🔍 Make your ML models more interpretable with beautiful variable importance plots
vip provides a unified framework for constructing variable importance plots from virtually any machine learning model in R. Stop juggling different importance() functions across packages – vip gives you one consistent interface for interpretable ML.
- Universal interface: Works with 40+ model types from different packages
- Multiple methods: Model-specific, permutation, SHAP, and variance-based importance
- Beautiful plots: Publication-ready visualizations with ggplot2
- Efficient algorithms: Optimized with parallel processing support
- Extensible design: Easy to add support for new model types
- Well-documented: Comprehensive guides and academic backing
# Install from CRAN (stable)
install.packages("vip")
# Install development version (latest features)
# install.packages("pak")
pak::pak("koalaverse/vip")library(vip)
library(randomForest)
# Fit a model
model <- randomForest(Species ~ ., data = iris)
# Get importance scores
vi_scores <- vi(model)
print(vi_scores)
#> # A tibble: 4 × 2
#> Variable Importance
#> <chr> <dbl>
#> 1 Petal.Length 32.4
#> 2 Petal.Width 31.3
#> 3 Sepal.Length 9.51
#> 4 Sepal.Width 6.75
# Create a beautiful plot
vip(model)| Method | Description | Use case | Function |
|---|---|---|---|
| Model-specific | Extract built-in importance | Fast, model-native | vi(model, method = "model") |
| Permutation | Shuffle features, measure impact | Model-agnostic, robust | vi(model, method = "permute") |
| Shapley values | Game theory attribution | Detailed explanations | vi(model, method = "shap") |
| Variance-based | FIRM approach | Feature ranking | vi(model, method = "firm") |
Tree-based models
Linear models
Neural networks
Meta-frameworks
Specialized models
library(ranger)
# Fit model
rf_model <- ranger(mpg ~ ., data = mtcars, importance = "none")
# Permutation importance with custom metric
vi_perm <- vi(
rf_model,
method = "permute",
train = mtcars,
target = "mpg",
metric = "rmse",
nsim = 50, # 50 permutations for stability
parallel = TRUE # Speed up with parallel processing
)
# Create enhanced plot
vip(vi_perm, num_features = 10, geom = "point") +
labs(title = "Permutation-based Variable Importance",
subtitle = "RMSE metric, 50 permutations") +
theme_minimal()library(xgboost)
# Prepare data
X <- data.matrix(subset(mtcars, select = -mpg))
y <- mtcars$mpg
# Fit XGBoost model
xgb_model <- xgboost(data = X, label = y, nrounds = 100, verbose = 0)
# SHAP-based importance
vi_shap <- vi(
xgb_model,
method = "shap",
train = X,
nsim = 30
)
# Beautiful SHAP plot
vip(vi_shap, geom = "col", aesthetics = list(fill = "steelblue", alpha = 0.8)) +
labs(title = "SHAP-based Variable Importance") +
theme_light()We welcome contributions! Here's how to get involved:
# Clone the repo
git clone https://github.com/koalaverse/vip.git
cd vip
# Open in RStudio or your favorite editor
# The project includes a comprehensive CLAUDE.md development guideWe use tinytest for lightweight, reliable testing:
# Run all tests
tinytest::test_package("vip")
# Test specific functionality
tinytest::run_test_file("inst/tinytest/test_vip.R")- Check issues: Look for good first issues
- Create branch:
git checkout -b feature/awesome-feature - Write tests: Follow TDD principles (see
CLAUDE.md) - Run checks:
R CMD checkand tests - Submit PR: With clear description
Adding support for new models is straightforward:
# Add S3 method to R/vi_model.R
vi_model.your_model_class <- function(object, ...) {
# Extract importance from your model
importance_scores <- your_model_importance_function(object)
# Return as tibble
tibble::tibble(
Variable = names(importance_scores),
Importance = as.numeric(importance_scores)
)
}See CLAUDE.md for detailed instructions!
- Package website - Comprehensive documentation
- R Journal paper - Academic foundation
- IML book - Theory background
- Development guide - Contributing guidelines
- ✅ ggplot2 S7 compatibility - Future-proof plotting
- lightgbm support - Popular gradient boosting
- Enhanced yardstick integration - Better metrics
- Improved documentation - Clearer examples
See NEWS.md for complete version history and migration notes.
- Bug reports: GitHub Issues
- Feature requests: GitHub Discussions
- Questions: Stack Overflow (tag:
vip)
If you use vip in your research, please cite:
@article{greenwell2020variable,
title={Variable Importance Plots—An Introduction to the vip Package},
author={Greenwell, Brandon M and Boehmke, Bradley C},
journal={The R Journal},
volume={12},
number={1},
pages={343--366},
year={2020},
doi={10.32614/RJ-2020-013}
}GPL (>= 2) © Brandon M. Greenwell, Brad Boehmke
⭐ Star us on GitHub if vip helps make your models more interpretable! ⭐
Built with ❤️ by the koalaverse team