Select galaxies for shear measurement:

Once you have a shape catalogue you need to decide which galaxies to use in your analysis.  With nearly all shape measurement methods this selection is rather ‘black magic’ and your choice usually has a significant impact on your results.  The exception to this is the Lensfit method which outputs a probability distribution of the shear for each galaxy. 

For KSB your selection should be based on the following

1. •Remove all galaxies that are within a mask
   

2. •Remove galaxies which have a size close to the PSF size.  These galaxies are usually hard to completely remove the PSF distortion from.  I find rg>1.15 r_star works well
   

3. •Remove low signal-to-noise galaxies which are hard to measure shapes from.  I find S/N>10 works well
   

4. •Remove really noisy shape measurements.  I find |γ| < 1 works well. 
   

The above selection criteria are used in gal_correct to write out the average shear in the catalogue.

How should you choose your selection criteria?

If you have wide-field data you know that on average <γ> = 0.  This should be true for all galaxy sizes, all magnitudes and all S/N.  gal_select.f90 bins and averages your shape catalogue by size, magnitude and S/N.  You can then identify “problem” galaxies which are biasing your shear measurement.

Should you fit Pgamma?

Many KSB users do fit Pgamma with a polynomial to reduce noise in the final shear catalogue.  My experience is that this fit biases your result, however I include the ability to do this fit with the gal_select -fit 1 option.

The resulting simple final shear catalogue

The output file is very simple

#  1 x          SExtractor x  

#  2 y          SExtractor y

#  3 g1         KSBf90 output reduced shear_1

#  4 g2         KSBf90 output reduced shear_2

Running gal_select.f90 from the command line:

gal_select.a -in shape.cat -param KSBf90.param -fit 0 -nbins 10 -out finalshear.cat

-in : Your SExtractor catalogue with KSB shape parameters appended output by gal_correct

-param:  The name and location of your parameter file

-fit : Do you want to fit Pgamma with a polynomial?  Yes - fit 1,  No -fit 0

-out:  The output of this code; your final selected shear catalogue

Page maintained by Catherine Heymans, IfA, University of Edinburgh