jaxspec

:warning: jaxspec is still in early release: expect bugs, breaking API changes, undocumented features and lack of functionalities

jaxspec is an X-ray spectral fitting library built in pure Python. It can currently load an X-ray spectrum (in the OGIP standard), define a spectral model from the implemented components, and calculate the best parameters using state-of-the-art Bayesian approaches. It is built on top of JAX to provide just-in-time compilation and automatic differentiation of the spectral models, enabling the use of sampling algorithm such as NUTS.

jaxspec is written in pure Python, and has no dependancy to HEASoft, and can be installed directly using the pip command.

Documentation : https://jaxspec.readthedocs.io/en/latest/

Installation

We recommend the users to start from a fresh Python 3.10 conda environment.

conda create -n jaxspec python=3.10
conda activate jaxspec

Once the environment is set up, you can install jaxspec directly from pypi

pip install jaxspec --upgrade

Citation

If you use jaxspec in your research, please consider citing the following article

@ARTICLE{2024A&A...690A.317D,
       author = {{Dupourqu{\'e}}, S. and {Barret}, D. and {Diez}, C.~M. and {Guillot}, S. and {Quintin}, E.},
        title = "{jaxspec: A fast and robust Python library for X-ray spectral fitting}",
      journal = {\aap},
     keywords = {methods: data analysis, methods: statistical, X-rays: general},
         year = 2024,
        month = oct,
       volume = {690},
          eid = {A317},
        pages = {A317},
          doi = {10.1051/0004-6361/202451736},
       adsurl = {https://ui.adsabs.harvard.edu/abs/2024A&A...690A.317D},
      adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}