[daip] Weighting of the UV data

[Eric Greisen]

On 09/27/2017 11:54 AM, Feng Gao wrote:
> Dear Eric,
> 
> Sorry to bother you with more AIPS related questions.
> 
> I wonder how should I assign the weight for each UV point data properly? I looked at some VLBI data and noticed the weight for "Amp" and "Phas" is around 1, and the weight for Real and Imaginary part are 10.
> 
> Right now I calculated the weight as 1/(2*amp_err^2), here the amp_err is the error in the fringe amplitude measurements. It is between 0 and 1, so the weight I got are much larger and didn't show up properly in AIPS. I only see something like ********* in PRTUV. So I'm not sure if my weight calculation is okay.
> 
> Attached is my sample data, for your reference.
> 

The FITS file says that the RA is

3996.252269922912

degrees.  That cannot be what you intended.

PRTUV has options in DPARM if the parameter ends up as ***.
Set DPARM(2)=9999 and you will see that your weights are between 1000 
and 4000.  We use 1/(amp_err^2) although maybe the 1/2 is also
appropriate.  The weights printed by tasks do not change between 
real/imaginary and amp/phase.  The scaling of the display may however be 
misleading.  Your weights suggest an amp error of 0.011 to 0.022.

ERic Greisen