolor}Out[{\color{outcolor}23}]:} date 0
serial\_number 0
model 0
capacity\_bytes 0
failure 0
read\_error\_rate 0
throughput\_performance 6579501
spin\_up\_time 0
start\_stop\_count 0