varImpScores: Variable importances as defined by Hernandez et al. (2018)

In bartBMA: Bayesian Additive Regression Trees using Bayesian Model Averaging

Description

This measure defines the importance of a variable as the model-probability weighted sum of the number of splits on the variable of interest, divided by the sum over all variables of such weighted counts of splits.

Usage

varImpScores(object)

Arguments

object

A bartBMA object obtained using the barBMA function.

Value

A vector of variable importances. The variables are ordered in the same order that they occur in columns of the input covariate matrix used to obtain the input bartBMA object.

Examples

#set the seed
set.seed(100)
#simulate some data
N <- 100
p<- 100
epsilon <- rnorm(N)
xcov <- matrix(runif(N*p), nrow=N)
y <- sin(pi*xcov[,1]*xcov[,2]) + 20*(xcov[,3]-0.5)^2+10*xcov[,4]+5*xcov[,5]+epsilon
epsilontest <- rnorm(N)
xcovtest <- matrix(runif(N*p), nrow=N)
ytest <- sin(pi*xcovtest[,1]*xcovtest[,2]) + 20*(xcovtest[,3]-0.5)^2+10*xcovtest[,4]+
  5*xcovtest[,5]+epsilontest

#Train the object 
bart_bma_example <- bartBMA(x.train = xcov,y.train=y,x.test=xcovtest,zero_split = 1, 
                            only_max_num_trees = 1,split_rule_node = 0)
#Obtain the variable importances
varImpScores(bart_bma_example)

bartBMA documentation built on March 13, 2020, 5:06 p.m.