Changelog

Source: NEWS.md

postlink 0.1.1

New Features

• Updated the priors argument for the Bayesian Mixture functions (adjMixBayes(), glmMixBayes(), survregMixBayes()) so that intercept priors and slope coefficient priors are now decoupled.

Architecture & S3 Methods

• Refactored S3 class assignment system to remove false inheritance from base glm, lm, and coxph classes. Fitted objects now use dedicated package-level classes (e.g., plmodel, plglm, plcoxph).
• Added standard generic extractors (i.e., coef(), vcov(), confint(), df.residual()) for compatibility with tools like lmtest::coeftest().
• Added custom methods to indicate unsupported standard likelihood and residual-based generics (e.g., logLik(), profile(), anova(), extractAIC(), cooks.distance(), rstudent()).
• Consolidated shared printing behaviors for adjustment objects (adjELE, adjMixture, adjMixBayes) using a unified base print.adjustment() method via NextMethod(), reducing code duplication.
• Added an error message to clarify that the gaussian family in glmMixture is intended to be used only with the identity link for now.

Documentation & Testing

• Standardized all Roxygen manual page titles to Title Case. Implemented a GitHub Actions workflow (format-titles.yml) for documentation styling in continuous integration.
• Removed VignetteBuilder: knitr from the DESCRIPTION file to resolve a CRAN NOTE, as the extended package articles are hosted only via pkgdown for now.
• Updated the testthat suite to align with the newly refactored S3 class structures and console outputs.

postlink 0.1.0

CRAN release: 2026-04-15

• Initial CRAN release.
• Implements a suite of statistical tools (weighting or mixture modeling) for secondary analysis of linked data accounting for mismatch errors.
• Added support for adjusting generalized linear models, Cox proportional hazards models, parametric survival modeling, and contingency tables.