v5.9.1

What's Changed

New Features 🎉

• Allow batched parameters in MvNormal and MvStudentT distributions by @ricardoV94 in #6897
• Logprob derivation of Max for Discrete IID distributions by @Dhruvanshu-Joshi in #6790
• Support logp derivation of power(base, rv) by @LukeLB in #6962

Bugfixes 🪲

• Make Model.str_repr robust to variables without monkey-patch by @ricardoV94 in #6942
• Fix bug in GP Periodic and WrappedPeriodic kernel full method by @lucianopaz in #6952
• Fix rejection-based truncation of scalar variables by @ricardoV94 in #6923

Documentation 📖

• Add expression for NegativeBinomial variance by @ricardoV94 in #6957

Maintenance 🔧

• Add constant and observed data to nutpie idata by @Y0dler in #6943
• Improve multinomial moment by @aerubanov in #6933
• Fix HurdleLogNormal Docstring by @amcadie in #6958
• Use numpy testing utilities instead of custom close_to* by @erik-werner in #6961
• Include more PyTensor functions in math module by @jaharvey8 in #6956
• Improve blackjax sampling integration by @junpenglao in #6963

New Contributors

• @Y0dler made their first contribution in #6943
• @amcadie made their first contribution in #6958
• @erik-werner made their first contribution in #6961

Full Changelog: v5.9.0...v5.9.1