gpjax
https://github.com/jaxgaussianprocesses/gpjax
Python
Gaussian processes in JAX.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported1 Subscribers
Add a CodeTriage badge to gpjax
Help out
- Issues
- Natural/Expectation variational families lack NGD machinery and can leave the PD cone
- State-space RTS smoother is covariance-form despite 'square-root' docstring
- Support quasi-periodic (Periodic × Matérn) product kernels in the state-space SDE registry
- Add a Student-t likelihood for robust regression
- Extend pathwise (decoupled) sampling to sparse variational posteriors
- Extract the duplicated sparse-predict projection block in variational families
- Delete the dead `_gram` computation method (RFF computes features twice)
- OILMM `predict` defaults to a dense O(M N² P²) covariance
- Merge `_newton_girard` and `_newton_girard_matrices` (OAK/Sobol)
- State-space GP: dense joint predictive covariance (follow-up to diagonal-only v1)
- Docs
- Python not yet supported