pollux

DESC package to create weak lensing and large-scale structure simulations from N-body simulations for the Higher-Order Statistics Topical Team.

Features

• Interfaces with HACC and SLICS N-body simulations.
• Construction of particle and halo lightcones with replicated boxes and random isometries (translations, rotations and flips), typically identical for groups of shells of width given by the box size.
• Lensing in the Born approximation (BornLensingBuilder), including tomography from user-defined redshift tomographic distributions.
• Generation of tomographic galaxy catalogs using uniform sampling (CatalogSourceUniform) or clustered sampling via (linearly biased) sampling of the density field with Multinomial (CatalogSourceClusteredMultinomial, fixed total number of galaxies) or Poisson (CatalogSourceClusteredPoisson, fixed density) samplings.
• Efficient parallel computation using the Parsl framework to distribute tasks over computing resources.
• Use of DESC tools: CCL, healsparse and NaMaster.

Organisation of the code

• reader.py implements Reader objects to read particle data from N-body simulation snapshots, and provides basic lightcone information (particle density, shell binning, etc.)
• map_builder.py implements MapBuilder abstract objects with their basic functionalities: I/O, spherical harmonic transforms, visualization.
• shells.py implements ShellBuilder objects (subclass of MapBuilder) to generate particle and halo lightcone shells.
• lensing.py implements LensingBuilder objects (subclass of MapBuilder) to generate lensing maps from lightcone shells.
• catalog.py implements CatalogGenerator objects to generate tomographic galaxy catalogs from the lightcone shells and lensing maps.
• utils.py contains a nuber of helper functions for maps, Cl's computations and array manipulations.
• cosmo.py contains fixed cosmologies.