plotRule: Plot values for a rule.

In mategarb/R.ROSETTA: R.ROSETTA: an interpretable machine learning framework

Description

Represent rule as a heatmap or boxplot.

Usage

plotRule(dt, rules, type="heatmap", discrete=FALSE, ind=1, label=c())

Arguments

dt	A data frame of decision table. The last column is decision.
rules	A data frame of rule table in a form of rosetta $main output.
discrete	Logical. Set TRUE for non-discrete data. Default is FALSE.
type	A character indciating plot type. Heatmap or boxplot.
ind	A numeric value of index of the rule in the table.
label	A character vector indciating new labels for the rule discrete levels.

Value

heatmap	A heatmap representing objects distribution in the model for the rule.
boxplot	A boxplots representing objects distribution in the model for the rule.

Author(s)

Mateusz Garbulowski

Examples

library(R.ROSETTA)

out <- rosetta(autcon)
rules <- out$main
recRules <- recalculateRules(autcon, rules)

# heatmap
plotRule(autcon, recRules, type="heatmap", discrete=FALSE, ind=1, label = c("high","low"))
plotRule(autcon, recRules, type="heatmap", discrete=FALSE, ind=7, label = c("high","high"))


# boxplot
plotRule(autcon, recRules, type="boxplot", discrete=FALSE, ind=1, label = c("high","low"))
plotRule(autcon, recRules, type="boxplot", discrete=FALSE, ind=7, label = c("high","high"))

mategarb/R.ROSETTA documentation built on April 2, 2021, 12:28 a.m.