PSI
(the Phase Space Intersector) is a Python module for post-processing the outputs of cosmological simulations. PSI
applies the Phase Space Sheet discretization (see Abel et al. 2012) to a set of particles, then volume-samples the density and velocity fields (as well as higher-order moments) onto a Cartesian grid. Anyone can use PSI
to produce smooth and continuous density and velocity maps from particle data! The volume-sampling procedure is geometrically exact, and converves mass down to machine precision.
The low-level volume-sampling routines are based on the r3d software, described in Powell & Abel (2015) and LA-UR-15-26964.
-
First,
git clone
the repository andcd
into it -
Then,
python setup.py install
. -
To use in Python, simply
import PSI as psi
or similar -
See
tutorial/Tutorial.ipynb
for a guide to using PSI!
PyPSI is free to use under the MIT License. See source files for the full license text.
If you use PyPSI for your research, please cite:
-
Devon Powell and Tom Abel. An exact general remeshing scheme applied to physically conservative voxelization. Journal of Computational Physics, 297:340–356, 2015a (link)
-
Devon M. Powell. r3d: Software for fast, robust geometric operations in 3D and 2D. Los Alamos National Lab technical report, LA‑UR‑15‑26964, 2015a (link)
Devon M. Powell (Stanford, MPA Garching)
Tom Abel (Stanford, SLAC)
Goran Jelic-Cizmek (University of Geneva)