Skip to contents

Calculate Multivariate Environmental Similarity

similarity.Rd

NOTICE: This function was borrowed from the rmaxent package written by John Baumgartner (https://github.com/johnbaums/rmaxent/).

Calculate Multivariate Environmental Similarity and most dissimilar/similar variables with respect to a reference dataset, for a set of environmental variables.

Usage

similarity(x, ref, full = FALSE)

Arguments

x

a `Raster*`, `list`, `matrix`, or `data.frame` where each layer/column/element represents focal values of an environmental variable.

ref

a `list`, `matrix`, or `data.frame` where each column/element represents reference values for an environmental variable (corresponding to those given in `x`).

full

(logical) should similarity values be returned for all variables? If `FALSE` (the default), then only the minimum similarity scores across variables will be returned.

Value

If `x` is a `Raster*` object, this function returns a list containing: - `similarity`: a `RasterStack` giving the environmental similarities for each variable in `x` (only included when `full=TRUE`); - `similarity_min`: a `Raster` layer giving the minimum similarity value across all variables for each location (i.e. the MESS); - `mod`: a factor `Raster` layer indicating which variable was most dissimilar to its reference range (i.e. the MoD map, Elith et al. 2010); and - `mos`: a factor `Raster` layer indicating which variable was most similar to its reference range.

If `x` is a `list`, `matrix`, or `data.frame`, the function will return a list as above, but with `RasterStack` and `Raster` objects replaced by matrix and vectors.

Details

`similarity` uses the MESS algorithm described in Appendix S3 of Elith et al. 2010.

References

Elith, J., Kearney, M., and Phillips, S. (2010) The art of modelling range-shifting species. Methods in Ecology and Evolution, 1: 330-342. doi:10.1111/j.2041-210X.2010.00036.x

Examples

library(dismo)
library(raster)
ff <- list.files(system.file('ex', package='dismo'), '\\.grd$', 
                 full.names=TRUE )
predictors <- stack(grep('biome', ff, value=TRUE, invert=TRUE))
occ <- read.csv(system.file('ex/bradypus.csv', package='dismo'))[, -1]
ref <- extract(predictors, occ)
mess <- similarity(predictors, ref, full=TRUE)

if (FALSE) {
library(rasterVis)
library(RColorBrewer)
levelplot(mess$mod, col.regions=brewer.pal(8, 'Set1'))
levelplot(mess$mos, col.regions=brewer.pal(8, 'Set1'))
}