[daip] Re: CALIB issues

[Eric Greisen]

The counting of failed solutions is complicated in that there are many
reasons for failure - missing data being one of them.  In July 2005, I
changed CALIB (version 31DEC05) in an attempt to differentiate between
the no data case and true failures.  By no data however one means no
data for any antenna at that IF and polarization.  The solutions that
have apparent SNR less than SNRMIN are counted as too few not failed.

When you say you tried it with 31DEC05 you must be careful to say
whether it was a development version and if so to what date that
version is complete.  You could look at the last date in the
$AIPS_ROOT/31DEC05/HIST/CHANGE.DOC to determine that.

The logic behind this counting is difficult and I believe that the
people who coded it did not expect this to be some detailed exact
count.  It should be better now than it was but I cannot invest time
looking at it without more knowledge of what date your version was
current to.

btw - please use daip at nrao.edu to address your queries.

Eric Greisen