Skip to contents

Generate background points

Source: R/background_points.R
sample_background.Rd

A constructor function to generate background points for various purposes (e.g. computing evaluation scores for presence-only data). It exclude NA cells from the sample, and eventually the observed locations if needed.

Usage

sample_background(
  mask,
  points = NULL,
  n = 1000,
  method = "random",
  cells = FALSE,
  xy = TRUE,
  as.points = FALSE,
  na.rm = TRUE,
  ...
)

Arguments

mask

SpatRaster object to be used as mask (preferably, the predicted intensity or habitat suitability).

points

Spatial points (data.frame, sf or SpatVector objects) that would be excluded from the background sample.

n

integer. The number of pseudo-absence points to sample for presence-only data. The default is 1000.

method

character. The sampling technique to select pixels from the raster mask (see spatSample). It defaults to random.

cells

logical. If TRUE, sampled cells numbers will be returned. The default is FALSE.

xy

logical. If TRUE, the locations of sampled cells will be returned. The default is TRUE.

as.points

logical. If TRUE, spatial points object will be returned. The default is FALSE.

na.rm

logical. If TRUE, NA values will be excluded from the raster mask. It defaults to TRUE.

...

Additional arguments passed on to the internal spatSample function.

Value

An S3 object with class BackgroundPoints, containing the modified SpatRaster object and the generated background points.

See also

Other BackgroundPoints methods: BackgroundPoints-methods

Examples

if (FALSE) { # \dontrun{
library(terra)
set.seed(123)
r <- rast(nrows = 100, ncols = 100, xmin = 0, xmax = 10, ymin = 0, ymax = 10)
terra::values(r) <- runif(ncell(r))
pts <- spatSample(r, size = 100, xy = TRUE, values = FALSE)

# Requesting few points with their x and y coordinates
set.seed(235)
bg_sample1 <- sample_background(r, points = pts, n = 500, xy = TRUE, cells = FALSE)
plot(bg_sample1)
print(bg_sample1)

# Requesting points more than available non-NA cells
bg_sample2 <- sample_background(r, points = pts, n = 10000, xy = TRUE, cells = FALSE)
dim(bg_sample2$bg)
plot(bg_sample2)
} # }