shapr: Prediction Explanation with Dependence-Aware Shapley Values

Complex machine learning models are often hard to interpret. However, in many situations it is crucial to understand and explain why a model made a specific prediction. Shapley values is the only method for such prediction explanation framework with a solid theoretical foundation. Previously known methods for estimating the Shapley values do, however, assume feature independence. This package implements methods which accounts for any feature dependence, and thereby produces more accurate estimates of the true Shapley values. An accompanying 'Python' wrapper ('shaprpy') is available through the GitHub repository.

Version: 1.0.3
Depends: R (≥ 3.5.0)
Imports: stats, data.table (≥ 1.15.0), Rcpp (≥ 0.12.15), Matrix, future.apply, methods
LinkingTo: RcppArmadillo, Rcpp
Suggests: ranger, xgboost, mgcv, testthat (≥ 3.0.0), knitr, rmarkdown, roxygen2, ggplot2, gbm, party, partykit, waldo, progressr, future, ggbeeswarm, vdiffr, forecast, torch, GGally, coro, parsnip, recipes, workflows, tune, dials, yardstick, hardhat, rsample, rlang, cli
Published: 2025-03-26
DOI: 10.32614/CRAN.package.shapr
Author: Martin Jullum ORCID iD [cre, aut], Lars Henry Berge Olsen ORCID iD [aut], Annabelle Redelmeier [aut], Jon Lachmann ORCID iD [aut], Nikolai Sellereite ORCID iD [aut], Anders Løland [ctb], Jens Christian Wahl [ctb], Camilla Lingjærde [ctb], Norsk Regnesentral [cph, fnd]
Maintainer: Martin Jullum <Martin.Jullum at nr.no>
BugReports: https://github.com/NorskRegnesentral/shapr/issues
License: MIT + file LICENSE
URL: https://norskregnesentral.github.io/shapr/, https://github.com/NorskRegnesentral/shapr/
NeedsCompilation: yes
Language: en-US
Citation: shapr citation info
Materials: README NEWS
In views: MachineLearning
CRAN checks: shapr results [issues need fixing before 2025-04-29]

Documentation:

Reference manual: shapr.pdf
Vignettes: Asymmetric and causal Shapley value explanations (source)
'shapr': Explaining individual machine learning predictions with Shapley values (source)
Shapley value explanations using the regression paradigm (source)
More details and advanced usage of the 'vaeac' approach (source)

Downloads:

Package source: shapr_1.0.3.tar.gz
Windows binaries: r-devel: shapr_1.0.3.zip, r-release: shapr_1.0.3.zip, r-oldrel: shapr_1.0.3.zip
macOS binaries: r-release (arm64): shapr_1.0.3.tgz, r-oldrel (arm64): shapr_1.0.3.tgz, r-release (x86_64): shapr_1.0.3.tgz, r-oldrel (x86_64): shapr_1.0.3.tgz
Old sources: shapr archive

Reverse dependencies:

Reverse suggests: XAItest

Linking:

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