Changelog

Source: NEWS.md

metalite.ae 0.1.5

Improvements

• Speed up the bisection confidence-interval search in rate_compare_sum() (and therefore rate_compare() and extend_ae_specific_inference()) by reusing shared interval endpoints and hoisting loop-invariant quantities out of the objective function; results are unchanged up to floating-point tolerance (#253, thanks to @yihui).
• Further speed up unstratified confidence intervals in extend_ae_specific_inference() by computing every term’s interval in one vectorized pass: the bisection grid is evaluated across all terms at once and every sign-change bracket is refined by a single vectorized bisection loop, instead of scanning the grid separately for each term. Stratified inputs still use rate_compare_sum() per term. Results are unchanged up to floating-point tolerance (#257, thanks to @yihui).
• Speed up prepare_ae_specific() by dropping an unused order() pre-sort of the full observation table when collecting the distinct SOC/preferred-term pairs; the downstream merge() re-sorts by key and the row order is derived from factor levels, so the sort never affected the output. Results are unchanged for valid MedDRA data (one system organ class per preferred term) (#259, thanks to @yihui).
• Trim per-call allocations in the Miettinen-Nurminen variance helper (replacing ifelse() with sign arithmetic and indexed assignment) and reset row names before the Total-group rbind() in prepare_ae_specific() to avoid a costly make.unique(). Results are unchanged (#260, thanks to @yihui).

metalite.ae 0.1.4

CRAN release: 2026-08-29

New features

• The new gt_ae_specific() function is developed to output AE specific tables in gt format (#235, thanks to @LittleBeannie).
• The new gt_ae_summary() function renders formatted adverse event summaries as gt tables, including titles, treatment-group spanners, footnotes, and source notes (#215, thanks to @ginnaram13).
• rate_compare() and rate_compare_sum() now accept a bisection argument to control the number of sections used by the bisection method (#221, thanks to @lm3388).
• The analysis argument is now required by all tlf_*() functions so that table titles are generated from the selected analysis metadata (#207, thanks to @lm3388).
• meta_ae_example() is no longer exported. Package examples now construct metadata directly from the example datasets in forestly (#229, thanks to @LittleBeannie).
• Added interactive AE listing workflow support with new format_ae_listing() and react_ae_listing() functions for structured listing formatting and reactable-based rendering (#239, thanks to @LittleBeannie).

Bug fixes

• format_ae_summary() now uses parameter labels when custom parameters do not define summ_row, preventing row-name and summary-data length mismatches (#224, thanks to @LittleBeannie).
• gt_ae_summary() now supports missing source notes and footnotes without passing NULL to gt::md() (#226, thanks to @LittleBeannie).
• prepare_ae_specific() now preserves SOC and preferred-term row ordering when the data exceed the previous fixed ordering placeholder (#214, thanks to @ginnaram13).

Improvements

• The metalite_ae_adexsum example dataset has been expanded for adverse event forest plot workflows (#220, thanks to @lm3388).
• Unused arguments and R CMD check warnings in gt_ae_summary() and prepare_ae_specific() have been removed (#232, thanks to @fukuhiro2023).
• Package examples now use direct, reproducible metadata definitions based on forestly example data (#229, thanks to @LittleBeannie and @fukuhiro2023).

Documentation

• The adverse event vignettes and pkgdown site have been reorganized, with new articles for custom columns, filtering and sorting, and mock tables (#204, #231, thanks to @lm3388 and @LittleBeannie).
• Added a new vignette for interactive AE listing tables and updated listing workflow documentation and pkgdown references for the new listing functions (#239, thanks to @LittleBeannie).
• Added a new vignette demonstrating how one metadata object can generate multiple AE outputs (summary, specific, listing, and forest plot), and updated pkgdown article navigation for this workflow (#239, thanks to @LittleBeannie).

metalite.ae 0.1.3

CRAN release: 2024-10-23

New features

• format_ae_specific() now has new arguments filter_method and filter_criteria to control displayed rows in a table, sort_order and sort_column to sort an output table, and hide_soc_stats to control display of statistics for SOC rows (#191, #192, #203, thanks to @fukuhiro2023).
• Add data/metalite_ae_adesxum.rda (#189).
• Add event counts to avg_count() and extend_ae_specific_events() (#194).
• Add empty_table() to return an empty table object when data has no population (#200).

Bug fixes

• Fix bug to correctly display the order of means of duration and means of event count in extend_ae_specific_events() and extend_ae_specific_duration()(#194, #199).
• Fix bug to avoid an error when a treatment group has no observation to calculate event counts and duration in avg_event() and avg_duration() (#199).
• Fix bug to avoid an error when there is no observation to summarize event counts and duration in avg_event(), avg_duration(), extend_ae_specific_events() and extend_ae_specific_duration() (#200).

Improvements

• Update fmt_est() to change display of missing values from a character string of “NA” to a blank in an output table (#194).
• Update extend_ae_specific_inference() to pass additional arguments to rate_compare_sum() (#197).

metalite.ae 0.1.2

CRAN release: 2024-04-16

New features

• tlf_ae_summary() now has a new argument title for a user-customized title (#133).
• Add new functions: extend_ae_specific_subgroup(), prepare_ae_specific_subgroup(), format_ae_specific_subgroup(), tlf_ae_specific_subgroup(), extend_ae_summary_eaer(), format_ae_exp_adj() and tlf_ae_exp_adj() (#150, #161, #174, #180).
• Add data/metalite_ae_adex.rda (#164).
• Add new vignettes: vignette("ae-specific-subgroup") and vignette("exposure-adjusted-event-rate") (#167, #169).
• Add System Organ Classes (SOC) information to outdata in prepare_ae_specific() (#186).

Bug fixes

• Fix bug to avoid warning messages when no comparison variables are requested in format_ae_specific() (#135).
• Fix bug to align the row order of output data with the input one in avg_event() and avg_duration() (#136).
• Fix bug to check missing grouping variable after a subset of input data in prepare_ae_specific() (#137).
• Fix bug to display item if AE term has NA in prepare_ae_specific() (#166).
• Fix bug to correctly assign treatment labels when a treatment variable has NA (#179).

Improvements

• Update GitHub Actions workflow (#131, #163, #181).
• Add styler workflow (#134).
• Add CRAN download badge (#140).
• Remove dependencies on dplyr and tidyr from avg_event(), avg_duration() and fmt_val() (#136).
• Change the default value of the title argument in tlf_ae_specific() (#138).
• Remove R/n_subject.R and change to use metalite::n_subject() in prepare_ae_specific() (#144).
• Race values for adsl and adae are converted to title case in meta_ae_exmple() (#158).
• Organize files for utility functions (#159)
  • Rename R/rtf_output.R to R/utility.R.
  • Remove rate_compare_sum.R, R/to_mock.R and R/to_sentence.R.
  • Move to_rate_compare_sum() to R/rate_compare.R, and move to_mock() and to_sentence() to R/utility.R.
• Remove library calls from tests (#183).
• Add R/outdata.R to control outdata object (#161).
• Add an installation section to README.md (#177).
• Improve style and formatting for vignettes and code (#158, #168).

metalite.ae 0.1.1

CRAN release: 2023-02-24

• Updated the DESCRIPTION file to reformat the reference.
• Uncommented the previously commented code example.

metalite.ae 0.1.0

• Initial submission to CRAN.
• Added a NEWS.md file to track changes to the package.