NEWS

---

DataSum 1.0.0 (2026-07-13)

Breaking changes

• Replaces the prototype API (Datum(), DataSumm(), getmode(), and shapiro_normality_test()) with a clean research-facing API: summarize_vector(), summarize_data(), profile_data(), datasum_report(), and run_datasum_app().
• Normality diagnostics now return the test name, statistic, p-value, alpha, and decision instead of a single label.

New features

• Adds NA-aware summaries for numeric, character, factor, logical, date, datetime, and unsupported columns.
• Adds robust statistics: quartiles, IQR, MAD, outlier counts, missingness diagnostics, skewness, and excess kurtosis.
• Adds grouped summaries through summarize_data(..., by = ...).
• Adds profile_data() for dataset-level diagnostics and warnings.
• Adds datasum_report() to create reproducible Quarto diagnostic reports.
• Adds run_datasum_app() as a Shiny interface for uploaded CSV files, summaries, warnings, plots, and report download.
• Adds testthat coverage, GitHub Actions, pkgdown configuration, citation metadata, and an introductory vignette.