[daip] New Ticket - [!XFC-796735]: Problem with AVGTIM in SPLAT?

[Adam T Deller]

New Ticket: Problem with AVGTIM in SPLAT?

Hi guys,

I think there is a problem with AVGTIM in SPLAT.  I have a dataset which I created by DBCON'ing two observations together (that were taken many days apart).  I use IMAGR to locate the peak emission within 4" (this is VLBI data) and then UVFIX to that position.  Now I try to run SPLAT to average in frequency and time to save some space.  At this point SPLAT complains with:

SPLAT1: Task SPLAT  (release of 31DEC12) begins
SPLAT1: Averaging all selected channels in frequency
SPLAT1: Nothing to assemble because the data are single source
SPLAT1: Doing no flagging this time
SPLAT1: Create MJV00159    .SPLAT .   1 (UV)  on disk  1  cno 1521
SPLAT1: AVGTIM: TOO MANY BASELINES FOR BUFFER AT 4 -  9
SPLAT1: SPLCOP: ERROR    1 WRITING OUTPUT FILE
SPLAT1: Destroyed  1 extension files of type NX
SPLAT1: Destroyed UV image file: catno=   1521 disk= 1
SPLAT1: PROBLEM WITH SOURCE:
SPLAT1: FAILED TO SELECT VALID DATA
SPLAT1: Purports to die of UNNATURAL causes
SPLAT1: dop235 31DEC12 TST: Cpu=      0.1  Real=      0  IO=        27
AIPS 1: Resumes

If I turn the time averaging off, everything works fine:

>solint 0
>go
SPLAT1: Task SPLAT  (release of 31DEC12) begins
SPLAT1: Averaging all selected channels in frequency
SPLAT1: Nothing to assemble because the data are single source
SPLAT1: Doing no flagging this time
SPLAT1: Create MJV00159    .SPLAT .   1 (UV)  on disk  1  cno 1521
AIPS 1: Resumes
>SPLAT1:     35103 Visibilities written
SPLAT1: New NX table written
SPLAT1: Updating FREQ in AN tables
SPLAT1: Reformatted AN file from vol/cno/vers  1 6346   1 to  1 1521   1
SPLAT1: Appears to have ended successfully
SPLAT1: dop235 31DEC12 TST: Cpu=      0.1  Real=      0  IO=        59

I noted that there seemed to be some discussion about 5 years ago about whether AVGTIM was overrunning some buffers (see http://listmgr.cv.nrao.edu/pipermail/daip/2007-November/012051.html).  Is this possibly related to another manifestation of a similar problem?  Or is it more likely that the big time gap is causing problems?

I tried writing the data out after spectral averaging only, and then reading it back in, and the SPLAT problem with time averaging persists.  Accordingly, I've attached the uvfits file here.  SPLAT should fail when run with solint = 0.3333.

Let me know if I can provide any more details/info.

Cheers,
Adam



Ticket Details
===================
Ticket ID: XFC-796735
Department: AIPS Data Processing
Priority: Default
Status: Open
Link:  https://help.nrao.edu/staff/index.php?_m=tickets&_a=viewticket&ticketid=1713
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dbconavgtest_uv.fits
Type: image/x-fits
Size: 4737600 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/daip/attachments/20120419/9fd1c01f/attachment.bin>