[daip] dbcon weirdness

[Eric Greisen]

On 08/06/2015 08:38 AM, bob at physics.ucdavis.edu wrote:
> hi, dbcon does weird stuff if the # of subarrays goes over 20.
> here is an example.  1st a file with 20 subarrays with imhead and listr

What are you doing?  Please note that 20 subarrays raises the time over 
100 which makes the accuracy rather minimal.  Is there a need to make 
the various bits of data separate subarrays (i.e. their antenna files 
are different) or can some be combined together as single arrays?

I have found a limitation in the index table routines to 20 subarrays.
I can fix that - it is not too expensive in memory although it will 
force a vast recompile/relink.

But you should rethink your strategy too - times especially cause issues 
but some code only does 1 subarray and almost all code does only 1 
subarray at a time even if it can loop internally over subarrays.

ERic Greisen