cpath

Explaining and visualizing black-box models through counterfactual paths

Paper: https://arxiv.org/abs/2307.07764.

Software documentation: https://github.com/pievos101/cpath/blob/main/cpath_documentation.pdf.

Installation

The {cpath} R-package can be installed using {devtools}.

```r install.packages("devtools") library(devtools)

devtools::install_github("pievos101/cpath") library(cpath)

```

Usage

Lets apply {cpath} on a random forest model.

```r library(caret) library(cpath)

Generate simulated data

res = sim() data = res$data target = as.factor(res$target)

Train-test split

80% of the sample size

smp_size <- floor(0.80 * nrow(data))

trainind <- sample(seqlen(nrow(data)), size = smp_size)

train <- data[trainind, ] test <- data[-trainind, ]

targettrain = as.factor(target[trainind]) targettest = as.factor(target[-trainind])

TRAIN = as.data.frame(cbind(train, targettrain)) TRAIN$targettrain = as.factor(TRAIN$target_train)

10 folds repeat 3 times

control <- trainControl(method='repeatedcv', number=10, repeats=3)

Metric compare model is Accuracy

metric <- "Accuracy" set.seed(123)

Number randomly variable selected is mtry

mtry <- sqrt(ncol(train)) tunegrid <- expand.grid(.mtry=mtry) model <- train(target_train~., data=TRAIN, method='rf', metric='Accuracy', tuneGrid=tunegrid, trControl=control) print(model)

Get the counterfactual paths

P = cpath::cpaths(model, test, k=4, n_paths = 1000)

P = cpath::cpathsmc(model, test, k=4, npaths = 1000) #multi-core

Build transition matrix

T = cpath::transition(P)

Get global feature importances

IMP = cpath::importance(T) print(IMP)

Get summary of the counterfactual paths

cpathsummary = getcpath_summary(P)

Plot the paths

plotpaths(cpathsummary)

```

Citation

If you find {cpath} useful please cite our paper:

@article{pfeifer2023explaining, title = {Explaining and visualizing black-box models through counterfactual paths}, author = {Bastian Pfeifer and Mateusz Krzyzinski and Hubert Baniecki and Anna Saranti and Andreas Holzinger and Przemyslaw Biecek}, year = {2023}, journal = {arXiv preprint, arXiv:2307.07764} }