A new version (v1.4.0) of priorsense: an R package for efficient prior and likelihood sensitivity checks has hit CRAN.
Priorsense provides tools for efficiently checking the influence of the prior and likelihood on the posterior without refitting the model.
The biggest news is that priorsense is now a Stan package! Documentation now lives on the Stan webpage (Prior Diagnostics and Sensitivity Analysis • priorsense) and development now happens at the stan-dev GitHub.
Other news since the last CRAN release:
- getting started vignettes have been completely rewritten. It should now be much easier to get started with priorsense, regardless of if you are using Stan or another PPL
- priorsense was peer-reviewed by ROpenSci and received a gold badge
- a JOSS paper was published
- various usability improvements and bugfixes
Feature requests, issue reports, and questions are welcome as always! Either on the GitHub or here on Discourse.