Changelog

GLMMcosinor (development version)

CRAN release: 2026-09-15

Refactor polar_plot() and autoplot() to avoid functions that are defined within other functions. Separate out into separate scripts where necessary.
Add example of using likelihood-ratio test to assess differential rhythmicity.
Use the mean of continuous variables as the reference level, and the first level for factors when creating newdata within autoplot().
Fix test-coverage GHA.
Use reformulas::findbars()/nobars() instead of the now-deprecated lme4::findbars()/nobars(), which were moved to the reformulas package and emitted a one-off deprecation warning that made test-cglmm.R order-dependent and flaky on R-devel. This was the cause of the CRAN archival on 2026-08-22.
Fix autoplot() legends being shown in the wrong order (colour before linetype/shape) when both ranef_plot and a grouping variable are plotted together, caused by a ggplot2 4.0 change to default guide ordering.

CRAN release: 2024-10-31

Fixed issues #14 and #15 relating to handling model formulas without amp_acro() components.
Update import from glmmTMB which previously failed with released version.
Perform the glmmTMB’s .onLoad() when GLMMcosinor is loaded by including an @importFrom glmmTMB in pkg documentation.
Plot labels now correspond to the name of the group(s), and they are more concise.
Component labels have been removed from polar_plot() when there is only one component in the model.

CRAN release: 2024-01-11

Successful peer review from rOpenSci!
Create a sigma() function which gets the dispersion parameter from a given cglmm model (by calling glmmTMB::sigma()).
Improve introduction to cosinor modeling in the getting-started vignette.
Fix small typos in vignettes/docs and make plot legend title consistent across autoplot() and polar_plot().
Rename functions to be snake_case and all S3 class names to camelCase.

First development version of GLMMcosinor for submission to rOpenSci.
Includes functions for fitting a cosinor model, similarly to the {cosinor} R package but using the glmmTMB modeling framework to allow more flexibility in terms of fitting generalized linear mixed cosinor models.