danielgruen / flaskpipe Goto Github PK

This is a set of scripts for creating many log-normal realizations of correlated matter/shear fields over the full sky with flask (https://github.com/hsxavier/flask).

1. Installation You need flask https://github.com/hsxavier/flask and a python environment with some common (numpy) and less common (healpy) packages.

2. Configuration To describe your setup of fields and their mutual cross-correlations,

• edit config.rc to contain the correct directories, bias/density for your galaxy fields, name of healpix mask (possibly containing several realizations in one sky), prefix to healpix masks of individual realizations (only required for optional measurement scripts), NSIDE of mask/output maps and LMAX of your input Cl files a RUN name for your configuration
• place these masks of individual realizations and their stack in this directory in healpix .fits.gz format some tools for creating masks: see below
• edit ${RUN}-info.tab to list all galaxy fields (I do that as 1 1 ..., 1 2 ..., 1 3 ...) and shear fields (as 2 1 ..., 2 2 ..., ...) see example-info.tab for an example for two foreground fields and to lensing source bins
• copy example.config to ${RUN}.config; likely you don't have to edit anything here (many options are set by 02runflask.sh on the fly)
• create a ${RUN}Cl-XY.dat file containing the C_l for each combination of fields X(=f?z?) and Y in your setup see exampleCl-* for an example; note that LMAX in config.rc and the maximum ell in these files must match
• create a ${RUN}pz-f1.dat - just copy the example if you never care for the actual redshift of individual sources in any of your fields

3. Running run.sh shows how to produce a single realization of outputs. As is usually a good idea for job queue systems, this will copy over files to a work directory, run flask there, and copy back catalogs/maps of density, convergence and shear for storage. If you want to make any measurements of data vectors, add these between steps 02runflask.sh and 04save.sh.

4. Creating masks Starting from a single healpix mask of your survey, first resample it to your desired Nside with healpix_resample.py you can generate a few rotated masks with healpix_freerotate.py and co-add the rotated masks with healpix_coadd.py

To generate 8 rotations of a DES Y1 SPT footprint m1.fits in a single sky, e.g. use healpix_freerotate.py m1.fits 180 0 m2.fits healpix_freerotate.py m1.fits 0 60 m3.fits healpix_freerotate.py m1.fits 90 60 m4.fits healpix_freerotate.py m1.fits 180 60 m5.fits healpix_freerotate.py m1.fits 270 60 m6.fits healpix_freerotate.py m1.fits 0 120 m7.fits healpix_freerotate.py m1.fits 180 120 m8.fits