New software tool for modeling dark matter and galaxy movement
The PIANISTpy package uses JAX to build models of dark matter haloes and galaxies.
Researchers created PIANISTpy, a JAX-based Python package. It builds spherical, non-rotating initial conditions for astrophysical models.
The tool handles velocity-anisotropic astrophysical equilibrium models. Velocity anisotropy is a non-trivial imprint on haloes and galaxies from mergers and anisotropic mass accretion.
PIANISTpy uses an analytical framework that performs truncation in phase space. This framework supports arbitrary double-power-law analytical radial density profiles.
The package is GPU-accelerated. It can create systems with non-analytical anisotropy profiles using a linear superposition of analytical models as basis functions.
It outputs 3D velocity profiles and line-of-sight-integrated observables. These outputs feature end-to-end differentiability for rapid gradient-based Monte Carlo inference.
The tool provides self-consistent local dark matter velocity distributions. This goes beyond the unphysical Maxwell-Boltzmann simplification used in terrestrial detection experiments.
Key facts
- PIANISTpy is a JAX-based Python package for constructing initial conditions.
- The framework supports arbitrary double-power-law analytical radial density profiles.
- The tool provides higher numerical solver accuracy and superior dynamical stability compared to galpy and AGAMA.