Nx2pt is designed to be a simple to use wrapper around the LSST DESC tools NaMASTER and SACC to calculate potentially large sets of cross-correlations and their covariance.
Installation should be as simple as
git clone https://github.com/ajouellette/nx2pt
cd nx2pt
python -m pip install . [--user] [-e]
Most uses of Nx2pt will simply involve running the script run_nx2pt which takes a YAML file that defines all the tracers of interest and specifies which cross-spectra and covariances to calculate.
A very simple example of a pipeline.yaml file is given below:
nside: 1024
# binning scheme for cross-spectra
binning:
kind = "linear"
delta_ell = 50
# where to save namaster workspaces containg the mode coupling matrices
workspace_dir: "./workspaces"
tracers:
# define first tracer (some galaxy sample, for example)
gal1:
name: "Galaxy sample 1"
data_dir: "galaxies1"
healpix:
map: "galaxy1_delta.fits"
mask: "galaxy1_mask.fits"
# define second tracer (some other galaxy sample)
gal2:
name: "Galaxy sample 2"
data_dir: "galaxies2"
healpix:
map: "galaxy2_delta.fits"
mask: "galaxy2_mask.fits"
cross_spectra:
# calculate all auto- and cross-spectra
list:
- [gal1, gal2]
- [gal1, gal1]
- [gal2, gal2]
# calculate full covariance
covariance: True
# save everything to a .npz file
save_npz: "galaxy_3x2pt_spectra.npz"It is very easy to setup significantly more complicated pipelines with arbitrary number of tracers (each possibly with tomographic bins and/or spin). More pipeline examples are given in examples.
After specifying the tracers and various configuration options, all of the cross-spectra and covariances can be calculated by simply running:
run_nx2pt pipeline.yaml
If you use this code in a publication, make sure to cite all the relevant papers that describe NaMASTER and SACC. Thank you to all of the LSST DESC members who have developed and maintain these tools.