Skip to content

Return negative infinity after quiet factorization failures - #209

Open
rpinho wants to merge 1 commit into
exoplanet-dev:mainfrom
rpinho:fix-quiet-likelihood
Open

rpinho wants to merge 1 commit into
exoplanet-dev:mainfrom
rpinho:fix-quiet-likelihood

Conversation

@rpinho

@rpinho rpinho commented Nov 18, 2026

Copy link
Copy Markdown

Calling log_likelihood after compute with quiet=True can raise an attribute error or return NaN or positive infinity when factorization fails. A GP computed at repeated coordinates without observation noise reproduces the failure. Set the normalization to negative infinity and guard the quadratic form when the factorization has failed in NumPy, JAX, and PyMC. Regression tests check rejection before and after a prior factorization and compare likelihoods after recovery with reference calculations.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant