martinju / lime

Local Interpretable Model-Agnostic Explanations (R port of original Python package)

lime

This is an R port of the Python lime package (https://github.com/marcotcr/lime) developed by the authors of the lime (Local Interpretable Model-agnostic Explanations) approach for black-box model explanations. All credits for the invention of the approach goes to the original developers.

The purpose of lime is to explain the predictions of black box classifiers. What this means is that for any given prediction and any given classifier it is able to determine a small set of features in the original data that has driven the outcome of the prediction. To learn more about the methodology of lime read the paper and visit the repository of the original implementation.

The lime package for R does not aim to be a line-by-line port of its Python counterpart. Instead it takes the ideas laid out in the original code and implements them in an API that is idiomatic to R.

An example

Out of the box lime supports models created using the caret and mlr frameworks. Support for other models are easy to achieve by adding a predict_model and model_type method for the given model.

The following shows how a random forest model is trained on the iris data set and how lime is then used to explain a set of new observations:

library(caret)
library(lime)

# Split up the data set
iris_test <- iris[1:5, 1:4]
iris_train <- iris[-(1:5), 1:4]
iris_lab <- iris[[5]][-(1:5)]

# Create Random Forest model on iris data
model <- train(iris_train, iris_lab, method = 'rf')

# Create an explainer object
explainer <- lime(iris_train, model)

# Explain new observation
explanation <- explain(iris_test, explainer, n_labels = 1, n_features = 2)

# The output is provided in a consistent tabular format and includes the
# output from the model.
head(explanation)
#>       model_type case  label label_prob  model_r2 model_intercept
#> 1 classification    1 setosa          1 0.4341912       0.2420850
#> 2 classification    1 setosa          1 0.4341912       0.2420850
#> 3 classification    2 setosa          1 0.4424828       0.2341992
#> 4 classification    2 setosa          1 0.4424828       0.2341992
#> 5 classification    3 setosa          1 0.4446400       0.2291437
#> 6 classification    3 setosa          1 0.4446400       0.2291437
#>   model_prediction      feature feature_value feature_weight
#> 1        0.7016979  Sepal.Width           3.5   -0.015758934
#> 2        0.7016979  Petal.Width           0.2    0.475371862
#> 3        0.7082935 Sepal.Length           4.9   -0.008450071
#> 4        0.7082935  Petal.Width           0.2    0.482544423
#> 5        0.7211259  Sepal.Width           3.2    0.005927646
#> 6        0.7211259  Petal.Width           0.2    0.486054549
#>               feature_desc               data prediction
#> 1        3.3 < Sepal.Width 5.1, 3.5, 1.4, 0.2    1, 0, 0
#> 2       Petal.Width <= 0.4 5.1, 3.5, 1.4, 0.2    1, 0, 0
#> 3      Sepal.Length <= 5.2 4.9, 3.0, 1.4, 0.2    1, 0, 0
#> 4       Petal.Width <= 0.4 4.9, 3.0, 1.4, 0.2    1, 0, 0
#> 5 3.0 < Sepal.Width <= 3.3 4.7, 3.2, 1.3, 0.2    1, 0, 0
#> 6       Petal.Width <= 0.4 4.7, 3.2, 1.3, 0.2    1, 0, 0

# And can be visualised directly
plot_features(explanation)

lime also supports explaining text model and putting the explanation in the context of the original text input. It even includes a shiny application for interactively exploring text models:

Installation

lime is still a work in progress and is thus not available on CRAN yet. In order to try it out install it directly from GitHub:

# install.packages('devtools')
devtools::install_github('thomasp85/lime')

Scope

The current version of lime has support for tabular and text data. The Python implementation has additional support for image data, which will be added to this package in time. In addition to the capabilities discussed in the "Why Should I Trust You?": Explaining the Predictions of Any Classifier article, this package also support regression model explanations (this has been added to the Python library as well). The global model explanation using submodular picks that the article discusses is not supported in either packages. It might get support once it appears in the Python version and it is clear how the authors envision it.