[evlatests] No end-of-scan phase problem!

[Rick Perley]

    I have previously reported on an apparent problem at the end of 
scans, wherein the last two records showed an apparent non-closing phase 
of a few degrees. 

    To reduce the possibility that this claimed effect was due to the 
calibration software, Steve loaded my test data from yesterday into 
CASA, and following a calibration procedure the same as mine, reported 
that he could see no such effect.

    I then used the closure triangle plotting program CLPLT to 
demonstrate that the data showing the apparent end-of-scan spread of 
phases has in fact no unusual closure errors, either in the 'raw' nor 
'calibrated' versions.  (It would be quite surprising to see these two 
version with different closure relations, given that all calibration is 
antenna-based). 
    Furthermore, the closure properties at the ends of the scans are the 
same as the data in the middle of the scans. 

    It would thus appear that the problem noted yesterday (and on days 
prior to this) were somehow generated by the calibration software.  
Further studies are ongoing ...