Replacing raster values with closest value from list using R?

Question

The task: I'm creating a function to assess flood damage in R. The inputs are:

• Dataframe with water depth values and corresponding damage values in percentage.
• Raster representing water depth

The raster values must be approximated to the closest value in the first column of the df and then create a new raster with damage values using corresponding values from the second column).

library(raster)
damage <- cbind(c(0.0,0.2,0.4,0.6,0.8,1.0),c(0, 5, 10, 40, 70, 100))
r.wd <- raster(myraster.tif)

I've tried using which.min(abs(damage[,1]-r.wd)) but doesn't work.

Answer 1

Here's an option using calc.

library(raster)

# Create test data
r <- raster()
r[] <- rnorm(n = ncell(r), mean = 0.5, sd = 0.1)

damageDf <- data.frame(depth = c(0.0,0.2,0.4,0.6,0.8,1.0),
                       damage = c(0, 5, 10, 40, 70, 100))

# Function to be passed to calc (gets damageDf from global environment)
depth2damage <- function(x) {
    damageDf$damage[which.min(abs(damageDf$depth - x))]
}

# Apply function to depth raster
# You may omit forceapply = TRUE but since we know that the function
# only works for single values we might as well explicitly tell calc about it
damageRaster <- calc(r, depth2damage, forceapply = TRUE)