gpyrn

Modelling stellar activity with Gaussian process regression networks

gpyrn is a Python package implementing a GPRN framework for the analysis of RV datasets.
A GPRN is a model for multi-output regression which exploits the structural properties of neural networks and the flexibility of Gaussian processes.

The GPRN was originally proposed by Wilson et al. (2012).

Documentation

Documentation is available here.

Authors

The gpyrn package was developed at IA, in the context of the PhD thesis of João Camacho, with contributions from João Faria and Pedro Viana.

Cite

If you use this package in your work, please cite the following publication (currently under review)

@ARTICLE{gpyrn2022,
    author = {{Camacho}, J.~D. and {Faria}, J.~P. and {Viana}, P.~T.~P.},
        title = "{Modelling stellar activity with Gaussian process regression networks}",
    journal = {arXiv e-prints},
    keywords = {Astrophysics - Earth and Planetary Astrophysics, Astrophysics - Solar and Stellar Astrophysics},
        year = 2022,
        month = may,
        eid = {arXiv:2205.06627},
        pages = {arXiv:2205.06627},
archivePrefix = {arXiv},
    eprint = {2205.06627},
primaryClass = {astro-ph.EP},
    adsurl = {https://ui.adsabs.harvard.edu/abs/2022arXiv220506627C},
    adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}

License

Copyright 2022 Institute of Astrophysics and Space Sciences.
Licensed under the MIT license (see LICENSE).