NEWS

---

ggadjustedforest 0.1.1

Changes

• cumulative = TRUE with model_type = "logistic" or "coxph" now emits a warning explaining that odds ratios and hazard ratios are non-collapsible effect measures. A shifting OR/HR across sequential models cannot be cleanly attributed to confounding; cumulative display is most interpretable for collapsible measures (risk differences, risk ratios). The feature is retained for descriptive and predictive modelling contexts.

Documentation

• Updated vignette and README examples to use the rotterdam breast cancer dataset (from survival) for the Cox model demonstration.
• Installation instructions updated to reflect CRAN availability.
• Long plot titles now wrap automatically to prevent clipping.

ggadjustedforest 0.1.0 (2026-07-17)

Initial release

• gg_adjusted_forest(): main function producing forest plots and tables for unadjusted and adjusted exposure effects.
• forest_table(): standalone table extraction helper.
• Supports "logistic", "linear", "poisson", and "coxph" model types.
• Cumulative adjustment mode (cumulative = TRUE).
• Optional cluster-robust standard errors via sandwich / lmtest.
• Optional survey weights.
• patchwork-based combined plot + table display.
• S3 print and plot methods for ggadjustedforest objects.