lime_extractor: LIME wrapper

In mstaniak/egalitaRian: Compare Validation and Training Datasets and Explore Differences

Description

LIME wrapper

Usage

lime_extractor(instance, target, training_set, model,
  predict_function = predict, ...)

Arguments

instance	data frame that contains the observation to explain
target	name of the response variable
training_set	data on which the explained 'model“ was trained
model	model to be explained
predict_function	predict function for 'model' which returns classes. Defaults to 'predict'.
...	additional parameters to 'lime::explain' function

Value

list that consists of

explanation_model	object returned by 'lime::explain' function
variables	character vector of names of variables that contributed to the wrong prediction

mstaniak/egalitaRian documentation built on Aug. 26, 2019, 11:11 p.m.