Summary of chop-nod analysis telecon of Dec 10, 2008

Mike, Giles, Lero, Tristan, Martin

_____________________________________________________________

Mike's analysis of NGC 1333 IRAS 4:

Mike has addressed several concerns raised by John and Giles previously:

He made a change to his chi2 program: Now, when the -update flag is used, 
the errors are only changed when the reduced chi squared is greater than 
one.  Otherwise they are left unaltered.

(The new chi2 is now installed on zamin but not puuoo.)

Mike reran chi2 on his data, and then reran polsharp on the resulting 
inflated-error file, as before.  The result is that only two vectors are 
lost, compared to the previous map.  See new map sent to sharpsoftware 
list today.  In this map the -eff flag is used and the -onep flag is used 
to draw circles for points with 2-sigma upper limits below 1%.

To address concerns about the sharp_integ flags he used, Mike re-ran 
sharpinteg on all ~110 files, this time without the -nc flag, without the 
-w flag, and with the -em flag.  He ran chi2 on this stuff, and the chi2 
results were roughly the same.  From this he judged that his use of the -w 
flag is probably not too harmful.  Regarding the -nc flag, he found that 
only four files were rejected due to chop/nod issues but the worst 
discrepancy was 4 arcseconds.  So his inclusion of these marginally 
problematic files is not a problem.

We agreed that since Mike looked by eye at all his I-maps, his omission of 
the -em flag is not a serious problem.  The I-map spikes that are removed 
by the -em flag would have been obvious.  Note that the 8 files Mike 
discarded may in fact be good data, if processed with the -em flag.  In 
fact, his third posting on the analysis logbook suggests this.  But still, 
we decided not to redo the analysis, since its a lot of work for a small 
amount of data.

________________________________________________________________

Other issues:

There was a brief discussion of the super-bin/sub-bin idea.  Martin will 
study John's E-mail on this.

Giles reviewed work by Mi Zou ("Mimi") on H-V differences measured in the 
sky.  Mimi finds that for the Sept. 2008 run, the H-V misalignment can be 
measured to an rms accuracy of 1/20 of a pixel using sky data.  (The 
misalignment she finds is about a third of a pixel.)  We discussed 
conventions that should be used in discussing these misalignment values.  
We decided that for this purpose GAP likes H-V, not V-H, and that GAP 
likes sharpinteg map coordinates, not sky right coordinates.

________________________________________________________________