Priorsense 1.0 is now on CRAN

Publicity
1

priorsense 1.0 has been released and is now available on CRAN.

priorsense is an R package for efficiently checking whether the posterior is sensitivity to changes to the prior or likelihood (see the website and paper for details). It works with Stan (brms, rstan or cmdstanr) fits, or just posterior draws, as long as the log prior density and log likelihood evaluations (at the locations of the posterior draws) are available.

For Python users, the primary diagnostic is also available in ArviZ.

While reaching 1.0 and releasing on CRAN represents a significant milestone, development continues. If there are any suggestions for features, issues or bug reports, don’t hesitate to make an issue on the GitHub issue tracker or get in touch directly.

17 Likes
2

Congratulations, this looks incredibly useful! FYI I think it could be worthwhile to highlight which variables should be included in the Stan program. It appears that it’s log_lik and lprior? Thanks!

3

Looks great!

4

Yes. Although you can also use other names, and you can have several variables for different parts of prior

5

Congratulations!!