curves: Model-Agnostic Response Curves for Fitted Models

Create model-agnostic response-curve diagnostics for fitted prediction models. Supports profile curves, partial dependence, individual conditional expectation, and accumulated local effects; univariate curves, bivariate surfaces, ensemble summaries across multiple models, ALE-based interaction ranking, and optional raster-linked exploration with 'terra' and 'shiny'. Static displays are returned as 'ggplot2' plots. For more details on the methods see Molnar (2025) <https://christophm.github.io/interpretable-ml-book/>.

Version:	0.4.0
Depends:	R (≥ 3.5.0)
Imports:	cowplot, ggplot2 (≥ 3.3.6)
Suggests:	disdat, knitr, mgcv, plotly, randomForest, rmarkdown, shiny, terra (≥ 1.6-41), testthat (≥ 3.0.0)
Published:	2026-05-11
DOI:	10.32614/CRAN.package.curves
Author:	Roozbeh Valavi [aut, cre]
Maintainer:	Roozbeh Valavi <valavi.r at gmail.com>
BugReports:	https://github.com/rvalavi/curves/issues
License:	GPL-3
URL:	https://github.com/rvalavi/curves
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	curves results

Documentation:

Reference manual:	curves.html , curves.pdf
Vignettes:	Species Distribution Response Curves (source, R code)

Downloads:

Package source:	curves_0.4.0.tar.gz
Windows binaries:	r-devel: curves_0.4.0.zip, r-release: curves_0.4.0.zip, r-oldrel: curves_0.4.0.zip
macOS binaries:	r-release (arm64): curves_0.4.0.tgz, r-oldrel (arm64): curves_0.4.0.tgz, r-release (x86_64): curves_0.4.0.tgz, r-oldrel (x86_64): curves_0.4.0.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=curves to link to this page.