niche.Model.Build {NicheBarcoding}

Ecological niche model building using the randomForest classifier

Description

Build a niche model for a given species according to its distribution data.

Usage

niche.Model.Build(
  prese = NULL,
  absen = NULL,
  prese.env = NULL,
  absen.env = NULL,
  model = "RF",
  en.vir = NULL,
  bak.vir = NULL
)

Arguments

prese

Data frame, longitude and latitude of the present data of a species (can be absent when providing prese.env parameter).

absen

Data frame, longitude and latitude of the absent data of a species.(can be absent when providing absen.env or back parameter).

prese.env

Data frame, bioclimate variables of present data. (can be absent when providing prese parameter).

absen.env

Data frame, bioclimate variables of absent data. (can be absent when providing absen or back parameter).

model

Character, string indicating which niche model will be used. Must be one of "RF" (default) or "MAXENT". "MAXENT" can only be applied when the java program paste(system.file(package="dismo"), "/java/maxent.jar", sep=”) exists.

en.vir

RasterBrick, the global bioclimate data output from "raster::getData" function.

bak.vir

Matrix, bioclimate variables of random background points.

Value

randomForest/MaxEnt, a trained niche model object.

A vector including the specificity, sensitivity and threshold of the trained model.

Author(s)

Cai-qing YANG (Email: yangcq_ivy(at)163.com) and Ai-bing ZHANG (Email:zhangab2008(at)cnu.edu.cn), Capital Normal University (CNU), Beijing, CHINA.

References

Breiman, L. 2001. Random forests. Machine Learning 45(1):5-32.

Liaw, A. and M. Wiener. 2002. Clasification and regression by randomForest. R News, 2/3:18-22.

Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones and A. Jarvis. 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology, 25(15):1965-1978.

Examples

data(en.vir)
data(bak.vir)
#envir<-raster::getData("worldclim",download=FALSE,var="bio",res=2.5)
#en.vir<-raster::brick(envir)
#back<-dismo::randomPoints(mask=en.vir,n=5000,ext=NULL,extf=1.1,
#                          excludep=TRUE,prob=FALSE,
#                          cellnumbers=FALSE,tryf=3,warn=2,
#                          lonlatCorrection=TRUE)
#bak.vir<-raster::extract(en.vir,back)

data<-data.frame(species=rep("Acosmeryx anceus",3),
                 Lon=c(145.380,145.270,135.461),
                 Lat=c(-16.4800,-5.2500,-16.0810))
present.points<-pseudo.present.points(data,10,2,1,en.vir)
NMB.out<-niche.Model.Build(prese=present.points,absen=NULL,
                           prese.env=NULL,absen.env=NULL,
                           model="RF",
                           en.vir=en.vir,bak.vir=bak.vir)
NMB.out


prese.env<-raster::extract(en.vir,present.points[,2:3])
prese.env<-as.data.frame(prese.env)
NMB.out2<-niche.Model.Build(prese=NULL,absen=NULL,
                            prese.env=prese.env,absen.env=NULL,
                            model="RF",
                            en.vir=en.vir,bak.vir=bak.vir)
NMB.out2