Skip to content

A Bayesian Python code to fit the axion-photon parameter space to cosmological data.

License

Notifications You must be signed in to change notification settings

ChenSun-Phys/cosmo_axions

Repository files navigation

Note:

This is a modified version that adds the options to generate a random ICM magnetic domain realization for each of the galaxy cluster, as well as allowing magnetic field in ICM to have random orientations in each domain. For the version that reproduces the results in the published article, see our original version here at this repository.

cosmo_axions

A Bayesian Python code to fit the axion-photon parameter space to cosmological data.

Written by Manuel A. Buen-Abad and Chen Sun, 2020

Requirements

  1. Python
  2. numpy
  3. scipy
  4. emcee
  5. corner

How to run

In the terminal:

$ python cosmo_axions_run.py -L likelihoods/ -o path/to/your/chain/output/ -i inputs/the_param_file.param -N number_of_points -w number_of_walkers

After the runs are finished, you can analyze them with:

$ python cosmo_axions_analysis.py -i path/to/your/chain/output/

Once the analysis is done, if you wanna output the contours in ma-ga space from the frequentist likelihood ratio test, do:

$ python bin_chi2.py -c path/to/your/chain/output/ -b number_of_ma-ga_bins

where the argument with flag -b bins the ma-ga parameter space in order to minimize the chi2 in each bin. A value of ~50 is good enough.

Bibtex entry

If you use this code or find it in any way useful for your research, please cite Buen-Abad, Fan, & Sun (2020). The Bibtex entry is:

@article{Buen-Abad:2020zbd,
    author = "Buen-Abad, Manuel A. and Fan, JiJi and Sun, Chen",
    title = "{Constraints on Axions from Cosmic Distance Measurements}",
    eprint = "2011.05993",
    archivePrefix = "arXiv",
    primaryClass = "hep-ph",
    month = "11",
    year = "2020"
}

About

A Bayesian Python code to fit the axion-photon parameter space to cosmological data.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published