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 will be available soon
The importable package lives in src/gpyrn. Example scripts and generated
figures live in examples, and bundled observational data lives in data.
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.
If you use this package in your work, please cite the following publication (currently under review)
@article{10.1093/mnras/stac3727,
author = {Camacho, J D and Faria, J P and Viana, P T P},
title = "{Modelling stellar activity with Gaussian process regression networks}",
journal = {Monthly Notices of the Royal Astronomical Society},
year = {2022},
month = {12},
issn = {0035-8711},
doi = {10.1093/mnras/stac3727},
url = {https://doi.org/10.1093/mnras/stac3727},
note = {stac3727},
eprint = {https://academic.oup.com/mnras/advance-article-pdf/doi/10.1093/mnras/stac3727/48453218/stac3727.pdf},
}
Copyright 2026 João Camacho
Licensed under the MIT license (see LICENSE).