gdal_sd: Cell-wise raster standard deviation
In johnbaums/things: Mostly spatial things

Description Usage Arguments Details Value Examples

Calculate a raster layer whose cells' values are the standard deviations across the bands of a single raster, or across multiple rasters.

1	gdal_sd(infile, outfile, return_raster = TRUE, quiet = TRUE)

`infile`	A `RasterStack` or `RasterBrick`, a file path to a single, multiband raster file, or a vector of paths to multiple raster files. If a vector of paths is provided, then if any of the rasters have multiple bands, the first band will be used. A maximum of 26 bands or files can be provided to `infile`.
`outfile`	A path to the desired output raster file, which will be created if it doesn't already exist (though the containing directory must exist). If not provided and `return_raster` is `TRUE`, a temporary output file will be created.
`return_raster`	Logical. Should the output raster be read back in to R as a `Raster` object?
`quiet`	Logical. Should GDAL output messages be suppressed?

The quantity calculated is the corrected sample standard deviation, calculated as sqrt(sum((x - mean(x))^2)/(length(x)-1)).

Unless return_raster is TRUE, gdal_sd creates the file specified at outfile but returns NULL, invisibly, to R. If return_raster is TRUE, the standard deviation Raster object is returned to R.

library(raster)
s <- stack(replicate(5, raster(matrix(rnorm(100), ncol=10))))

# Input is a RasterStack
gdal_sd(s)

# Input is multiple raster files
writeRaster(s, ff <- paste0(tempfile(), 1:5, '.tif'), bylayer=TRUE)
gdal_sd(ff)

# Input is a raster file with multiple bands 
writeRaster(s, f <- tempfile(fileext='.tif'))
gdal_sd(f)

# Input is a RasterBrick with multiple bands
b <- brick(f)
gdal_sd(b)