pdp (version 0.8.1)

topPredictors: Extract Most "Important" Predictors (Experimental)

Description

Extract the most "important" predictors for regression and classification models.

Usage

topPredictors(object, n = 1L, ...)
# S3 method for default
topPredictors(object, n = 1L, ...)
# S3 method for train
topPredictors(object, n = 1L, ...)

Arguments

object

A fitted model object of appropriate class (e.g., "gbm", "lm", "randomForest", etc.).

n

Integer specifying the number of predictors to return. Default is 1 meaning return the single most important predictor.

...

Additional optional arguments to be passed onto varImp.

Details

This function uses the generic function varImp to calculate variable importance scores for each predictor. After that, they are sorted at the names of the n highest scoring predictors are returned.

Examples

# NOT RUN {
#
# Regression example (requires randomForest package to run)
#

Load required packages
library(ggplot2)
library(randomForest)

# Fit a random forest to the mtcars dataset
data(mtcars, package = "datasets")
set.seed(101)
mtcars.rf <- randomForest(mpg ~ ., data = mtcars, mtry = 5, importance = TRUE)

# Topfour predictors
top4 <- topPredictors(mtcars.rf, n = 4)

# Construct partial dependence functions for top four predictors
pd <- NULL
for (i in top4) {
  tmp <- partial(mtcars.rf, pred.var = i)
  names(tmp) <- c("x", "y")
  pd <- rbind(pd,  cbind(tmp, predictor = i))
}

# Display partial dependence functions
ggplot(pd, aes(x, y)) +
  geom_line() +
  facet_wrap(~ predictor, scales = "free") +
  theme_bw() +
  ylab("mpg")

# }