PhaseSpace (phasespace Python package)

PhaseSpace (phasespace) is an open-source Python software package for n-body phase space generation implemented using TensorFlow. It is part of the Scikit-HEP and zfit ecosystem for high-energy physics (HEP) data analysis. The package is maintained on GitHub at github.com/zfit/phasespace and documented at phasespace.readthedocs.io.

The library implements the Raubold and Lynch method for generating n-body decay events, a standard technique in particle and nuclear physics for Monte Carlo simulation. Unlike the ROOT-based TGenPhaseSpace tool it can complement or replace, phasespace runs purely in Python with TensorFlow as the computational backend, enabling GPU acceleration and integration with modern machine learning workflows.

Key capabilities include simulation of sequential and complex decay chains (e.g., B meson decays into multiple daughter particles), support for resonances with variable mass (non-fixed width), and compatibility with the DecayLanguage package for defining decay topologies. Results for simple n-body decays are validated against TGenPhaseSpace; sequential decays are verified against RapidSim, a fast Monte Carlo generator for heavy-quark hadron decays.

This project is a software library for academic and research use in particle physics and is not related to robotics, industrial automation, or any robotics hardware or software company. It was listed in the RobotToday directory under the URL phasespace.readthedocs.io, but this URL corresponds solely to this physics software package. The project is open-source and community-maintained with no commercial entity behind it.