[daip] Flagging with VPLOT

[Eric Greisen]

Andrew Biggs wrote:
> Hi there. I'm trying to flag some GMRT data using VPLOT, but this is failing with the following error message:
> 
> localh> VPLOT1: Task VPLOT  (release of 31DEC11) begins
> localh> VPLOT1: Copied FG file from vol/cno/vers  2    3   5 to  2    3   6
> localh> VPLOT1: Averaging from channel    1- 128
> localh> VPLOT1: ZOPEN: LUN = 30 ALREADY OPENED IN FTAB
> localh> VPLOT1: TABINI: I/O ERROR FROM ISTAB ON AN
> localh> VPLOT1: UVGET: SUBARRAY  1 ERROR   3 OPENING AN TABLE
> localh> VPLOT1: SCALVB: ERROR  3 INIT VIS FILE
> localh> VPLOT1: Appears to have ended successfully
> localh> VPLOT1: pc016987 31DEC11 TST: Cpu=      0.8  Real=      2  IO=       173
> 
> I only get this if I set APARM(1) = 1 - otherwise VPLOT works very well. I tried restarting AIPS, but this didn't help. I've attached my inputs below - sorry for the bad formatting.
> 
> By the way, it seems that the ability to flag data points per baseline/IF/pol based on the deviation of a single data point from a nearby average in sigma units is only offered by VPLOT. Is this correct? FLAGR offers similar functionality, but not this specific one which I personally would find very useful.

The error arose from using LUN 30 (was 27 but that was changed) which 
collides when DOCAL et al are true or when flagging the data as they are 
read.  Changed the LUNs for this option.

Yes in some ways this option is unique.  RFLAG and TVFLG and SPFLG have 
very similar options but the clip levels are in Jy not in sigmas.  Note 
that these use real and imaginaries not amplitudes which is probably 
better (Gaussian noise not Ricean).  RFLAG started out clipping in 
sigmas but was changed since things with low sigma would lead to 
flagging of good data.  (It has an option to flag things with sigmas 
that are way too low - i.e. data that are not zero but are oddly low in 
amplitude.)

Thanks

Eric Greisen