/DarkHistory

A code package for calculating modified cosmic ionization and thermal histories with dark matter and other exotic energy injections

Primary LanguageJupyter NotebookMIT LicenseMIT

DarkHistory v2.0, with improved treatment of low energy electrons and spectral distortions

The branch containing the upgraded treatment for low energy electrons and spectral distortions can be found here. In additional to the data files needed for v1.0, this upgrade requires additional data files.

The upgrades are described in a paper available at arXiv:2303.07366, and examples of applications are given in arXiv:2303.07370. Please cite these as well as arXiv:1904.09296 if you use this version of DarkHistory in a scientific publication.

DarkHistory v1.1 with Neural Network transfer functions

Added Neural Network transfer functions to optionally replace large tabulated transfer functions. Requires Tensorflow 2.0 in addition to v1.0 dependencies, and a compact dataset to use the Neural Network transfer functions. (To upgrade from v1.0, one can simply add the compact dataset to the existing data directory). To use the tabulated transfer functions, a full dataset is required. (This version of DarkHistory also works with v1.0 dataset with setting use_v1_0_data=True in config.py.)

The update is described in a paper available at arXiv:2207.06425. Please cite this paper as well as arXiv:1904.09296 if you use this version of DarkHistory in a scientific publication. The release for this version can be found here. For more information, please visit our webpage here.

DarkHistory is a Python code package that calculates the global temperature and ionization history of the universe given an exotic source of energy injection, such as dark matter annihilation or decay. DarkHistory is described in a paper available at arXiv:1904.09296. Please cite this paper if you use DarkHistory in a scientific publication. The data files for required for this version can be found here. The release for this version can be found here. For more information, please visit our webpage here.