attStats: Extract attribute statistics

In Boruta: Wrapper Algorithm for All Relevant Feature Selection

Extract attribute statistics

Description

attStats shows a summary of a Boruta run in an attribute-centred way. It produces a data frame containing some importance stats as well as the number of hits that attribute scored and the decision it was given.

Usage

attStats(x)

Arguments

x	an object of a class Boruta, from which attribute stats should be extracted.

Value

A data frame containing, for each attribute that was originally in information system, mean, median, maximal and minimal importance, number of hits normalised to number of importance source runs performed and the decision copied from finalDecision.

Note

When using a Boruta object generated by a TentativeRoughFix, the resulting data frame will consist a rough-fixed decision.

x has to be made with holdHistory set to TRUE for this code to run.

Examples

## Not run: 
library(mlbench); data(Sonar)
#Takes some time, so be patient
Boruta(Class~.,data=Sonar,doTrace=2)->Bor.son
print(Bor.son)
stats<-attStats(Bor.son)
print(stats)
plot(normHits~meanImp,col=stats$decision,data=stats)

## End(Not run)

Boruta documentation built on Nov. 12, 2022, 9:06 a.m.