[daip] [!6387]: aips - FITLD does not read UVFITS source tables correctly

[Eskil Varenius]

Eskil Varenius updated #6387
----------------------------

FITLD does not read UVFITS source tables correctly
--------------------------------------------------

           Ticket ID: 6387
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/6387
           Full Name: Eskil Varenius
               Email: varenius at chalmers.se
             Creator: User
          Department: AIPS Data Processing
       Staff (Owner): -- Unassigned --
                Type: Issue
              Status: Open
            Priority: Default
                 SLA: NRAO E2E
      Template Group: Default
             Created: 27 March 2015 12:54 PM
             Updated: 27 March 2015 12:54 PM
                 Due: 31 March 2015 12:54 PM (4d 0h 0m)
      Resolution Due: 06 April 2015 12:54 PM (10d 0h 0m)



Dear support,
Earlier this week I have been able to load a UVFITS file of eMERLIN data with FITLD. This file was written by AIPS last year and contain multiple sources. When I load the file today it looks in LISTR as if all sources are present:  
 1 NGC4418         : 0000         12:26:54.6126 -00:52:39.372  134805
   2 J1232           : 0000         12:32:00.0160 -02:24:04.794  204885
   3 J1256           : 0000         12:56:11.1666 -05:47:21.525   55965
   4 1407+284        : 0000   CAL   14:07:00.3944  28:27:14.690  107850
   5 0319+415        : 0000   CAL   03:19:48.1601  41:30:42.103   53925
   6 1331+305        : 0000   ACAL  13:31:08.2873  30:30:32.959   53805
But, I discovered I can no longer select the source J1232 in any task, as I could earlier this week when I loaded this FITS file. In fact, I cannot select any source except the first one. But, when running e.g. POSSM without any source selection (source = '') I see strong signal visibilities which cannot be for the first target source, they are from the calibrators. So, the data is there, but somehow is not read properly. I tried re-running INDXR on the data, assuming something had changed and the old NX table had to be remade. After creating a new NX table I get this from LISTR:
   1 NGC4418         : 0000         12:26:54.6126 -00:52:39.372  611235
   2 J1232           : 0000         12:32:00.0160 -02:24:04.794       0
   3 J1256           : 0000         12:56:11.1666 -05:47:21.525       0
   4 1407+284        : 0000   CAL   14:07:00.3944  28:27:14.690       0
   5 0319+415        : 0000   CAL   03:19:48.1601  41:30:42.103       0
   6 1331+305        : 0000   ACAL  13:31:08.2873  30:30:32.959       0
i.e. INDXR thinks all visibilities are for source no 1. So, it seems to me that FITLD stores all visibilities as source number 1. If this is an error when FITLD reads the data, when it writes it into AIPS, or some fundamental error when the other tasks e.g. INDXR tries to access the data, I cannot say. I don't know what more info to give, but in case you want me to check some specific thing I'd be happy to do so. 
Kind regards,
Eskil

------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff