NEWS

---

ExperimentalDesignGeneratorandRandomiser 0.1.0 (2026-09-02)

Initial native R port of EDGAR.

• Added nine experimental designs implemented natively in R: cr_eq, cr_uneq, rcb, rcb_uneq, two_factor_rcb, latin, split_plot, variable_blocks, alpha.
• Added public API: list_designs(), generate_design(type, ..., seed = 0L), validate_design(type, ...), plus design-specific convenience functions (design_cr(), design_rcb(), design_alpha(), and so on).
• Added S3 edgar_design class with print, as.data.frame, total_units, has_layout, and as_layout_frames methods.
• Added exporters write_edgar_csv() (no extra dependencies), write_edgar_json() (requires jsonlite), write_edgar_xlsx() (requires openxlsx).
• Added propose_alpha_structures() helper for alpha design feasibility checks. choose_design() is provided as an alias matching the upstream Python API.
• Added pure-R reimplementation of CPython's random.Random Mersenne Twister seeding and Fisher-Yates shuffle, giving byte-identical cross-language reproducibility for the same integer seed.
• Added isolated RNG that does not modify the global .Random.seed.
• Added testthat test suite covering structural invariants, validation failures, determinism, and byte-identical parity against the upstream Python edgar-design package.
• Added two vignettes: "Introduction to EDGAR" and "Alpha designs and structure proposal".

EDGAR was originally developed by the Biometrics team at Rothamsted Research as Excel workbooks (http://www.edgarweb.org.uk/). The algorithms were subsequently re-implemented in the Python project rotsl/edgar, documented at https://rotsl.github.io/edgar/ and distributed as edgar-design on PyPI. This R package is a native R port of that Python implementation. Alpha designs follow Patterson and Williams (1976), doi:10.1093/biomet/63.1.83.