mvmetrics 0.1.0
April 2026 — Initial release
New features
- Five metric families (A–E) with full roxygen2 documentation and examples.
- Novel E-CVWMD (Metric E) with CV-derived weights and adaptive α* via multivariate distance-correlation test (
energy::dcov.test).
evaluate_models() — main wrapper returning scores, rankings, recommendation, and bootstrap CIs in a single call.- Non-parametric bootstrap CIs via
bootstrap_metrics() with pairwise CI overlap test.
- Four ggplot2 visualisation types via
plot_metrics(): bars, forest, heatmap, decomp.
- Rule-based
recommend_metric() detecting scale heterogeneity, flat Metric A, and dependence signal.
- 20+ unit tests via testthat.
- Calibrated simulation example (
use_simulation.R) with Valle del Cauca Clausius-Clapeyron parameters.
Dependencies
- Imports: stats, utils, methods (base R only)
- Suggests: energy ≥ 1.7.10, ggplot2 ≥ 3.4.0, MASS, testthat, knitr