[daip] problems with flag table

[Eric Greisen]

Jacqueline van Gorkom writes:
 > 
 > Eric I am mystefied by that. Today I repeated the whole exercise and
 > again when I do clip, with aparm 2.5, 0, 0.5, 0 and then a uvcop,
 > it does not seem to do anything useful. If I try to make a uvplt
 > of the output file it tells me to again run UVCOP because there
 > are too many flags in the table.
 > The only manner in which I can make it work, is to set aparm(9) and
 > thus flag whole spectra in a given IF. This works, the bad data are gone,
 > but unfortunately we lose too many data that way.
 > 
 > Since it works for you, I am now worried about me making mistakes.
 > I will look at this in the coming days and get back to you.

I asked two specific questions.  First, did you compile your aips
locally or take a binary distribution?  If it is the former, then what
version of what compiler did you use?  This is a very basic question
since some compiler versions produce code that does not work
correctly.

Second, did you EXTDEST any flag file versions?  In the UVCOP print
out you sent me, you said that you set FLAGVER to 3.  Without an
EXTDEST, it should then copy FG table versions 1 and 2 but it did not
report having done so.  I ask because users do not usually bother with
EXTDEST now that disk is plentiful.

When you run UVCOP you must set FLAGVER to the desired version.  It
will then not copy that version.  It will copy any other FG table
versions you might have.  If UVPLT now is complaining about too
many flags at a specific time in time-ordered data, then that large FG
table must have been copied not used.

Eric Greisen