mardom / bajes Goto Github PK

bajes [baɪɛs] is a Python software for Bayesian inference developed at Friedrich-Schiller-Universtät Jena and specialized in the analysis of gravitational-wave and multi-messenger transients. The software is designed to be state-of-art, simple-to-use and light-weighted with minimal dependencies on external libraries.

bajes is compatible with Python v3.7 (or higher) and it is built on modules that can be easily installed via pip. The mandatory dependencies are numpy, scipy and astropy. However, the user might need to download some further packages. See INSTALL for more information.

bajes provides an homonymous Python module that includes:

• bajes.inf: implementation of the statistical objects and Bayesian workflow,
• bajes.obs: tools and methods for data analysis of multi-messenger signals. For more details, visit gw_tutorial.

The bajes package provides a user-friendly interface capable to easily set up a Bayesian analysis for an arbitrary model. Providing a prior file and a likelihood function, the command

python -m bajes -p prior.ini -l like.py -o /path/to/outdir/

will run a parameter estimation job, inferring the properties of the input model. For more details, visit inf_tutorial or type python -m bajes --help.

The bajes infrastructure allows the user to set up a pipeline for parameters estimation of multi-messenger transients. This can be easily done writing a configuration file, that contains the information to be passed to the executables. Subsequently, the following command,

bajes_pipe.py config.ini

will generates the requested output directory, if it does not exists, and the pipeline will be written into a bash executable (/path/to/outdir/jobname.sub). For more details, visit conifg_example.

bajes is developed by Matteo Breschi at the Friedrich-Schiller-Universität Jena with the contribution of Rossella Gamba and Sebastiano Bernuzzi.

If you find bajes useful in your research, please include the following citation in your publication,

@article{Bajes:2021,
         author         = "Breschi, Matteo and Gamba, Rossella and Bernuzzi, Sebastiano",
         title          = "${\tt bajes}$: Bayesian inference of multimessenger astrophysical data, 
                          methods and application to gravitational-waves",
         eprint         = "2102.00017",
         archivePrefix  = "arXiv",
         primaryClass   = "gr-qc",
         month          = "1",
         year           = "2021"}

See CREDITS for more information.

bajes has benefited from open source libraries, including the samplers,

• cpnest
• dynesty
• emcee

and the gravitational-wave analysis packages,

• bilby
• gwbinning
• lalsuite
• pycbc

We also acknowledge the LIGO-Virgo-KAGRA Collaboration for maitaining the GWOSC archive.