*** Last Change 16JUN94
Fix typos in parameter file for SQSKY. The user can select the type of statistic to apply to the data (norm_stat) and whether to offset or scale the data to a common statistic ("norm_opt") prior to imcombine.
Handle case where do transformation file present for TRANSMAT
Separated "nircombine" into nircombine, which supports only current (>= 2.10.2) imcombine and "onircombine" which supports only older (<= 2.20.1) imcombine.
NIRCOMBINE work
Modify stack database; submit intensity offsets to imcombine rather than perform imarith prior tro submission.
Fix bug in recovering intensity offsets from imcombine log: When less than 10 images are combined, the format is imname[n] median adjust xoff yoff However when 10 or more images are combined, the format is: imname[ n] median adjust xoff yoff ; n < 10 imname[nn] median adjust xoff yoff ; n >= 10 Since the field count varies for ten or more images, the assumption that field 3 contained the "adjust" value fails on images 1-9 (median is extracted instead).
Replace fscan with scan from a pipe to get database parameters.
Add apply_zero flag to allow choice of: (1) applying zero-point shifts via imarith prior to imcombine, so that zeropoint adjusted results will be saved via make_stack+ or save_images+. (2) submitting intensity offsets file directly to imcombine, so that zeropoint unadjusted results will be saved via make_stack+ or save_images+.
Note: IRAF V2.10EXPORT April 1992 imcombine has an inconsistency in the sign convention on applied zeropoint. The sign of the file based zeropoint assumes the value to be subtracted, whereas the reported number is the value to be added. These have been reconciled to "value to be added" in the next release and current IRAFX zero_invert flag toggles these situations.
Tested results against original code. Results are indistinguishable. However, note that recorded log from IMCOMBINE can lose precision due to formatting. (This will be fixed inside next release IRAF imcombine.)
Fix so frame_nums = ""," ","all" work to correctly signify "use all COM lines".
Major "bug" fix:
Replaced internal writes to file which used "type" with "concatenate" or "copy" as appropriate. Within the core SQIID database routines, lines longer than a given length (e.g. 80 characters) were being wrapped to 80 characters with the rest of the line appearing as a new line in an enviromentally sensitive way: stty gterm worked OK stty xterm or stty xgterm exhibited the "wrap" behavior The above modification fixes this bug in the following routines: closure, getcenters, ircombine, linklaps, mergecom, nircombine, sqmos, transmat, xyadopt, xyget, xylap, xytrace, zget Apparently "type" is context sensitive to the tty even when writing to files.
Some changes/additions:
The tasks getcombine.cl, compose.cl, getmatch.cl, getoffsets.cl, which have not been kept current, have been dropped from the current distribution.
Sub-directory "prior" contains earlier versions of tasks: 1) which were changed in the current release. 2) which were frozen as compatible with earlier IRAF releases, such as the *.cl0 tasks mentioned below.
Some changes/additions:
chlist' and colorlist' have been updated to allow the new WILDFIRE
naming convention for IRIM/CRSP/COB (id_color = "seq"), which does not
append a channel designator (j|h|k|l) to the end of the image_name.
(The old FIRE convention is id_color = "predot" for SQIID.)
(The current WFIRE convention is id_color = "end" for SQIID.)
(The current WFIRE convention is id_color = "seq" for IRIM|COB|CRSP.)
sqmos' has been modified to allow wildcarded images like sqproc', sqky', sqdark', and `sqflat' already allow. Thus:
sqmos pn1_001_*j.imh mosj 3 3
will put the 9 files pn1_001_001j thru pn1_001_009j into the mosaic "mosj".
NB: IRAF2.10.2 contains a change to the IMCOMBINE task which requires changes to all tasks which use imcombine. The versions which work with 2.10.0 and 2.10.1 are labeled *.cl0 and the ones which work with 2.10.2 and later are labeled *.cl. To use the *.cl0 files, delete (or rename) the .cl versions and rename the *.cl0 files *.cl.
A preliminary discussion on how to use the "sqiid" package, by K. Michael Merrill and John Mackenty is in "DOC/sqiidpkg.doc". There are no help pages yet. The latest copy of the SQIID observation manual is in "DOC/sqiid.aug92".
Some changes/additions:
chlist' and colorlist' have been updated to allow the new WILDFIRE naming
convention (id_color = "end"), which appends the SQIID channel designator
(j|h|k|l) to the end of the image_name. (The old FIRE convention is
id_color = "predot")
nircombine' -- incorporates "frame_num" list to combine a designated subset of the data list and incorpates imstack' as a means of getting around
the 102 frame limit for "imcombine".
sqproc' -- the general list-process data task has undergone minor revisions and some new data processing tasks have been generated: sqnotch' uses a subset of a datalist within a specified distance of
each frame (the "notch") to generate a running SKY frame which is
subtracted from that frame.
sqtriad' - list-process SQIID raw image data of the form ON OFF ON ... ON OFF list# 1 2 list# 3 2 list# 4 5 list# 6 5 list# 7 8 list# 9 8 sq9pair' - list-process SQIID raw image data of the form ON OFF OFF ON ...
ON OFF
list# 1 2
list# 4 3
list# 5 6
list# 8 7
list# 9 10
list# 12 11
list# 13 14
list# 16 15
list# 17 18
show1' -- display and optionally sky subtract an image show4' -- display and optionally sky subtract all 4 channels of an image
show9' -- display and optionally sky subtract a 3x3 grid of image data show9pair' -- display sky subtracted data taken as noted above
which' -- interactive task which outputs the path_ids ("COM_XXXX") that were nircombined' to produce the image at a given image cursor
position. The resultant image must be in the display.
xyadopt' -- a number of changes, including an "all_images" parameter to over-ride "nraster" within a mosaic to allow all the images to be used. Previously, one was limited to the number of images within the "nraster" line of the database. Since one could have generated a longer listing (using mergecom'), the task
was amended to allow a larger number.
xyget' -- output modified to skip images which have not been adequately centroided. Previously, all images were included. Only on careful reading of verbose+ output, could one realize that some of the off-sets might be ill-defined. zget' -- minor changes
================================================================================
Assuming you copy the package somewhere in a directory called "sqiid", you would put the following definitions in you loginuser.cl file:
set sqscripts = "your_pathname/sqiid/"
task $sqiid = sqscripts$sqiid.cl
You also need to edit the sqiid.cl in the "sqiid" package to file to point to your path instead of my path!
Sample files needed to GEOTRAN images in one color to that of another color are included in the geotran subdirectory as follows: files for GEOTRAN the raw (S up E left) JH to K: n10[jh]tok.geo[co|db] files for GEOTRAN the raw (S up E left) HK to J: n10[hk]toj.geo[co|db] files for GEOTRAN the raw (S up E left) JK to H: n10[jk]toh.geo[co|db] files for GEOTRAN the oriented (N up E left) JH to K: n10[jh]toktr.geo[co|db] files for GEOTRAN the oriented (N up E left) HK to J: n10[hk]tojtr.geo[co|db] files for GEOTRAN the oriented (N up E left) JK to H: n10[jk]tohtr.geo[co|db]
Both (.geoco & .geodb) files are needed in the directory where images are GEOTRANed. See the README file inside the geotran directory for more details.
Since a number of the scripts call the UNIX "awk" command to do their fancy formatted output, "awk" must be available if the parameter format=yes is selected.
A preliminary user manual for the sqiid package is in "DOC/sqiid.doc"
The sqiid IRAF package is available from a tar file called "sqiid.tar" within the ftp-anonymous area at mira.tuc.noao.edu. You can retrieve it by using:
ftp mira.tuc.noao.edu ( aka 140.252.3.85) logging in as "anonymous" using your last name as password cd pub/sqiid to see what's there get sqiid.tar to get tar file bye
Contact me (merrill@noao.edu;602-325-9319) when you have successfully retrieved the programs and if you have further questions.
Program-related problems are most easily addressed if the appropriate info are included with the problem report:
copy of command line used to call the task
lpar of the task in question ( lpar task_name >> errfile )
copies of the ascii files called by that task (type filenames >> errfile)
version of IRAF used (= cl.version)
Michael Merrill