% README File FOR AIPS Release 15JUL94 % This File describes the patches to known bugs in 15JUL94 Tasks and other things for which patches have been made: =============================================================================== Task: Large angles in full-sky projections Summary: The low level routine DIRCOS did not handle large apparent differences beween RA and RA0. It should just rotate RA by 360 degrees but it didn't. Files: /15JUL94/APL/SUB/DIRCOS.FOR Fix: (1) Get the file and place it in $APLSUB (2) Compile it to the libraries with COMRPL $APLSUB/DIRCOS (3) Rebuild programs that need it with, e.g., COMLNK $AIPPGM/AIPS COMLNK $YPGM/*.FOR $YPGNOT/*.FOR COMLNK $QPGNOT/SDGRD Documentor: Eric W. Greisen Date: 1994.08.24 ------------------------------------------------------------------------ Task: UVDIF Summary: UVDIF failed to report differences when one data set was flagged and the other not - a case about which one would very much like to hear. Files: /15JUL94/APL/PGM/UVDIF.FOR Fix: (1) Fetch the file and put it in $APLPGM (2) Compile and link it with COMLNK $APLPGM/UVDIF Documentor: Eric W. Greisen Date: 1994.09.12 ------------------------------------------------------------------------ Task: SPLIT, CALIB, HORUS, FRING Summary: These tasks refused to apply requested flag tables to single-source uv data files. CALIB and FRING also refused to apply requested calibration tables. Other calibration tasks do honor these requests, so I turned off these arbitrary prohibitions. If the tables are present and requested, they will be used. Files: /15JUL94/APL/PGM/NOTST/SPLIT.FOR /15JUL94/Q/PGM/NOTST/CALIB.FOR /15JUL94/Q/PGM/NOTST/HORUS.FOR /15JUL94/Q/PGM/NOTST/FRING.FOR Fix: (1) Fetch the first file and put it in $APGNOT (2) Fetch the others and put them in $QPGNOT (2) Compile and link them with COMLNK $APGNOT/SPLIT $QPGNOT/HORUS COMLNK $QPGNOT/CALIB $QPGNOT/FRING Documentor: Eric W. Greisen Date: 1994.09.13 ------------------------------------------------------------------------ Task: LINK shell script Summary: The use of separate versions of LIBR.DAT for shared or non- shared libraries was not working correctly. For Sun4, there are two files LIBR.DAT and LIBR.DAT.SHARED supplied, and if the installer selects the use of shared libraries, the former is renamed LIBR.DAT.UNSHARED and the latter to LIBR.DAT. The generation of the list of libraries by LINK was not taking this into account. It does now. Files: /15JUL94/SYSTEM/UNIX/LINK Fix: (1) Fetch the file and put it in $SYSUNIX (2) Compile and link any tasks that failed to build before with COMLNK. Documentor: Patrick P. Murphy Date: 1994.09.13 ------------------------------------------------------------------------ Task: Single-dish software: part 1 of several Summary: Errors: (1) The single-dish data reading subroutines could not handle compressed data and mishandled channel and IF selection. (2) The print task PRTSD had minor bugs related to starting sample. The display of times was improved to show 0.1's of a second. Files: /15JUL94/APL/SUB/NOTST/SDGET.FOR /15JUL94/APL/SUB/NOTST/DGETSD.FOR /15JUL94/APL/PGM/NOTST/PRTSD.FOR Fix: (1) Get the first 2 files, placing them in $APLNOT. (2) Get the third file, placing it in $APGNOT. (3) Rebuild them with COMRPL $APLNOT/SDGET $APLNOT/DGETSD COMLNK $APGNOT/SELSD $APGNOT/PRTSD Documentor: Eric W. Greisen Date: 1994.08.23 ------------------------------------------------------------------------ Task: Single-dish software: part 2 of several Summary: Improvements: A new task called UVF was submitted by Jeff Mangum and revised by me. It reads SDD format data making a FITS disk file for UVLOD to read. This task requires the fitsio libraries (available, e.g. via anon ftp to legacy.gsfc.nasa.gov in /software/fitsio/). Files: /15JUL94/CONTRIBU/UVF.HLP /15JUL94/CONTRIBU/UVFAIP.FOR /15JUL94/CONTRIBU/UVF.C /15JUL94/CONTRIBU/UVF.OPT Fix: (1) Get all four files, placing them in YOUR OWN WORK AREA (do NOT put them in APLCONTR!!!). (2) If your system is not SUN4 architecture, modify the UVF.OPT file, replacing APLSUN with, e.g. APLSOL for Solaris, APLIBM for AIX, APLDEC for Alpha, etc. (3) Build them with COMRPL UVFAIP COMLNK UVF UVF.OPT This will only work if the fitsio libraries are available from your system link edit areas. (4) Copy all 4 files above to $APLCONTR. (5) Also cp UVF.HLP $HLPFIL cp UVF.EXE $LOAD Notes: (1) OTFUV removes most of the need for this task. Documentor: Eric W. Greisen Date: 1994.08.23, updated 1994.08.24, 26, 29, 30 ------------------------------------------------------------------------ Task: Single-dish software: part 3 of several (SEE ALSO 6 OCTOBER ENTRY BELOW) Summary: Improvements (1) A new gridding task called SDGRD was written. It handles coordinate projection, sorting, and multi-channel gridding all in one task. This should make SELSD and GRIDR superfluous. (2) A new task called CONPL was written to display convolution functions available in AIPS. (3) A new task called OTFUV reads the OTF data in SDD format directly into AIPS. Files: /15JUL94/APL/PGM/OTFUV.FOR /15JUL94/Q/PGM/NOTST/SDGRD.FOR /15JUL94/QY/PGM/NOTST/CONPL.FOR /15JUL94/APL/DEV/ZFIO.FOR /15JUL94/APL/DEV/UNIX/ZFI2.C /15JUL94/HELP/SDGRD.HLP /15JUL94/HELP/OTFUV.HLP /15JUL94/HELP/CONPL.HLP Fix: (1) Get the first file, placing it in $APLPGM. (2) Get the second file, placing it in $QPGNOT. (3) Get the third file, placing it in $QYPGNOT. (4) Get the fourth file, placing it in $APLGEN. (5) Get the fifth file, placing it in $APLUNIX. (6) Get the remaining files, placing them in $HLPFIL. (7) Rebuild them with COMRPL $APLGEN/ZFIO $APLUNIX/ZFI2 COMLNK $QPGNOT/SDGRD $QYPGNOT/CONPL COMLNK $APLPGM/OTFUV Notes: (1) Be sure to apply the DIRCOS, SDGET, and DGETSD patches before doing this "patch". (2) If you apply these patches to earlier releases of AIPS, you will have to remove references to the variable LOCNUM. Comment out the line in SDGRD that sets LOCNUM to 1 and change "(LOCNUM)" and ",LOCNUM" to nothing at all in the several places they occur in DIRCOS and CONPL. Documentor: Eric W. Greisen Date: 1994.09.16 OTFUV corrected 09.23 ------------------------------------------------------------------------ Task: DBCON and single-dish data Summary: This task was forcing the projection code (-SIN) to be filled into the "U" and "V" random parameters. It was also rescaling U and V to move the reference channel to 1. These are good for real interferometer data but bad for single-dish data. Files: /15JUL94/APL/PGM/NOTST/DBCON.FOR Fix: (1) Copy the file to $APGNOT (2) Rebuild it with COMLNK $APGNOT/DBCON Documentor: Eric W. Greisen Date: 1994.09.22 ------------------------------------------------------------------------ Task: All tasks - affects OTFUV most seriously Summary: AIPS disk file open routine assumes that one wants read/write access to the file. This is not always permitted. Changed to detecct this access error and to try read only if exclusive use is not requested. Files: /15JUL94/APL/DEV/UNIX/ZDAOPN.C Fix: (1) Copy the file to $APLUNIX (2) Compile the file with COMRPL $APLUNIX/ZDAOPN (3) Include this upgrade as needed, e.g., COMLNK $APLPGM/OTFUV Documentor: Eric W. Greisen Date: 1994.09.23 ------------------------------------------------------------------------ Task: WFCLN Summary: WFCLN set the number of channel to average in the SUM mode to a wrong and sometimes negative number. This affects the averaging of multi-channel data only, having no effect when BCHAN = ECHAN. This negative number led to a 0 divide and God knows what else. Files: /15JUL94/Q/SUB/OOP/PBUTIL.FOR /15JUL94/Q/PGM/OOP/WFCLN.FOR Fix: (1) Fetch the first file placing it in $QOOP (2) Fetch the second file placing it in $QPGOOP (3) Rebuild with COMRPL $QOOP/PBUTIL COMLNK $QPGOOP/WFCLN Documentor: Eric W. Greisen Date: 1994.09.23 ------------------------------------------------------------------------ Task: SDGRD Summary: Improvements to the speed, eliminating excess IO and computation. New uniform weighting option. Files: /15JUL94/Q/PGM/NOTST/SDGRD.FOR /15JUL94/HELP/SDGRD.HLP /15JUL94/Q/DEV/PSAP/QGRIDA.FOR /15JUL94/Q/DEV/PSAP/QGRD5.FOR /15JUL94/Q/DEV/PSAP/QGRD6.FOR Fix: (1) Fetch the first file placing it in $QPGNOT (2) Fetch the second file, placing it in $HLPFIL (3) Fetch the remaining 3 files, placing them in $QPSAP (4) Compile the new Q routines with COMRPL $QPSAP/QGRIDA $QPSAP/QGRD5 $QPSAP/QGRD6 (5) Link the task with COMLNK $QPGNOT/SDGRD Documentor: Eric W. Greisen Date: 94.10.06 ------------------------------------------------------------------------ Task: Data flagging improvement (not a bug fix) Summary: The data flagging tasks refused previously to use flag tables with single-source data sets. Changed that and added the ability to handle compressed data to UVFLG. Files: /15JUL94/APL/PGM/NOTST/UVFLG.FOR /15JUL94/Y/PGM/NOTST/TVFLG.FOR /15JUL94/Y/PGM/NOTST/SPFLG.FOR /15JUL94/Y/PGM/NOTST/IBLED.FOR /15JUL94/HELP/UVFLG.HLP /15JUL94/HELP/TVFLG.HLP /15JUL94/HELP/SPFLG.HLP /15JUL94/HELP/IBLED.HLP Fix: (1) Fetch the first file, placing it in $APGNOT. (2) Fetch the second, third, and fourth files, placing them in $YPGNOT. (3) Fetch the remaining four files, placing them in $HLPFIL. (4) Build the new capability with COMLNK $APGNOT/UVFLG $YPGNOT/IBLED COMLNK $YPGNOT/SPFLG $YPGNOT/TVFLG Documentor: Eric W. Greisen Date: 94.10.11 ------------------------------------------------------------------------ Task: UVLSF (not bug fix, new capabilities) Summary: Added the ability to shift phases and to flag data based on the fit of a linear spectral baseline to each complex visibility spectrum. These are new capabilities, making this task competitive with, but different from, UVLIN. UVLSF will be featured prominently in the new Chapter 10 of the CookBook (on spectral-line reduction). Files: /15JUL94/APL/PGM/UVLSF.FOR /15JUL94/HELP/UVLSF.HLP Fix: (1) Fetch the first file, placing it in $APLPGM (2) Fetch the SECOND file, placing it in $HLPFIL. (3) Build the new capability with COMLNK $APLPGM/UVLSF Documentor: Eric W. Greisen Date: 94.10.13 ------------------------------------------------------------------------ Task: CALIB, SNPLT Summary: The option to restrain the average gain to be 1.0 had some problems. CALIB did not average over sub-arrays and, instead, wrote the value for the last sub-array into the SN table. CALIB also looped over the numebr of spectral channels when averaging and computing solutions when it should have looped only over the number of IFs being "averaged". SNPLT applied the mean gain with a multiply when a divide is correct. Files: /AIPS/15JUL94/Q/PGM/NOTST/CALIB.FOR /AIPS/15JUL94/Y/PGM/NOTST/SNPLT.FOR Fix: (1) Fetch the two files, placing the first in $QPGNOT and the second in $YPGNOT. (2) Link the two with COMLNK $QPGNOT/CALIB $YPGNOT/SNPLT Documentor: Eric W. Greisen Date: 94.11.03 ------------------------------------------------------------------------ Task: OTFUV Summary: Added the ability to concatenate output files. This will avoid the multiple use of DBCON, but it may make flagging harder. Files: /15JUL94/APL/PGM/OTFUV.FOR /15JUL94/HELP/OTFUV.HLP /15JUL94/APL/SUB/CATINI.FOR Fix: (1) Fetch the first file, placing it in $APLPGM. (2) Fetch the second file, placing it in $HLPFIL. (3) Fetch the third file, plaing it in $APLSUB. (3) Build with COMRPL $APLSUB/CATINI COMLNK $APLPGM/OTFUV Documentor: Eric W. Greisen Date: 94.11.07 ------------------------------------------------------------------------ Patches above broadcast on bananas (11/10 - PPM), below newer than that ------------------------------------------------------------------------ Task: Single dish tasks and UVFLG for single-source files Summary: The data selection routines have used new parameters for some time which were not being set in the single-dish data reading routine and which affected the use of flagging. UVFLG was putting 300 copies of each flag command into the FG table for single-source tables. Files: /15JUL94/APL/SUB/NOTST/SDGET.FOR /15JUL94/APL/PGM/NOTST/UVFLG.FOR Fix: (1) Get the first file, placing it in $APLNOT (2) Get the second file, placing it in $APGNOT (3) Compile and link as: COMRPL $APLNOT/SDGET COMLNK $APGNOT/UVFLG $APGNOT/PRTSD COMLNK $APGNOT/SELSD $APGNOT/SDCAL COMLNK $QPGNOT/SDGRD Documentor: Eric W. Greisen Date: 94.11.13 ------------------------------------------------------------------------ Task: SYSETUP [script for setting up AIPS system areas] Summary: In the mode where you do not specify a particular host to set up, and the script tried to check all hosts in the HOSTS.LIST file, it added far more IC and ID (Image Catalog and TV lock files) than were needed. This fixed version only adds exactly the right number based on the information in HOSTS.LIST and TVDEVS.SH. Update 94.12.20: generalized for either 15JUL94 or 15JAN95 data (C/D format); the earlier version placed here never worked. Files: /15JUL94/SYSTEM/UNIX/SYSETUP Fix: (1) Get the file, place it in $SYSUNIX. Documentor: Patrick P. Murphy Date: 1994.11.14, Updated 1994.12.20 ------------------------------------------------------------------------ Task: OTFUV and SDGRD Summary: The 12m SDD format records the right ascension offsets not in right ascension units but in angles on the sky. Corrected the output RAs by dividing the offsets by the cos(center declination). (ALL PREVIOUS DATA SETS NEED TO BE RECOMPUTED.) Corrected a minor problem in SDGRD where BIF = 2 could cause the output axis increment to be 0 on the frequency axis. Corrected a minor problem in comparing headers for concatenation in OTFUV. Files: /15JUL94/APL/PGM/OTFUV.FOR /15JUL94/Q/PGM/NOTST/SDGRD.FOR Fix: (1) Fetch the first program and place it in $APLPGM. (2) Fetch the second program and place it in $QPGNOT. (3) Rebuild with: COMLNK $APLPGM/OTFUV $QPGNOT/SDGRD Documentor: Eric W. Greisen Date: 94.11.15, 94.11.23 ------------------------------------------------------------------------ Task: OTFUV and SDGRD (enhancements) Summary: To simplify flagging channels known in advance to be bad, added the CHANSEL adverb to OTFUV to specify channels to flag through the full input data set. Changed SDGRD to create a second scratch file only in the unlikely case that the data require sorting. It was requiring the extra disk space and then not using it most of the time. Files: /15JUL94/APL/PGM/OTFUV.FOR /15JUL94/Q/PGM/NOTST/SDGRD.FOR /15JUL94/HELP/OTFUV.HLP Fix: (1) Fetch the first program and place it in $APLPGM. (2) Fetch the second program and place it in $QPGNOT. (3) Fetch the help file and place it in $HLPFIL. (4) Rebuild with: COMLNK $APLPGM/OTFUV $QPGNOT/SDGRD Documentor: Eric W. Greisen Date: 94.12.15 ------------------------------------------------------------------------ Task: Single dish improvements Summary: Changed OTFUV to set the pseudo-baseline "beam" parameter to 257 (baseline 1-1) instead of a highly variable number. Made numerous changes to improve the gridding to the master file in SPFLG, including changes for shorter intervals such as those used in OTF. Files: /15JUL94/APL/PGM/OTFUV.FOR /15JUL94/Y/PGM/NOTST/SPFLG.FOR Fix: (1) Fetch the first program and place it in $APLPGM. (2) Fetch the second program and place it in $YPGNOT. (3) Rebuild with: COMLNK $APLPGM/OTFUV $YPGNOT/SPFLG Documentor: Eric W. Greisen Date: 94.12.23 ------------------------------------------------------------------------