[daip] Re: CLIPM and zero-UV points

[Eric Greisen]

Thank you very much for your help with this problem to date.

Looking at your data I found:

1. UVCOP did not copy FG records correctly - fixing FGSEL

2. CLIPM did not handle the PFLAGS correctly when the data were only
   LL polarization - fixing CLIPM.

3. The AIPS calibration package did not warn the user when it was
   trying to calibrate a multi-source data set that did not have an NX
   table.  In this case, the internals think of the source number as 0
   and when a flag record said some other source number it previously
   thought that the flag record did not apply.  I have changed it to 
   ignore the source ID on a flag record when the internal source
   number is zero and also to ignore the user specified list of
   sources in this case.  And I have added a warning since BP cal
   requires correct source info for the VLBA and Pol cal also needs
   some source info.  Ignoring the source info in data selection and
   flagging is also often wrong.  The flagging worked fine before all
   these fixes after INDXR was run.


Now - the fact that your data have a high flux as the least
significant bit suggests that you must back up to FITLD and load the
data uncompressed.  Find the bad channels (very high fluxes), erase
them with UVFLG and UVCOP, and then procede with real data not values
from = 0,1,2, or 3 milliJy.  These may be compressed to save space.

Thanks again,

Eric