gini_process: computes Gini index
In rfVarImpOOB: Unbiased Variable Importance for Random Forests

gini_process

R Documentation

computes Gini index

Description

computes Gini index

Usage

gini_process(classes, splitvar = NULL)

Arguments

`classes`	vector of factors/categorical vars
`splitvar`	split variable

Value

Gini index

Author(s)

Markus Loecher <Markus.Loecher@gmail.com>

Examples



#Test binary case:





#50/50split


gini_process(c(rep(0,10),rep(1,10)))#0.5 CORRECT !


#10/90split


gini_process(c(rep(0,1),rep(1,9)))#0.18= CORRECT ! 


#0/100split


gini_process(factor(c(rep(0,0),rep(1,10)), levels=c(0,1)))#0








#Test binary case:





#25/25/25/25 split


gini_process(factor(c(rep(0,5),rep(1,5),rep(2,5),


                      rep(3,5)), levels=c(0:3)))#0.75 = 4*0.25*0.75 CORRECT !


#10/10/10/70 split


gini_process(factor(c(rep(0,1),rep(1,1),rep(2,1),


                      rep(3,7)), levels=c(0:3)))#0.48 = 3*0.1*0.9+0.7*0.3  CORRECT !


#0/0/0/100 split


gini_process(factor(c(rep(0,0),rep(1,0),rep(2,0),


                      rep(3,20)), levels=c(0:3)))#0. CORRECT !

rfVarImpOOB documentation built on July 1, 2022, 5:05 p.m.