[evlatests] WIDAR bronze dataset - C band dual pol.

[Michael Rupen]

>>  As I mentioned to Frazer, one might do better by deriving rms from the
>>  scatter over channels than over time, for WIDAR data.  Assuming no lines,
>>  ignoring RFI, etc. :}

>    I don't that will work unless there are no sources.

Why?  Remove the delay and the bandpass, do a linear fit in frequency a la
UVLSF/UVLIN to take care of the source over a subband (maybe after subtracting 
a model, as Ed suggests), calculate the rms, possibly do an antenna-based
decomposition, and use the results to figure out the weights.  You have
to account for source structure (and spectral index as a secondary effect)
of course.

Admittedly if we have good Tcal(freq) we should be able to get good
Tsys(freq) and that's much easier.

                 Michael