Code to compute small-scale power spectra and isotropic bispectra for cosmological simulations and galaxy surveys of arbitrary shape, based on the configuration space estimators of Philcox & Eisenstein (2019, MNRAS, arXiv) and Philcox (2020, submitted, arXiv). This computes the Legendre multipoles of the power spectrum, 
or bispectrum, 
by computing weighted pair counts over the simulation box or survey, truncated at some maximum radius 
. This fully accounts for window function effects, does not include shot-noise or aliasing, and is optimized for small-scale spectrum computation in real- or redshift-space. Both the power spectrum and bispectrum estimators have efficiency 
for 
particles, and become faster at large 
.
The code can be run either in 'aperiodic' or 'periodic' mode, for galaxy surveys or cosmological simulations respectively. The 'periodic' mode contains various optimizations relating to the periodic geometry, as detailed in the second paper. HIPSTER also supports weighted spectra, for example when tracer particles are weighted by their mass in a multi-species simulation. Generalization to anisotropic bispectra is straightforward (and requires no additional computing time) and can be added on request.
Full documentation of HIPSTER is available on ReadTheDocs.
To compute a power spectrum from particles in a periodic simulation box (data.dat) up to 
with pair-counts truncated at radius using -space binning file binning.csv and 4 CPU cores, run:
./hipster_wrapper_periodic.sh --dat data.dat --l_max L -R0 R0 -k_bin binning.csv --nthreads 4
To compute a power spectrum from galaxies in a non-periodic survey (data.dat), defined by a set of randoms (randoms.dat), up to , truncating pair-counts at and using -space binning file binning.csv, with 4 CPU-cores, run:
./hipster_wrapper.sh --dat galaxies.dat --ran_DR randoms.dat --ran_RR randoms.dat -l_max L -R0 R0 -k_bin binning.csv --nthreads 4
To compute an isotropic bispectrum from particles in a periodic simulation box (data.dat) up to with pair-counts truncated at radius using -space binning file binning.csv and 4 CPU cores, using 3 times as many random points as data points, run:
./hipster_wrapper_bispectrum.sh --dat data.dat --l_max L -R0 R0 -k_bin binning.csv --nthreads 4 --f_rand 3
For any queries regarding the code please contact Oliver Philcox.
New for version 2: Optimizations for periodic N-body simulations
New for version 3: A new pair-count estimator for the periodic bispectrum