Creating explanations

These functions are the bread and butter of lime and is used to create an explainer from a model and apply it to an observation.

lime()

Create a model explanation function based on training data

explain()

Explain model predictions

predict_model() model_type()

Methods for extending limes model support

as_classifier() as_regressor()

Indicate model type to lime

Investigating explanations

While an explanation can be inspected through its tabular output format it is often much more powerful through different visualisations.

plot_features()

Plot the features in an explanation

plot_explanations()

Plot a condensed overview of all explanations

plot_image_explanation()

Display image explanations as superpixel areas

plot_text_explanations()

Plot text explanations

interactive_text_explanations() text_explanations_output() render_text_explanations()

Interactive explanations

Miscellaneous

This set of functions are sometimes needed in more specialised tasks

default_tokenize()

Default function to tokenize

plot_superpixels()

Test super pixel segmentation

stop_words_sentences

Stop words list

test_sentences

Sentence corpus - test part

train_sentences

Sentence corpus - train part

lime-package

lime: Local Interpretable Model-Agnostic Explanations