How to extract xy-coordinates from raster where its highest value is located within a polygon?

Question

given is a raster as well as a SpatialPolygonsDataframe. In order to retrieve the highest value of the raster within the area of a polygon, raster::extract can be used. It works fine.

How to get additionally the coordinates of the extracted highest value of the raster within the area of a polygon?

# create raster
r <- raster(ncol=36, nrow=18)
r[] <- runif(ncell(r))
# create SpatialPolygons from GridTopology
grd <- GridTopology(c(-150, -50), c(40, 40), c(8, 3))
Spol <- as(grd, "SpatialPolygons")
# create SpatialPolygonsDataFrame
centroids <- coordinates(Spol)
x <- centroids[,1]
y <- centroids[,2]
SPDF <- SpatialPolygonsDataFrame(Spol, data=data.frame(x=x, y=y, row.names=row.names(Spol)))
# extract max value of raster for each SpatialPolygon
ext <- raster::extract(r, SPDF, fun=max)

*example code is taken from R-documentation

Answer 1

You can use the cellnumbers=TRUE argument in extract, followed by a sapply to get the cell number:

ext <- raster::extract(r, SPDF, cellnumbers=TRUE)
v <- t(sapply(ext, function(i) i[which.max(i[,2]), ] ))

#      cell     value
# [1,]  185 0.9303460
# [2,]  188 0.9821190
# [3,]  154 0.9926290
# [4,]  232 0.8907819
# [5,]  234 0.9998510

To get the coordinates:

xyFromCell(r, v[,1])

#         x   y
# [1,] -135  35
# [2,] -105  35
# [3,]  -85  45
# [4,]  -25  25
# [5,]   -5  25