esheldon/galsim_extra

A package of modules that can be loaded in galsim config files

galsim_extra

A package of modules that can be loaded in galsim config files using the modules option.

To use these in a GalSim configuration file you will need to do the following:

1. Download or clone this repo. e.g.

   git clone git@github.com:esheldon/galsim_extra.git
   

2. Type python setup.py install (possibly with sudo or --prefix=~ or similar as you prefer)

3. Add the following somewhere in your configuration file

   modules:
       - galsim_extra
   

4. Use any of the galsim_extra module types in your config file.

Notable modules available in this repo

• cosmos_sampler is an input type that reads in the real_galaxy_25.2_fits.fits COSMOS catalog shipped with GalSim and calculates the joint PDF for flux and sizes. Then you can sample from that with value types CosmosR50 and CosmosFlux which will produce appropriately correlated values for your galaxy population. cf. examples/focal.yaml

• FocalPlane is an output type that lets you have parameters that are randomly generated once per focal plane to be used consistently across multiple CCD images. It adds a top-level field called meta_params, which generates values that can be used in eval statements elsewhere in the config file. e.g. PSF parameters that set the overall PSF pattern, but which change randomly each exposure.

  It also automatically calculates some parameters that may be helpful about the overall focal plane geometry. e.g. fov_minra, fov_maxra, pointing_ra, pointing_dec, etc.

  Finally, it adds exp_num as an additional index_key that you can use to set values to only update each new exposure, rather than each file or image. cf. examples/focal.yaml

• MixedScene is a stamp type that lets you have several different kinds of objects in your scene with different probabilities. e.g. stars, bright galaxies, faint galaxies, etc. Each object gets its own top-level field, which can be named anything you like, to replace the normal single gal field. (You can still have gal as one of your fields if you like.) cf. examples/focal.yaml

• OffChip is a bool value type that checks if an object is (significantly) off the image currently being worked on. Useful in conjunction with the FocalPlane output type. cf. examples/focal.yaml

• LogNormal is a float value type that samples from a log normal distribution.

• ExcludedRandom is an int value type that samples from a given range (min to max), but excluding a specified list of integers. Useful for DES CCD numbers to avoid the bad CCDs.

• des_wcs is an input type to enable the DES_Local wcs type. It returns the local Jacobian wcs for an arbitrary ccd number and image position from a given DES WCS solution. cf. examples/des/meds.yaml in the GalSIm repo.

• WrongWCS is an image type that is identical to the normal Scattered image type, except that it changes the wcs to something different (output_wcs) before writing the image to disk. This is useful for characterizing the impact of wcs errors on subsequent measurements. cf. examples/wrongwcs.yaml.

• all_files is an input type that gets the names of all the files in a directory matching a particular glob string. This is useful when the number of CCDs say for your various exposures is not constant. e.g. Some files may be excluded for not being part of a coadd set or might not have been written due to some kind of error upstream. This type lets you dynamically use only the files that are present in the directory. It is connected to value types NFiles, ThisFileName and ThisFileTag.

• More. This is not an exhaustive listing. There are other modules that were made for targeted investigations, which are not likely to be of wider interest. Although, of course feel free to peruse the files and use anything that seems like it might be helpful.