bestMatch — bestMatch • terra

Determine for each grid cell which reference it is most similar to. A reference consists of a SpatVector with reference locations, or a data.frame or matrix in which each column matches a layer name in the SpatRaster.

Similarity is computed with the mean absolute or the mean squared differences between the cell and the reference, or with an alternative function you provide. It may be important to first scale the input.

Usage

# S4 method for class 'SpatRaster,SpatVector'
bestMatch(x, y, labels=NULL, fun="squared", ..., filename="", overwrite=FALSE, wopt=list())

# S4 method for class 'SpatRaster,data.frame'
bestMatch(x, y, labels=NULL, fun="squared", ..., filename="", overwrite=FALSE, wopt=list())

# S4 method for class 'SpatRaster,matrix'
bestMatch(x, y, labels=NULL, fun="squared", ..., filename="", overwrite=FALSE, wopt=list())

Arguments

x: SpatRaster
y: SpatVector, data.frame or matrix
labels: character. labels that correspond to each class (row in y
fun: character. One of "abs" for the mean absolute difference, or "squared" for the mean squared difference. Or a true function like terra:::match_sqr
...: additional arguments passed to fun. For the built-in functions this can be na.rm=TRUE
filename: character. Output filename
overwrite: logical. If TRUE, filename is overwritten
wopt: additional arguments for writing files as in writeRaster

Value

SpatRaster

Examples

f <- system.file("ex/logo.tif", package = "terra")
r <- rast(f)

# locations of interest 
pts <- vect(cbind(c(25.25, 34.324, 43.003), c(54.577, 46.489, 30.905)))
pts$code <- LETTERS[1:3]

plot(r)
points(pts, pch=20, cex=2, col="red")
text(pts, "code", pos=4, halo=TRUE)


x <- scale(r)

s1 <- bestMatch(x, pts, labels=pts$code)
plot(s1)


# same result
e <- extract(x, pts, ID=FALSE)
s2 <- bestMatch(x, e, labels=c("Ap", "Nt", "Ms"))