6/13/2008
----------

Orig files: data/des_simlns/sepchips/chip27/

************************* Process:

- Took chip27 (size: 4104x2096 pixels) out of the full 62 chip set,
  ran sextractor over it using the decam.sex file (copy of this
  included in this dir, and it's simply a straight copy of the Subaru
  .sex telescope file we had)

- decam27.fwhmVSsmag.ps is the standard fwhm vs. magnitude plot -- we start with  about 450 obj's 

Issues: - the mags are very bright compared to normal, which normally
   fall from 18-25 in most exposures i've seen, but all i did is use
   the zeropoint 26.734317 listed in the fits header file, whereas
   when Julia uses the wrong zeropoint from the LBC she gets more
   reasonable range of mags, so i'm not sure what the deal is here

  - the star column is clear and prominent, *but* the objects in it
    don't have star class mostly > 0.9, but instead scattered all
    over, and i show this in: stellarityclass.starcolumn.ps  

    this is probably because the neural network classifier didn't
    work, because some of the parameters in the .sex file weren't
    optimal, i guess.. but i've never seen it be as scattered as this.

   - there are a large number of objects smaller than the stars, and
     bright -- these are apparently the cosmic rays, because the DES
     CCD's are 5-10 times thicker than normal ccd's, to be more
     red-sensitive.  usually objects smaller than stars aren't used in
     the shear analysis, so i discarded all these for the moment.

- The selection i initially used was:


stars:  fwhm > 4 , fwhm < 5.2 ,  smag < 15.5, smag > 11
gals:   fwhm > 5.2            ,  smag < 17, smag > 13 

115 stars, 109 gals, after this step.


- stars.noavg.ps is a straight plot of the sextractor ellipticities of the stars (made with simplestarplot.pro)

- The files stars.Npixelcells.ps are plots of the  sextractor ellipticities of the stars avg'd in cells of size NxN pixels.
(made with simplestaravgplot.pro)

- stars.psfgrid.ps is a plot of the psf from the 3rd order polynomial interpolation of the stars above. (used makegrid.pro, then gen2dpolynomial)


I did do the psf correction (psfc33) on the galaxies and then avg'd the
ellipticities of them afterwards, i find: (0.02, 0.002) for the shear
avg, but only 57 objs in the end.

Clearly, this is pretty far from the expected (.05, .05), but there are 
several issues we already have above, so i'm not so surprised by this.. (used awk code for this)



- stars.magLT14.psfgrid.ps: i then did a stricter star selection, then interpolated the psf, and it does look a little
smoother, as you can see in this file 

- Now i did a stars on stars test, and these files show the results:

---- 2D sextractor ellip dist:

stars.es.ps

---- Histos of the 2 ellips:

es1.hist.ps
es0.hist.ps


---- 2D post-deconv ellip dist, same stars:

stars.e.ps

---- Histos of the 2 ellips:

e1.hist.ps
e0.hist.ps


----- And numerical values of the means and standard deviations of the
      sextractor and post-deconv ellips:

esANDe.stars.avgANDsd