ensemble: Methods to assemble multiple algorithms in an ensemble SDM

Description

This is a method to assemble several algorithms in an ensemble SDM. The function takes as inputs several S4 Algorithm.SDM class objects returned by the modelling function. The function returns an S4 Ensemble.SDM class object containing the habitat suitability map, the binary map, and the uncertainty map (based on the between-algorithm variance) and the associated evaluation tables (model evaluation, algorithm evaluation, algorithm correlation matrix and variable importance).

Usage

ensemble(x, ..., name = NULL, ensemble.metric = c("AUC"),
  ensemble.thresh = c(0.75), weight = TRUE, thresh = 1001,
  uncertainty = TRUE, verbose = TRUE, GUI = FALSE)
# S4 method for Algorithm.SDM
ensemble(x, ..., name = NULL,
  ensemble.metric = c("AUC"), ensemble.thresh = c(0.75),
  weight = TRUE, thresh = 1001, uncertainty = TRUE, verbose = TRUE,
  GUI = FALSE)
# S4 method for Algorithm.SDM
sum(x, ..., name = NULL,
  ensemble.metric = c("AUC"), ensemble.thresh = c(0.75),
  weight = TRUE, thresh = 1001, format = TRUE, verbose = TRUE,
  na.rm = TRUE)

Value

an S4 Ensemble.SDM class object viewable with the

plot.model function.

Arguments

x, ...: SDMs. SDMs to be assembled.
name: character. Optional name given to the final Ensemble.SDM produced (by default 'Ensemble.SDM').
ensemble.metric: character. Metric(s) used to select the best SDMs that will be included in the ensemble SDM (see details below).
ensemble.thresh: numeric. Threshold(s) associated with the metric(s) used to compute the selection.
weight: logical. If TRUE, SDMs are weighted using the ensemble metric or, alternatively, the mean of the selection metrics.
thresh: numeric. A integer value specifying the number of equal interval threshold values between 0 and 1.
uncertainty: logical. If TRUE, generates an uncertainty map and an algorithm correlation matrix.
verbose: logical. If set to true, allows the function to print text in the console.
GUI, format, na.rm: logical. Do not take those arguments into account (parameters for the user interface and sum function).

Details

ensemble.metric (metric(s) used to select the best SDMs that will be included in the ensemble SDM) can be chosen from among:

AUC: Area under the receiver operating characteristic (ROC) curve
Kappa: Kappa from the confusion matrix
sensitivity: Sensitivity from the confusion matrix
specificity: Specificity from the confusion matrix
prop.correct: Proportion of correctly predicted occurrences from the confusion matrix

Examples

Run this code

if (FALSE) {
# Loading data
data(Env)
data(Occurrences)
Occurrences <- subset(Occurrences, Occurrences$SPECIES == 'elliptica')

# ensemble SDM building
CTA <- modelling('CTA', Occurrences, Env, Xcol = 'LONGITUDE', Ycol = 'LATITUDE')
SVM <- modelling('SVM', Occurrences, Env, Xcol = 'LONGITUDE', Ycol = 'LATITUDE')
ESDM <- ensemble(CTA, SVM, ensemble.thresh = c(0.6))

# Results plotting
plot(ESDM)
}