Busy week with loads of SDSS work

Haven't updated anything for the last week. Not because I was not doing any SDSS related work. But, because I literally had no time to update. Infact, I was only doing SDSS related work. Lots of work done in the last week.

Finally, solved the SoS ELG issue and tagged a new version of IDLSPEC2D/v5_10_2. This will mainly have updates for the SoS ELG program. Main changes are in the apo2d routine. Chased down one of the long standing issues in SoS. The SN2 values in the html page and the plots are different. The problem was chased down to the fact at one place we were using de-reddened magnitudes as compared to the other. The quicktrace routine didn't use the de-reddened magnitudes where as the plotsn routine used the de-reddened magnitudes. Fixed this by changing the quicktrace routine as changing the plotsn will have some implications for the full reduction. But, noticed that the SN2 values are still off by some amount. This was traced to the fact that plotsn was not hacking out the targets with a flag plugmap.objtype='NA' whereas quicktrace did that. It looks obvious to hack out these targets while fitting the SN2. So, implemented that in the plotsn routine. To address the SN2 values differently for ELG plates, wrote two new routines, quickextract_elg.pro and plotsn_elg.pro. The quickextract routine has the scale factor implemented for the R cameras and have hard wired the Blue cameras SN2 to be 4 per exposure. Had to parse a new variable "N_exp" to account for the total SN2 exposures in B cameras. Currently, the scale factor is set at 1.466. The RELEASE_NOTES for v5_10_2 is copied below.
-----------------------------------------------------------------------------------------------------------------------------------------
v5_10_2
VM (2016-09-09) SoS modifications to handle the ELG plates
-Based on keyword 'programname' in plugmap files, ELG plates will be treated differently for SN2-
measurements. Use sky free wavelengths in z-band to measure meansn2.
-Included a scale factor in quickextract_elg.pro and plotsn_elg.pro to tune the ELG depth.
-Included a constant (=0.83 corresponding to the scalefactor) in the equation on SN plots
-Updated opSNlimits.par to include z-band with a slope 0.6
-Updated new gain measurements post 2016 Summer shutdown
-Updated plotsn to exclude objtype='NA'

Adding:
- pro/apo2d/quickextract_elg.pro
- pro/spec2d/plotsn_elg.pro

Updating:
- pro/apo2d/aporeduce.pro
- pro/apo2d/quicktrace.pro
- pro/apo2d/apo_plotsn.pro
- pro/spec2d/sdssproc.pro
- opfiles/opSNlimits.par
-----------------------------------------------------------------------------------------------------------------------------------------


The platedesign of eboss22 results in a narrow observing windows for a lot of plates (19). These are the plates with HA=30. So, HA=30 seems to be a strong criteria for plates at stripe82. Asked Demitri to re-run these plates with HA=20. The resulting HA is slightly less than 27. But, that should be OK. Finally, submitted all the eboss22 plates for drilling.

Made the pixel flats and bad pixel masks using the post summer shutdown data. The pixelflats for B2 looks improved as it is the new camera. Rest looks normal. For the badpixel masks, I used the combinebpm.pro in the $IDLSPEC2D_DIR/specflat directory. I ran different sets with sigma=10 and count=[10,15]. The hardwired bad columns corresponding to bit 16 and 32 are difficult to set for the new B2 camera. Kyle suggested to make the bad pixel mask with out these and we will update the bad pixel mask as and when we notice new bad columns in the data.

That is it for now. Will update when I get time.