[daip] USUBA issue.

[Eric Greisen]

Craig Walker wrote:
> I have loaded the geodesy test (TG009) data to AIPS.  This was a 
> piggyback run with the VLBA stations recording MARK5C and MARK5A data 
> and three geodesy stations running MARK5B.  The correlator delivered 1 
> idifits file for the MARK5C/MARK5B and 3 (tried for 4 but one won't load 
> for lack of a CH table - an issue, but not the subject of this message) 
> for the MARK5A/MARK5B.  There are subarrays in the schedule.  It seems 
> that the multiple MARK5A files were to prevent subarrays in the idifits 
> files.  The AIPS versions of those data sets have NX tables.  But the 
> MARK5C file does not have an NX table or a CL table.  FITLD warned that 
> I would need to run USUBA and INDXR.  When I try to run USUBA after 
> doing a DEFAULT USUBA, IND 1, GETN 74 and setting OPCODE to 'SCAN' or 
> 'AUTO', I get:
> 
>  >go usuba
> USUBA1: Task USUBA  (release of 31DEC13) begins
> USUBA1: You are using a non-standard program
> USUBA1: CANNOT YET DEAL WITH INHOMOGENEOUS AN TABLES
> USUBA1: Purports to die of UNNATURAL causes
> USUBA1: noatak       31DEC13 TST: Cpu=       0.0  Real=       0
> AIPS 1: Resumes
>  >
> 
> There is only one AN table, and I don't know why it might be considered 
> inhomogeneous.  I looked at it with PRTAB and PRTAN and don't see 
> significant looking differences between, say, the VLBA and foreign stations.
> 
> Any idea what's going on?
> 
> This on is on noatak disk 1 so can be reached from other work machines.

Your antenna file had no entries for antenna numbers 2 and 5 so USUBA 
thought they were the same antenna (named = ' ').  Fixed it to name 
missing antennas ANT SSNN which will be unique and safe.  It now seems 
to work and - for you - tomorrow after the MNJ.

Eric Greisen

I cannot find any CH file anywhere so I need more info on that....