[daip] [!7296]: AIPS - OMFIT and strange chi2 values

[Eric Greisen]

Eric Greisen updated #7296
--------------------------

       Staff (Owner): Eric Greisen (was: -- Unassigned --)
              Status: Closed (was: Response Overdue)
                 Due: - Cleared - (was: 29 September 2015 10:52 PM)

OMFIT and strange chi2 values
-----------------------------

           Ticket ID: 7296
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/7296
           Full Name: Michael Bietenholz
               Email: mbieten at yorku.ca
             Creator: User
          Department: AIPS Data Reduction
       Staff (Owner): Eric Greisen
                Type: Issue
              Status: Closed
            Priority: Default
      Template Group: Default
             Created: 22 September 2015 08:35 PM
             Updated: 26 September 2015 09:08 PM



I have changed the help on NOISE(1) to say that it is any
additional correction to the weights and that 1.0 is the
default (now actually so) and is correct for weights that
are already 1/sigma^2 in 1/Jy^2.  I changed OMFIT.FOR
to make NOISE(1)<= 0 ->1.0 (not -1 as before) so the post-fit
rms will not enter any of the computations.  Then the CHI^2
"true" seems about right and the Ketan one is off by a factor
of the average weight.

I also found a place that tried to write 3 records into one string.
An old bug from previous overhauls of non-standard code.

Note that compiled with gfortran 5.1.0 in debug or optimize, there
has been no issue with dynamic memory.  Perhaps I need to look to
see if you have given me a test problem for that case.

Eric Greisen




------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff