More efficient raster reclassification

Question

I have a raster where the first column must be reclassified, leaving all other values as NA. I have written the code to do this (below) but it is very inefficient as the raster size grows much beyond this toy example.

Could someone recommend a more efficient way to re-write this code so that it can handle much larger raster files? I suspect apply() could be useful here but I'm unsure what direction to take this in.

r<-raster::raster(nrow=10, ncol=10)
r[]<-NA

  for(i in 1:r@nrows){
    r[(r@nrows*i)-(r@ncols-1)]<-2

  }

Answer 1

First a better way to do what you do

library(raster)
r <- raster::raster(nrow=10, ncol=10)
for(i in 1:nrow(r)){
    r[i, 1] <- 2
}

A much better alternative

r <- raster::raster(nrow=10, ncol=10)
r[,1] <- 2

But for larger rasters it is perhaps best do

r <- raster::raster(nrow=10, ncol=10)
xy <- cbind(xFromCol(r, 1), yFromRow(r, 1:nrow(r)))
r <- rasterize(xy, r, field=2)

And probably less efficient like this

r <- raster::raster(nrow=10, ncol=10)
r <- init(r, "col")
r <- reclassify(r, rbind(c(1,1,2), c(2,Inf,NA)), right=TRUE, include.lowest=TRUE)

Same idea, nicely concise

r <- subs(init(r, "col"), data.frame(from=1, to=2))