break_down: Model Agnostic Sequential Variable Attributions

In ModelOriented/breakDown2: Model Agnostic Instance Level Variable Attributions

Model Agnostic Sequential Variable Attributions

Description

This function finds Variable Attributions via Sequential Variable Conditioning. It calls either local_attributions for additive attributions or local_interactions for attributions with interactions.

Usage

break_down(x, ..., interactions = FALSE)

## S3 method for class 'explainer'
break_down(x, new_observation, ..., interactions = FALSE)

## Default S3 method:
break_down(
  x,
  data,
  predict_function = predict,
  new_observation,
  keep_distributions = FALSE,
  order = NULL,
  label = class(x)[1],
  ...,
  interactions = FALSE
)

Arguments

x	an explainer created with function explain or a model.
...	parameters passed to local_* functions.
interactions	shall interactions be included?
new_observation	a new observation with columns that correspond to variables used in the model.
data	validation dataset, will be extracted from x if it is an explainer.
predict_function	predict function, will be extracted from x if it's an explainer.
keep_distributions	if TRUE, then distribution of partial predictions is stored and can be plotted with the generic plot().
order	if not NULL, then it will be a fixed order of variables. It can be a numeric vector or vector with names of variables.
label	name of the model. By default it is extracted from the 'class' attribute of the model.

Value

an object of the break_down class.

References

Explanatory Model Analysis. Explore, Explain and Examine Predictive Models. https://ema.drwhy.ai

See Also

local_attributions, local_interactions

Examples

library("DALEX")
library("iBreakDown")
set.seed(1313)
model_titanic_glm <- glm(survived ~ gender + age + fare,
                       data = titanic_imputed, family = "binomial")
explain_titanic_glm <- explain(model_titanic_glm,
                           data = titanic_imputed,
                           y = titanic_imputed$survived,
                           label = "glm")

bd_glm <- break_down(explain_titanic_glm, titanic_imputed[1, ])
bd_glm
plot(bd_glm, max_features = 3)

## Not run: 
## Not run:
library("randomForest")
set.seed(1313)
# example with interaction
# classification for HR data
model <- randomForest(status ~ . , data = HR)
new_observation <- HR_test[1,]

explainer_rf <- explain(model,
                        data = HR[1:1000,1:5])

bd_rf <- break_down(explainer_rf,
                           new_observation)
head(bd_rf)
plot(bd_rf)

## End(Not run)

ModelOriented/breakDown2 documentation built on Jan. 5, 2024, 5:29 a.m.