[daip] AIPS questions

[Eric Greisen]

Bill Cotton wrote:
> Eric,
> 
>    What is your current usage of the NOPCAL keyword in the AN tables?  
> It used to be the number of polarization calibration parameters and 
> the code still describes it that way but it now seems to be something 
> else.  When I fiinally got AIPS/PCAL to run on my data it wrote 2 in 
> this keyword rather than 32 which caused VERY BAD things to happen.
>    Also, Rick is spreading rumors that you're working on finer poln 
> calibration solutions than per IF.  Any hints about how you're going to 
> implement this?
> 
> Thanks,
> Bill

NOPCAL is the number of constants per IF and the number of IFs is given 
to the ANTINI in a separate argument.  They were combined in the old 
days which meant that several tasks had to guess at the number of IFs
and the new column for FWHM did not have a defined dimension.  So I 
separated the two like in most tables.

I am working on a version of PCAL that will do either what it was 
supposed to do for the average of spectral channels on an IF basis
(including ICHANSEL, spectral index on PMODEL, allowing source 
polarization to be fixed, etc) or do a channel-by-channel solution
(with an additional spectral smoothing which respects ICHANSEL).
The latter will write its answers into a PD table (much like BP table 
but with terms like now go into the AN table) and also a PS table if it 
solves for source polarization as well.  These new tables will be 
plotable by POSSM and BPLOT and the PD table will be the first info 
sought when DOPOL is true.

This process is slow since PCAL had some very flimsy logic which I have 
finally just about fixed and the line stuff will take time to get right.
The application routines will also be slow to complete.

Eric