Martin Modrák

Custom c++ function may break when called with int parameters

2

#### Summary: Custom C++ functions may return slightly different results when called with seemingly equivalent inputs from Stan. The problem lies in not promoting `int` to `real`. #### Description: When...

Consolidate and refactor output from algorithms

5

#### Summary: This issues is to track progress on refactoring outputs. The design and other issues are discussed at forums: http://discourse.mc-stan.org/t/proposal-for-consolidated-output/4263/19 #### Description: This will be detailed once the design...

Support file_refit = "on_change" with brm_multiple

3

For first implementation of #1088 only the simpler case for single models was implemented. Support for `brm_multiple` has a WIP code at https://github.com/martinmodrak/brms/tree/extended_cache_brm_multiple but incorporation awaits better support for testing...

Centered and partially centered parametrization of varying effects/intercepts

1

Mike's case study on hierarchical models (https://betanalpha.github.io/assets/case_studies/hierarchical_modeling.html) mentions that for some situations a centered parametrization is superior to a non-centered one and that yet in other contexts centering only some...

#### Summary: If an error is thrown withing `rstan::config_argss`, it is reported simply as `'error in specifying arguments; sampling not done'`, without reproducing the error produced in `rstan::config_argss` #### Description:...

posterior_predict() fails with newdata for models estimated with family = mgcv::betar

4

#### Summary: Similar to #324, I can run `posterior_predict` and `posterior_linpred` for betareg models, but as soon as I specify `newdata = XX` I get an error. #### Description: Running...

beta_binomial_lpmf is unstable for large shape arguments

1

## Description As the shape arguments for the beta part get large (roughly > 1e14) the `beta_binomial_lpmf` function starts to return problematic values. ## Example ```c++ beta_binomial_lpmf(500, 1000, 0.5 *...

WOFF font files are not served with a correct MIME type (application/font-woff)

4

As WOFF fonts are (slightly) more widely supported by browsers then TTF fonts (see http://caniuse.com/#feat=woff versus http://caniuse.com/#feat=ttf) I wanted to use them in my app, but it turns out, elm-reactor...

#### Summary: The new warning page (https://mc-stan.org/misc/warnings.html) hopefully provides better context for understanding and diagnosing warnings from `diagnose` output (divergences, treedepth, ...). So the current short description and hints (e.g....

The paragraph on Placebo effects is IMHO outdated so I took the liberty of writing a paragraph summarizing more or less recent evidence + a reference. Hope you like it,...