[daip] AFARS problem BLC, TRC

[Eric Greisen]

Lawrence Rudnick wrote:
>   Eric & Leonia -  in 31DEC12, patched as of yesterday on my mac, there 
> is a big problem that has apparently been inserted into the BLC, TRC for 
> AFARS.
> 
> I am currently uploading FARSAM and FARSPH cubes to ftp area 
> (pub/staff/lrudnick) so you can run this yourself.  Here is the history 
> from running AFARS three times with three different sets of BLC,TRC (on 
> the RM axis).  The three maps are very different from each other, 
> because there has apparently been a constant offset put into some of the 
> RMs, while some of them are right.  Here is an IMVIM between two of the 
> pairs of maps, on an equal scale, so you can see. (I've only included 
> points with strong signals, so there should not be any effect from 
> including more or less of the RM spectrum).
>

I am not sure I understand what the task was doing since it seemed to 
get the right answer to x,y = 145,77 and then to be offset by BLC(1)
for the rest of the image.  I did find two places where the BLC(1) 
correction was not made.  I also found that it would set the peak pixel 
to 1.0 when the entire row was zero.  I changed that to have it take the 
blanked value instead.  It now seems to do the same on a SUBIM of the 
images or on a BLC,TRC in AFARS.

A MNJ now in 31DEC12 will fix it.

Eric