ecospat (version 3.2)

ecospat.maxentvarimport: Maxent Variable Importance

Description

Calculate the importance of variables for Maxent in the same way Biomod does, by randomly permuting each predictor variable independently, and computing the associated reduction in predictive performance.

Usage

ecospat.maxentvarimport (model, dfvar, nperm)

Arguments

model

The name of the maxent model.

dfvar

A dataframe object with the environmental variables.

nperm

The number of permutations in the randomization process. The default is 5.

Value

a list which contains a data.frame containing variables importance scores for each permutation run.

Details

The calculation is made as biomod2 "variables_importance" function. It's more or less base on the same principle than randomForest variables importance algorithm. The principle is to shuffle a single variable of the given data. Make model prediction with this 'shuffled' data.set. Then we compute a simple correlation (Pearson's by default) between references predictions and the 'shuffled' one. The return score is 1-cor(pred_ref,pred_shuffled). The highest the value, the more influence the variable has on the model. A value of this 0 assumes no influence of that variable on the model. Note that this technique does not account for interactions between the variables.

Examples

# NOT RUN {
library(dismo)
data('ecospat.testData')

# data for Soldanella alpina
data.Solalp<- ecospat.testData[c("Soldanella_alpina","ddeg","mind","srad","slp","topo")]

# copy maxent.jar file in the right folder
path.from<-system.file("extdata", "maxent.txt", package="ecospat")
path.to <- paste0(system.file(package="dismo"), "/java/maxent.txt")
path.to.renamed <- paste0(system.file(package="dismo"), "/java/maxent.jar")
file.copy(path.from,path.to,overwrite = TRUE)
file.rename(path.to, path.to.renamed)

if (file.exists(path.to.renamed) & require(rJava)) {
  me <- maxent(data.Solalp[,-1],data.Solalp[,1])
  ecospat.maxentvarimport (model=me, dfvar=data.Solalp[,-1], nperm=5)
  }
# }