powerbias

Pure Python code that computes power spectrum and/or bias from a simulated catalog of cosmological sources. Features:

fast interpolation of the density field in multiprocessing mode through the built-in python multiprocessing package or through MPI (suggested for large collections of simulated particles);
can be easily expanded to higher order interpolation schemes (currently supports order 2);
fast Fourier transform and deconvolution thanks to numpy;
defines a PowerSpectrum object to standardize operations;
save and load PowerSpectrum objects to disk in a standard binary format thanks to pickle;
can compute the cross-correlation power spectrum between two PowerSpectrum instances;
can compute the average power spectrum between multiple PowerSpectrum instances, together with their covariance matrix;
can compute the Feldman-Kaiser-Peacock estimator and iteratively optimize its weights for high accuracy;
can perform interlacing to mitigate aliasing effects on high frequency modes;
can sample a user-defined posterior distribution on a generic power spectrum model for Bayesian inference.

Documentation is currently being built.

Requires the following packages:

MatPeron/powerbias