[daip] SU table after UVAVG ?

[Eric Greisen]

Nissim Kanekar wrote:
> 
> Hi Eric,
> 
> I'm finding a strange problem in one of my EVLA datasets after running 
> UVAVG to smooth the UV data to 5-second resolution. Essentially, LISTR 
> (run with inext 'SU', opty 'SCAN') seems to show each record of the 
> .UVAVG file as a scan, thus producing a large number of entries per 
> source. This seems to only affect one of my datasets (of project AK724); 
> my other datasets, including others from the same project, are fine. 
> I've attached two examples of LISTR output; LISTR.OUT.1 is from the 
> original UV dataset, and LISTR.OUT.2 is from the UVAVG output. Do you 
> happen to know what's up?
> I was especially worried that tasks like BPASS and CALIB might treat 
> each 5-second record as a scan!
> 
> Thanks,
> Nissim
> 

This is strange - the automatic NX table writing routine must have seen 
some data with < 5 sec between samples and so thought that 5 sec was a 
real break.  Perhaps this comes from casa's insitence in have an int 
time random parameter.  I will look at it.

EXTDEST the offending NX table and use INDXR or UVCOP to make a new one.

It is wise to be cautious these days...

Eric Greisen