Boruta (version 1.6)

plotZHistory: Plotting history of importance

Description

plotZHistory visualises ZScoreHistory element of the Boruta object in some more straightforward way than plot.Boruta does. Namely, it plots each attributes' (including randomised meta-attributes, randMax, randMean, randMin) importance as a function of Random Forest run it was obtained in. The colour of each line corresponds to the final Boruta decision; see colCode argument note.

Usage

plotZHistory(x,colCode=c('green','yellow','red','blue'),showRounds=TRUE,
		col=NULL,type="l",lty=1,pch=0,
		xlab='Random Forest run',ylab='Importance',...)

Arguments

x

an object of class Boruta.

colCode

a vector containing colour codes for attribute decisions, respectively Confirmed, Tentative, Rejected, Random.

showRounds

if set, to TRUE, gray lines separating round will be drawn.

col

standard col attribute. If given, suppresses colCode.

type,lty,pch,xlab,ylab,...

additional graphical parameters that will be passed to matplot.

Value

• None.

Examples

library(mlbench); data(Sonar);
#Takes some time, so be patient
Boruta(Class~.,data=Sonar,doTrace=2)->Bor.son;
print(Bor.son);
plotZHistory(Bor.son);