gridsample: Draw a random sample from a grid laid out on a RasterLayer or...

In yinscapital/sat-locat-reference-team-lucc: TEAM land use and cover change data processing toolkit

Description

This function is used to subsample a RasterLayer or matrix by dividing the dataset into a grid of horizcells x vertcells, and by then drawing a sample of size nsamp from within each grid cell.

Usage

gridsample(x, horizcells = 10, vertcells = 10, nsamp = 10,
  rowmajor = FALSE, replace = FALSE)

Arguments

x	a matrix or RasterLayer to draw sample from
horizcells	how many cells to break the raster in horizontally (over the columns)
vertcells	how many cells to break the raster in vertically (over the rows)
nsamp	how many samples to draw from each grid cell
rowmajor	whether to return indices in row-major format (default is column-major). Row-major format is useful in conjunction with Raster* objects.
replace	whether to sample with replacement (within each grid cell)

Value

vector of sample indices

Note

TODO: Recode in C++ for speed.

Examples

# Make a 100x100 matrix
x <- matrix(1:10000, nrow=100)
# Subsample the matrix by breaking it up into a 10 x 10 grid, and take 10
# random samples from each grid cell without replacement (these are the
# default parameters).
y <- gridsample(x)

yinscapital/sat-locat-reference-team-lucc documentation built on May 14, 2019, 11:09 a.m.