Skip to content

Epidist 0.3.0

Latest
Compare
Choose a tag to compare
@seabbs seabbs released this 07 Mar 20:18
· 1 commit to main since this release
051d0cf

This release adds support for a wider range of distributions in the marginal model, improves documentation with new vignettes and FAQ sections, enhances the getting started guide with clearer examples of model comparison, and fixes several bugs related to parameter bounds and likelihood calculations.

Models

  • Added Stan-side support for fitting all distributions supported by primarycensored in the marginal model. See #540.
  • Added R-side analytical likelihood support for Lognormal, Gamma, and Weibull distributions. See #540.

Package

  • Remove caching of vignettes. See #533.

Documentation

  • Added a new vignette "Guide to the statistical models implemented in epidist". See #514.
  • Added a new FAQ section showcasing how to use the posterior package with epidist models, particularly for working with random variables (rvars) to propagate uncertainty in calculations. See #547.
  • Added a new FAQ section on how to use the marginaleffects package with epidist models. See #547.
  • Reduced the focus on simulating data in the getting started vignette to make it more accessible. See #549.
  • Made the entry to the package friendlier with clearer examples and improved documentation. See #549.
  • Added a schematic to explain right truncation more clearly to the getting started vignette. See #549.
  • Added a comparison of fitting naive and marginal models in the getting started vignette to highlight the importance of accounting for biases. See #549.
  • Added examples showing how to extract estimated parameters and plot them against true values to evaluate model performance. See #549.

Bugs

  • Fixed a vector length issue for censoring that was causing problems in some likelihood calls. See #540.
  • Fixed a bug in the preprocessing of the Weibull family. See #540.
  • Fixed a bug where bounds were not set for mu parameters in custom families. See #549.
  • Fixed a bug in predict_delay_parameters() where it couldn't detect brms families when used directly. See #549.

What's Changed

New Contributors

Full Changelog: v0.2.0...v0.3.0