Flexible, extensible, hardware-agnostic gravitational-wave population inference.
It provides:
- Simple use of GPU-acceleration via JAX and cupy.
- Implementations of widely used likelihood compatible with Bilby.
- A standard format for defining new population models.
- A collection of standard population models.
If you're using this on high-performance computing clusters, you may be interested in the associated pipeline code gwpopulation_pipe.
Please cite Talbot et al. (2019) if you use GWPopulation in your research.
@ARTICLE{2019PhRvD.100d3030T,
author = {{Talbot}, Colm and {Smith}, Rory and {Thrane}, Eric and {Poole}, Gregory B.},
title = "{Parallelized inference for gravitational-wave astronomy}",
journal = {\prd},
year = 2019,
month = aug,
volume = {100},
number = {4},
eid = {043030},
pages = {043030},
doi = {10.1103/PhysRevD.100.043030},
archivePrefix = {arXiv},
eprint = {1904.02863},
primaryClass = {astro-ph.IM},
}Additionally, please consider citing the original references for the implemented models which should be include in docstrings.