[daip] more on APCAL

[Jim Ulvestad]

Craig Walker wrote:
> 
> After talking to Eric a bit about my last message on APCAL, I took a
> closer look at my TY tables.  There are some rather curious
> characteristics to the one from calibration transfer, which is not the
> one I used for APCAL.  The data set (FQ table) only contains one FQ
> ID.  But the transfered TY table has FQID's 1, 2, and 3.  There were
> three frequencies observed, but I selected one in FITLD.  The TY table
> that I actually used in APCAL (TYVER=2) is one I loaded from scratch
> with ANTAB (easiest way to edit some bad Tsys's from EB and I needed to
> load VLA Tsys's).  It only has the one FQID.
> 
> Craig

I've never tried using cal transfer and FITLD together in this way.
I would have been surprised if it worked!

What I have done is run FITLD with all the frequencies loaded, then
UVCOP to select out a FQID.  This almost always works, although the
VLBA changes sources and bands fast enough that there is occasionally
some ambiguity about which FQID to associate a Tsys measurement with,
and you can get a bad value leaking through for this reason.

Yes, this takes more time and disk space, and isn't elegant.  But
I suspect that making FITLD fix this might take a fair amount of
effort.

ju