pctCorrect {rfPermute}

Percent Correctly Classified

Description

For classification models, calculate the percent of individuals correctly classified in a specified percent of trees in the forest.

Usage

pctCorrect(x, pct = c(seq(0.8, 0.95, 0.05), 0.99))

Arguments

x

a rfPermte or randomForest model object.

pct

vector of minimum percent of trees voting for each class. Can be 0:1 or 0:100.

Value

a matrix giving the percent of individuals correctly classified in each class and overall for each threshold value specified in pct.

Author(s)

Eric Archer [e-mail]

Examples

library(randomForest)
data(mtcars)

rf <- randomForest(factor(am) ~ ., mtcars, importance = TRUE)
pctCorrect(rf)