[daip] a few AIPS questions

[Eric Greisen]

Ashley Zauderer wrote:
> Dear Eric,
> I hope you are doing well!  I've been staying busy, but have managed to 
> find a nearby yoga studio which has helped to keep
> me from being too stressed.  Our paper on the tidal disruption event is 
> coming out this week, but I have to say that the
> best part of my three months in NM were yoga, basketball and the coffee 
> shop in the square.  :)
> 
> A student here, Tanmoy, is working on a VLBA project, looking at 
> variability of masers in a star forming region.
> We had a few questions come up that I think you could answer more 
> quickly and reliably than anybody.
> 
> (1)  JMFIT
> How are the error bars calculated for the fitted position?  
> 
> (2)  UVFIT vs. JMFIT
> What can you tell me/warn me about in terms of fitting in the uv-plane 
> instead of in the image plane?
> Is it ok to fit multiple sources in a plane/channel in the uv-domain?  
> 
> Looking for time variability of the masers, what is your advice in 
> fitting the flux, shape, size, position of the sources?
> Would you do it in the uv-plane or in the image-plane?  Why?  (The past 
> papers describe fitting in the uv-domain, but
> Tanmoy was getting some answers that did not quite make sense).

That does not surprise me - I have never believed in UVFIT and it is not 
used much (or at least I do not get questions about it).  I guess if one 
had a pretty good model from the image plane one could try to refine it
in the UV plane but it is tricky.  In the UV plane all points in the SD 
beam contribute to each vis which means one must model all emission at 
the particular frequency.

The errors in JMFIT are based on the image rms and follow a paper by 
Condon on such errors.  See the explain of JMFIT and (better) SAD for 
details and reference.

Eric