bm_PseudoAbsences {biomod2}

Select pseudo-absences

Description

This internal biomod2 function allows to select pseudo-absences according to 4 different methods : random, sre, disk or user.defined (see Details).

Usage

bm_PseudoAbsences(
  resp.var,
  expl.var,
  nb.rep = 1,
  strategy = "random",
  nb.absences = NULL,
  sre.quant = 0,
  dist.min = 0,
  dist.max = NULL,
  user.table = NULL
)

bm_PseudoAbsences_user.defined(resp.var, expl.var, ...)

## S4 method for signature 'ANY,SpatVector'
bm_PseudoAbsences_user.defined(resp.var, expl.var, user.table)

## S4 method for signature 'ANY,SpatRaster'
bm_PseudoAbsences_user.defined(resp.var, expl.var, user.table)

bm_PseudoAbsences_random(resp.var, expl.var, ...)

## S4 method for signature 'ANY,SpatVector'
bm_PseudoAbsences_random(resp.var, expl.var, nb.absences, nb.rep)

## S4 method for signature 'ANY,SpatRaster'
bm_PseudoAbsences_random(resp.var, expl.var, nb.absences, nb.rep)

bm_PseudoAbsences_sre(resp.var, expl.var, ...)

## S4 method for signature 'ANY,SpatVector'
bm_PseudoAbsences_sre(resp.var, expl.var, sre.quant, nb.absences, nb.rep)

## S4 method for signature 'ANY,SpatRaster'
bm_PseudoAbsences_sre(resp.var, expl.var, sre.quant, nb.absences, nb.rep)

bm_PseudoAbsences_disk(resp.var, expl.var, ...)

## S4 method for signature 'ANY,SpatVector'
bm_PseudoAbsences_disk(
  resp.var,
  expl.var,
  dist.min,
  dist.max,
  nb.absences,
  nb.rep
)

## S4 method for signature 'ANY,SpatRaster'
bm_PseudoAbsences_disk(
  resp.var,
  expl.var,
  dist.min,
  dist.max,
  nb.absences,
  nb.rep
)

Arguments

resp.var

a vector, SpatialPoints or SpatialPointsDataFrame object containing binary data (0 : absence, 1 : presence, NA : indeterminate) for a single species that will be used to find the pseudo-absences

expl.var

a matrix, data.frame, SpatialPointsDataFrame or SpatRaster object containing the explanatory variables (in columns or layers) that will be used to find the pseudo-absences

nb.rep

an integer corresponding to the number of sets (repetitions) of pseudo-absence points that will be drawn

strategy

a character corresponding to the pseudo-absence selection strategy, must be among random, sre, disk or user.defined

nb.absences

(optional, default NULL)
If strategy = 'random' or strategy = 'sre' or strategy = 'disk', an integer corresponding to the number of pseudo-absence points that will be selected for each pseudo-absence repetition (true absences included)

sre.quant

(optional, default 0)
If strategy = 'sre', a numeric between 0 and 0.5 defining the half-quantile used to make the sre pseudo-absence selection (see bm_SRE)

dist.min

(optional, default 0)
If strategy = 'disk', a numeric defining the minimal distance to presence points used to make the disk pseudo-absence selection (in meters)

dist.max

(optional, default NULL)
If strategy = 'disk', a numeric defining the maximal distance to presence points used to make the disk pseudo-absence selection (in meters)

user.table

(optional, default NULL)
If strategy = 'user.defined', a matrix or data.frame with as many rows as resp.var values, as many columns as nb.rep, and containing TRUE or FALSE values defining which points will be used to build the species distribution model(s) for each repetition

...

(optional, one or several of the above arguments depending on the selected method)

Details

Concerning random selection :

The idea is to select pseudo-absences randomly in spatial locations where the species has not been sampled. This method is the simplest one and the most appropriate if lacking information about the presence sampling (non-exhaustive, biased sampling, etc).

Concerning SRE selection (see bm_SRE) :

The idea is to select pseudo-absences in spatial locations whose environmental conditions are different from those of the presence points. This method is appropriate when most of the environmental space of the species has been sampled.

Concerning disk selection :

The idea is to select pseudo-absences, not too close from presence points, but not too far away either. This method is appropriate when most of the spatial range of the species has been sampled.

Concerning user defined selection :

The user can provide pseudo-absences locations through a table containing spatial locations in rows, pseudo-absences repetitions in columns, and TRUE/FALSE values indicating whether each point is to be considered as pseudo-absence or not for each dataset.

Value

A list containing the following elements :

• xy : the coordinates of the species observations

• sp : the values of the species observations (0, 1 or NA)

• env : the explanatory variables

• pa.tab : the corresponding table of selected pseudo-absences (indicated by TRUE or FALSE)

Author(s)

Wilfried Thuiller, Damien Georges

See Also

bm_SRE, BIOMOD.formated.data.PA, BIOMOD_FormatingData

Other Secundary functions: bm_BinaryTransformation(), bm_CrossValidation(), bm_FindOptimStat(), bm_MakeFormula(), bm_ModelingOptions(), bm_PlotEvalBoxplot(), bm_PlotEvalMean(), bm_PlotRangeSize(), bm_PlotResponseCurves(), bm_PlotVarImpBoxplot(), bm_RunModelsLoop(), bm_SRE(), bm_SampleBinaryVector(), bm_SampleFactorLevels(), bm_Tuning(), bm_VariablesImportance()

Examples

library(terra)

# Load species occurrences (6 species available)
data(DataSpecies)
head(DataSpecies)

# Select the name of the studied species
myRespName <- 'GuloGulo'

# Get corresponding presence/absence data
myResp <- as.numeric(DataSpecies[, myRespName])

# Get corresponding XY coordinates
myRespXY <- DataSpecies[, c('X_WGS84', 'Y_WGS84')]

# Load environmental variables extracted from BIOCLIM (bio_3, bio_4, bio_7, bio_11 & bio_12)
data(bioclim_current)
myExpl <- terra::rast(bioclim_current)




# --------------------------------------------------------------- #
# Create the different pseudo-absence datasets

# Transform true absences into potential pseudo-absences
myResp.PA <- ifelse(myResp == 1, 1, NA)
myResp.PA.vect <- vect(cbind(myRespXY, myResp.PA), geom = c("X_WGS84","Y_WGS84"))

# random method
PA.r <- bm_PseudoAbsences(resp.var = myResp.PA.vect,
                          expl.var = myExpl,
                          nb.rep = 4,
                          nb.absences = 1000,
                          strategy = 'random')

# disk method
PA.d <- bm_PseudoAbsences(resp.var = myResp.PA.vect,
                          expl.var = myExpl,
                          nb.rep = 4,
                          nb.absences = 500,
                          strategy = 'disk',
                          dist.min = 5,
                          dist.max = 35)

# SRE method
PA.s <- bm_PseudoAbsences(resp.var = myResp.PA.vect,
                          expl.var = myExpl,
                          nb.rep = 4,
                          nb.absences = 1000,
                          strategy = 'sre',
                          sre.quant = 0.025)

# user.defined method
myPAtable <- data.frame(PA1 = ifelse(myResp == 1, TRUE, FALSE),
                        PA2 = ifelse(myResp == 1, TRUE, FALSE))
for (i in 1:ncol(myPAtable)) myPAtable[sample(which(myPAtable[, i] == FALSE), 500), i] = TRUE
PA.u <- bm_PseudoAbsences(resp.var = myResp.PA.vect,
                          expl.var = myExpl,
                          strategy = 'user.defined',
                          user.table = myPAtable)

str(PA.r)
head(PA.r$pa.tab)
apply(PA.r$pa.tab, 2, table)

head(PA.d$pa.tab)
apply(PA.d$pa.tab, 2, table)

head(PA.s$pa.tab)
apply(PA.s$pa.tab, 2, table)

tail(PA.u$pa.tab)
apply(PA.u$pa.tab, 2, table)


# random method : different number of PA
PA.r_mult <- bm_PseudoAbsences(resp.var = myResp.PA.vect,
                               expl.var = myExpl,
                               nb.rep = 4,
                               nb.absences = c(1000, 500, 500, 200),
                               strategy = 'random')

str(PA.r_mult)
head(PA.r_mult$pa.tab)
apply(PA.r_mult$pa.tab, 2, table)