rankimp: Consensus Ranking of Variable Importance with Uncertainty

Variable importance rankings depend on the method, the random seed and the resample used to compute them. This package treats every source of importance as a judge expressing a ranking over the predictors, and synthesises those rankings into a Kemeny median ranking with ties. Uncertainty about the consensus is quantified through bootstrap rank confidence sets, top-k probabilities and clustering of disagreeing judges.

Version:	1.0.0
Depends:	R (≥ 3.5)
Imports:	ConsRank, ggplot2, stats, tibble, utils
Suggests:	cluster, covr, kernelshap, knitr, randomForest, ranger, rmarkdown, rsample, stabm, testthat (≥ 3.1.7)
Published:	2026-09-17
DOI:	10.32614/CRAN.package.rankimp (may not be active yet)
Author:	Agostino Gnasso [aut, cre, cph]
Maintainer:	Agostino Gnasso <agostino.gnasso at unina.it>
BugReports:	https://github.com/agostinognasso/rankimp/issues
License:	MIT + file LICENSE
URL:	https://github.com/agostinognasso/rankimp
NeedsCompilation:	no
Language:	en-GB
Materials:	README, NEWS
CRAN checks:	rankimp results

Documentation:

Reference manual:	rankimp.html , rankimp.pdf
Vignettes:	Set stability is not rank stability (source, R code) Robustly important variables in credit scoring (source, R code) When XAI methods disagree (source, R code) Consensus ranking of variable importance (source, R code) Every function, and when you want it (source, R code) How stable is a variable importance ranking? (source, R code) The theory the package rests on (source, R code)

Downloads:

Package source:	rankimp_1.0.0.tar.gz
Windows binaries:	r-devel: not available, r-release: not available, r-oldrel: not available
macOS binaries:	r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): not available, r-oldrel (x86_64): not available

Linking:

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