ravest

Radial Velocity fitting with Bayesian model comparison

Features:

• Model exoplanets and host stars, to simulate RV data for given orbital and instrumental parameters
• Fit RV data with MCMC to explore posterior distributions for parameters - including Gaussian Processes for stellar activity
• Bayesian Model Comparison using the Learned Harmonic Mean Estimator from harmonic
• Visualise/animate the star's orbit (coming soon!)

Check out the example notebooks at the online documentation to see examples of how to do all of these!

GitHub: github.com/ross-dobson/ravest

Installation

It should be as simple as

$ pip install ravest

JAX is a requirement for ravest (and harmonic), so you may want to consult the JAX installation docs if you want GPU or TPU support (tl;dr: install JAX first according to those instructions, then install ravest on top).

Usage

For an example of how to fit a model to RV data, see Example 1: fitting a single planet, where we fit ELODIE data for 51 Peg b.

For fitting a two-planet system, comparing circular and eccentric orbits, see Example 2: fitting a two-planet system, where we fit HIRES data for K2-24.

For an example of how to use a Gaussian Process to mitigate stellar variability, see Example 3: fitting with a Gaussian Process, where we use a quasiperiodic kernel on HARPS data for K2-229.

For an example on using the Learned Harmonic Mean Estimator from harmonic to compare two competing RV models by estimating the Bayesian evidence $\mathcal{Z}$ and Bayes factors, see Example 4: using the Learned Harmonic Mean Estimator, where we compare a one-planet and two-planet fit for TOI-544. It also demonstrates a multi-instrument fit, using both HARPS and HARPS-N data.

If you have any questions, check the Frequently Asked Questions, raise an issue on Github, or email me and I'll be happy to help.

Citing ravest

If you use ravest in your research, please cite Dobson et al. (2026).

If you use the Learned Harmonic Mean Estimator for Bayesian evidence estimation, then please also cite the harmonic team: see their GitHub repository for the relevant papers to cite, starting with McEwen et al. (2021).

Contributing

Interested in contributing? Check out the contributing guidelines. Please note that this project is released with a Code of Conduct. By contributing to this project, you agree to abide by its terms.

Acknowledgements

Ravest makes use of the following open-source packages:

• NumPy for numerical computing
• SciPy for scientific computing algorithms
• Matplotlib for plotting and visualisation
• Astropy for astronomical calculations and utilities
• pandas for data manipulation
• tqdm for progress bars
• emcee for MCMC sampling
• corner for visualising posterior distributions
• tinygp for Gaussian Process modelling
• JAX for accelerated array computing and 64-bit precision, also required by tinygp and harmonic
• harmonic for Bayesian evidence estimation via the Learned Harmonic Mean Estimator

License

ravest was created by Ross Dobson. It is licensed under the terms of the GNU General Public License v3.0 or later (GPL-3.0-or-later).