CODEWITHSUNDEEP
rr-raster
dww
dww

Reputation: 31452

subassignment of selected layers in raster brick

I want to modify a subset of layers in a raster brick by multiplying those layers by another raster.

For example, if we have a raster brick called 'r.brick' and we try to multiply its layers 2:4 by a raster, 'r.mult', with same row&column dimensions:

but, if I try to assign the result back into the subset layers I get an error

r.brick[[2:4]] = r.brick[[2:4]] * r.mult
# Error in value[] <- val : 
#   incompatible types (from S4 to double) in subassignment type fix

The error message suggests that the assignment is trying to assign the raster values rather than the raster itself. But if I try the assignment with getValues, I get a different error: 

r.brick[[2:4]] = getValues(r.brick[[2:4]] * r.mult)
# Error in .local(x, values, ...) : length(values) is not equal to ncell(x)

What is the correct way to do this?

Some reproducible data:

library(raster)
r.list = vector("list", 20)
set.seed(123)
for (i in 1:20) {
  r.list[[i]] = raster(vals=runif(100), nrows=10, ncols=10, ext=extent(c(0,25,0,25)))
}
r.brick = brick(r.list)
r.mult = raster(vals=sample(2,100,T), nrows=10, ncols=10, ext=extent(c(0,25,0,25)))

Upvotes: 1

Views: 1580

Answers (3)

Robert Hijmans
Robert Hijmans

Reputation: 47146

I think that is a missing feature. Thanks for pointing that out. I would go with the loop following maRtin (perhaps after creating a RasterStack, that may be more efficient). If the dataset is not too large, you can do

# example data
library(raster)
b <- brick(nrows=2, ncols=2, nl=6)
values(b) <- rep(1:4, 6) 
r.mult <- raster(vals=10, nrows=2, ncols=2)


values(b)[,3:4] <- values(b[[3:4]] * r.mult)
# values(b)

Upvotes: 2

maRtin
maRtin

Reputation: 6516

Then I only got a workaround (using a loop) for you:

layers <- 2:4
for(i in layers) {
  r.brick[[i]]  <- r.brick[[i]] * r.mult
}

Note: Apparently the assignment with the [] subsetting only works with single layers.

Upvotes: 3

danielson
danielson

Reputation: 1029

The following code with multiple layers 2-4 of r.brick raster values with r.mult values and assign the results to r.brick layers 2-4. 

attr(attr(r.brick, 'data'), 'values')[,2:4] = attr(attr(r.brick, 'data'), 'values')[,2:4] * attr(attr(r.mult, 'data'), 'values')

or 

attr(attr(r.brick, 'data'), 'values')[,2:4] = getValues(r.brick[[2:4]] * r.mult)

Upvotes: 1

Related Questions