plot.FeatureEffects: Plot FeatureEffect

In iml: Interpretable Machine Learning

Plot FeatureEffect

Description

plot.FeatureEffect() plots the results of a FeatureEffect object.

Usage

## S3 method for class 'FeatureEffects'
plot(x, features = NULL, nrows = NULL, ncols = NULL, fixed_y = TRUE, ...)

Arguments

x	A FeatureEffect object.
features	character For which features should the effects be plotted? Default is all features. You can also sort the order of the plots with this argument.
nrows	The number of rows in the table of graphics
ncols	The number of columns in the table of graphics
fixed_y	Should the y-axis range be the same for all effects? Defaults to TRUE.
...	Further arguments for FeatureEffect$plot()

Details

In contrast to other plot methods, for FeatureEffects the returned plot is not a ggplot2 object, but a grid object, a collection of multiple ggplot2 plots.

Value

grid object

See Also

FeatureEffects plot.FeatureEffect

Examples

# We train a random forest on the Boston dataset:
library("randomForest")
data("Boston", package = "MASS")
rf <- randomForest(medv ~ ., data = Boston, ntree = 50)
mod <- Predictor$new(rf, data = Boston)

# Compute the partial dependence for the first feature
eff <- FeatureEffects$new(mod)

# Plot the results directly
eff$plot()

# For a subset of features
eff$plot(features = c("lstat", "crim"))

# With a different layout
eff$plot(nrows = 2)

iml documentation built on May 29, 2024, 1:59 a.m.