auditor
modeloriented.github.io
- Free plan
- Windows
- macOS
- Linux
7.1 Club score Visit modeloriented.github.io
What the maker publishes
- Analysis
- It supports assessing model goodness of fit and performance, comparing residual similarity, and identifying outliers and influential observations. source
- Installation
- The site lists CRAN installation and a developer installation from GitHub. source
- Interactive plots
- Several plot types have interactive versions, including autocorrelation, influence, prediction, REC, and residual plots. source
- License
- The package is listed under the GPL license. source
- Model-agnostic methods
- The residual audit vignette says most methods can be used regardless of the expected residual distribution. source
- Model compatibility
- The site describes the package as usable to validate models of any classes. source
- Plots
- Available plot types include autocorrelation, influence, lift, model correlation, PCA, ROC, and residual plots. source
- Prerequisite workflow
- The demonstrated workflow creates an explainer using DALEX before constructing auditor model-residual objects. source
- Purpose
- auditor is an R package for model-agnostic validation, model comparison, and error analysis. source
- Regression and classification
- The plot overview marks supported plots for regression and classification; LIFT and ROC are marked classification-only. source
- Residual analysis
- The residual audit vignette says the package helps answer questions about model fit, relative performance, and model similarity. source
- Support
- The site directs users to GitHub issues to report bugs. source
- Validation approach
- Model examination uses diagnostic scores and visual verification. source
Checked against the maker's pages on 14 Nov 2026.