spatSample with option "stratified" returns error for rasters containing categories with a single pixel

[karlruf]

Hello Robert and all,

thanks for the awesome package, which I use almost daily!

I noticed that the spatSample function fails to sample categorical rasters which contain categories that are only represented by a single pixel.

This is the error message:
Error in sample.int(nrow(vv), min(size, nrow(vv)), prob = vv[, 3], replace = FALSE) :
length(n) == 1L is not TRUE

I believe the error is due to nrow(vv) returning NULL for a single row length sample.

Example:

rr <- rast(ncol=10, nrow=10, xmin=0, names="stratum")
values(rr) <- c(round(runif(ncell(rr)-1, 1, 3)),10)
freq(rr)
ss <- spatSample(rr,1, "stratified", xy=TRUE,  as.points=TRUE)

System information:
Operating System Windows 11
packageVersion("terra") = ‘1.7.71’
terra::gdal(lib="all") = gdal proj geos
"3.8.2" "9.3.1" "3.11.2"

Best

Karl

[rhijmans]

Thank you very much for reporting this and sorry for the very long wait. I think this has been fixed now.

library(terra)
rr <- rast(ncol=10, nrow=10, xmin=0, names="stratum")
set.seed(1)
values(rr) <- c(round(runif(ncell(rr)-1, 1, 3)),10)
spatSample(rr,1, "stratified", xy=TRUE,  as.points=TRUE)
# class       : SpatVector 
# geometry    : points 
# dimensions  : 4, 3  (geometries, attributes)
# extent      : 9, 171, -81, -27  (xmin, xmax, ymin, ymax)
# coord. ref. : lon/lat WGS 84 (CRS84) (OGC:CRS84) 
# names       :     x     y stratum
# type        : <num> <num>   <num>
# values      :     9    63       1
#                   9   -45       2
#                 135   -27       3