AurelienNicosiaULaval/CircularRegression: CircularRegression 0.4.0 - Release Notes

CircularRegression 0.3.0 This release aligns the package with the methodological framework of Rivest et al. (JRSS C, 2016), with a clearer API, a recommended two-step workflow, and multiple critical robustness fixes. Major changes (breaking changes) angular() now uses explicit reference handling via reference = c("auto","first","name"). consensus() replaces initbeta with initkappa (direct kappa_j parameterization). coef.consensus() now supports type = "kappa" (default) and type = "beta" (reference-based ratios). Wrappers meanDirectionModel(), decentredPredictorModel(), presnellModel(), and jammalamadakaModel() now return a pipeline object (consensus + homogeneous), not only a homogeneous fit. New features Added select_reference_angle() (article criterion: maximize mean(cos(y - x_j))). Added angular_two_step() implementing the recommended strategy: fit consensus, select reference angle, fit homogeneous model initialized from consensus ratios. pick_reference_angle() is kept as a deprecated transition wrapper. Important fixes Fixed maxiter = 0 crashes in both angular() and consensus(). Added explicit identifiability checks for homogeneous fits. Stabilized logLik.angular() numerically (scaled Bessel), improving compatibility with AIC/BIC/logLik (df and nobs attributes). Ensured consistency between stored BIC values and S3 methods. Made predict.angular_re() self-contained (no dependency on caller environment objects). Aligned NA/cluster handling in angular_re. Quality, tests, documentation Expanded testthat coverage for non-regression and new API behavior. Updated API documentation for the two-step workflow. Removed the vignette from the release tarball (in preparation for a dedicated R Journal-style paper). Refreshed README and added a finalized package hex logo. Quick migration notes consensus(..., initbeta = x) becomes consensus(..., initkappa = x). Prefer angular_two_step() as the default analysis workflow. Migrate from pick_reference_angle() to select_reference_angle(). Methodological reference Rivest, L.-P., Duchesne, T., Nicosia, A., & Fortin, D. (2016). A general angular regression model for the analysis of data on animal movement in ecology. JRSS C, 65(3), 445–463.