[AIPS] Patches to AIPS 15APR98 [NRAO]

This page most recently updated on $Date: 2001/12/28 20:53:39 $ (UT)

You will find patches to known bugs in 15APR98 here.

The following is a list of pointers to the different patches in this document.

  1. FITLD makes incorrect source tables and numbers. 1998-04-19 and 1998-04-23
  2. TEKSRV error for SGI and DEC computers. 1998-04-23
  3. DBCON reweights the wrong thing. 1998-04-28
  4. BPASS handles flagged channels and interpolation wrongly. 1998-05-05, corrected 1998-06-18
  5. IMAGR can get wrong frequency in output image. 1998-05-05, corrected 1998-05-06
  6. Digital Unix version 4.0D tape problems. 1998-06-05
  7. Y2K addition to FITS table readers. 1998-06-05
  8. DBCON has trouble with multiple sources. 1998-07-05
  9. FITLD labels LSB data incorrectly. 1998-07-06 and 1998-07-21 and 1998-08-18
  10. FTPGET for SUL gets library name wrong. 1998-07-29
  11. M3TAR handles double-precisions wrongly.. 1998-08-18

The remainder of this document is comprised of the patch descriptions pointed to by the above references.

Task:
FITLD
Summary:
FITLD was making source tables with multiple occurrences of essentially the same source and with multiple occurrences of the same source number attached to different sources. The VLBA Correlator treats the same source, with different CALCode as a different source. This causes AIPS to have things in the source table that it cannot properly address and is simply due to the VLBA antennas passing on CALCodes while non-VLBA ones do not. Changed FITLD to allow for source renumbering even while reading the first input file and changed the list handling to include current additions to the source list when checking the next input source. Dropped a test on source number when converting SO to SU files - these were supposed to be renumbered! The logic was backwards in handling the source number translation, complicating the process and forbidding any case in which 2 or more source numbers would both be translated into the same output source number. This arises: file 1 had a cal as source 1, 3C120 as sources 2 and 3 (w different CALCode). File 2 has 3C120 as sources 1 and 3 with the cal as sources 2 and 4, so both 1 and 3 have to become 2 and 2 and 4 become 1 on output. Also fixed up the formatting of times.
Added code to prevent CHTRIM writing beyond the end of the output string if the trimmed input string is longer than the output variable. This was causing failures in FITLD. The string will now be truncated if it doesn't fit the output variable.
A call to DFILL in GETDEL was in error and a buffer in FITLD was too small. These cause data over-write or bad address errors on some systems.
Data are in error:
UV files written by 15APR98 FITLD from January 27, 1998 until this patch are likely to have serious errors. Without the test on CALcode, a second file coming in might have two source numbers which should become one on output. Since this was not allowed, one of the two would pass through unchanged (BUT WRONG).
Files:
/15APR98/APL/PGM/NOTST/FITLD.FOR
/15APR98/APL/SUB/CHTRIM.FOR
/15APR98/APL/SUB/NOTST/GETDEL.FOR
Fix:
  1. Fetch the files listed above and place them in the corresponding areas on your system.
  2. COMRPL $APLSUB/CHTRIM
  3. COMRPL $APLNOT/GETDEL
  4. COMLNK $APGNOT/FITLD
Documentor:
Eric Greisen
Date:
1998-04-19 updated 1998-04-23
Task:
TEKSRV
Summary:
TEKSRV was modified to handle the Tektronix image catalogs, much as XAS handles the TV image catalog. Unfortunately, the $APLDEC and $APLSGI versions of the relevant routine were not modified. This error affects only DEC and SGI computers.
Files:
/15APR98/APL/DEV/UNIX/BELL/SGI/ZTKSRV.C
/15APR98/APL/DEV/UNIX/BERK/DEC/ZTKSRV.C
Fix:
  1. Fetch the files listed above and place them in the corresponding areas on your system.
  2. COMRPL $APLGEN/ZTKSRV
  3. COMLNK $AIPNOT/TEKSRV
Documentor:
Eric Greisen
Date:
1998-04-23
Task:
DBCON
Summary:
DBCON had an off-by-one error causing it to re-weight the imaginary part of the visibilities rather than the weights when the two data sets are in the same sort order. See also below.
Files:
/15APR98/APL/PGM/NOTST/DBCON.FOR
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMLNK $APGNOT/DBCON
Documentor:
Eric Greisen
Date:
1998-04-28.
Task:
BPASS
Summary:
BPASS was found to have several errors in handling flagged channels and the interpolation option to deal with them.

  1. The worst was an addressing error which meant that the address for weights did not get incremented when a channel was flagged. Then the weights for higher numbered channels were put in for lower numbered ones. This could lead to channels being eliminated, but did not seem to cause flagged channels to acquire false solutions. The excess channel elimination did lead to a message, so users should have noticed.

  2. The fit routine did not return an error code for no valid samples so the task did not know to invoke interpolation when a channel was fully flagged.

  3. The interpolation method used the wrong formula for weighting channels so that the interpolation was not quite right and it would fail to extrapolate a solution if the last couple of channels or more were failed.

  4. The corrected interpolation was backwards until June 18!

Files:
/15APR98/Q/PGM/NOTST/BPASS.FOR
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMLNK $QPGNOT/BPASS
Documentor:
Eric Greisen
Date:
1998-05-05 corrected 1998-06-18.
Task:
IMAGR
Summary:
The OOP utility routines had the usual OOP sort of error, this time in handling the reference pixel for spectral-line imaging. A low-level routine partly corrected for channel selection and averaging unbeknownst to a higher level routine which did the correction again, fully this time. If BCHAN and the frequency reference pixel in the input data set were both 1 (a common occurence) or if NPOINTS was 1 (no channel averaging) or large (all channels "averaged") then no error was made.
Files:
/15APR98/APL/SUB/OOP/UVUTIL.FOR
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMRPL $APLOOP/UVUTIL
  3. COMLNK $QPGOOP/IMAGR $QPGOOP/WFCLN
Documentor:
Eric Greisen
Date:
1998-05-05 Corrected 1998-05-06.
Task:
Digital Unix version 4.0D tape tasks
Summary:
The new release of the Digital Unix operating system version 4.0D generates a serious error condition when a back-file operation encounters the beginning-of-tape. Code to handle this condition has been patched into the basic tape manipulation subroutine. Apply this patch only if you have a DEC ALPHA computer running OSF/1 4.0D (look for "vmunix: Digital UNIX V4.0D" in /var/adm/messages).
Files:
/15APR98/APL/DEV/UNIX/BERK/DEC/ZTAP2.C
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMRPL $APLUNIX/ZTAP2
  3. Apply the next patch also, and then
  4. COMLNK $AIPPGM/AIPS
  5. COMLNK $AIPNOT/TPMON
  6. COMLNK $AIPNOT/GRITP
  7. COMLNK $AIPNOT/GR2TEX
  8. COMLNK $APLPGM/PRTTP
  9. COMLNK $APLPGM/AVTP
  10. COMLNK $APLPGM/TCOPY
  11. COMLNK $APGNOT/BAKLD
  12. COMLNK $APGNOT/BAKTP
  13. COMLNK $APGNOT/FILLM
  14. COMLNK $APGNOT/FILLR
  15. COMLNK $APGNOT/FITLD
  16. COMLNK $APGNOT/FITTP
  17. COMLNK $APGNOT/GSCAT
  18. COMLNK $APGNOT/IMLOD
  19. COMLNK $APGNOT/MK3IN
  20. COMLNK $APGNOT/MK3TX
  21. COMLNK $APGNOT/UVLOD
  22. COMLNK $APGNOT/M3TAR
Documentor:
Eric Greisen
Date:
1998-06-05.
Task:
FITLD, UVLOD, IMLOD
Summary:
Some FITS writers including AIPS and the VLBA correlator will begin writing dates as 'YYYY-MM-DD' rather than 'DD/MM/YY' beginning on 1 January 1999. 15APR98 installations may wish to prepare to receive such tapes next year. This patch should be enough. The TABHDR patch also disables "gain transfer" from the VLBA correlator which is not working properly.
Files:
/15APR98/APL/SUB/NOTST/TABHDR.FOR
/15APR98/APL/SUB/NOTST/GETKEY.FOR
Fix:
  1. Fetch the files listed above and place them in the corresponding area on your system.
  2. COMRPL $APLNOT/TABHDR $APLNOT/GETKEY
  3. Apply the previous patch if relevant, and then
  4. COMLNK $APGNOT/FITLD
  5. COMLNK $APGNOT/GSCAT
  6. COMLNK $APGNOT/IMLOD
  7. COMLNK $APGNOT/UVLOD
Documentor:
Eric Greisen
Date:
1998-06-05.
Task:
DBCON
Summary:
DBCON had an array addressing problem that caused it, when merging multi-source files, to assign new source numbers and names to sources that should have matched. See also above.
Files:
/15APR98/APL/PGM/NOTST/DBCON.FOR
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMLNK $APGNOT/DBCON
Documentor:
Eric Greisen
Date:
1998-07-06.
Task:
FITLD
Summary:
FITLD recalculated the reference frequency incorrectly for lower sideband VLBI data. This caused lower sideband data to be labelled incorrectly.
Files:
/15APR98/APL/PGM/NOTST/FITLD.FOR
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMLNK $APGNOT/FITLD
Documentor:
Chris Flatters
Date:
1998-07-06 corrected 1998-07-21 and 1998-08-18
Script:
FTPGET
Summary:
FTPGET incorrectly guesses APLSUL as the name of the architecture-specific object library for the SUL architecture. This is wrong; it should guess APLSOL.
File:
/15APR98/SYSTEM/UNIX/INSTALL/FTPGET
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system, preferably before running the FTPGET part of INSTEP1.
  2. Continue INSTEP1.
  3. Or: retrieve the APLSOL/SUBLIB library yourself (shift-click or download; that is a binary file!) yourself and then run the unfixed FTPGET or INSTEP1.
Documentor:
Pat Murphy
Date:
1998-07-29
Task:
M3TAR
Summary:
M3TAR used a simple copy for double precision IEEE floating-point numbers. A call to ZR64RL is required. Also it changed the values of constants such as 2 and 3 or tried to, depending on compiler.
Files:
/15APR98/APL/PGM/NOTST/M3TAR.FOR
Fix:
  1. Fetch the file listed above and place it in the corresponding area on your system.
  2. COMLNK $APGNOT/M3TAR
Documentor:
Eric Greisen
Date:
1998-08-18.
Patrick P. Murphy / comments