Raster multidimensional scaling (MDS)

Multidimensional scaling of raster values within an N x N focal window

An MDS focal function. If only one value provided for s, then a square matrix (window) will be used. If window.median = FALSE then the center value of the matrix is returned and not the median of the matrix

raster.mds(r, s = 5, window.median = FALSE, ...)

Arguments

r	Raster layer
s	Window size (may be a vector of 1 or 2) of n x n dimension.
window.median	(TRUE/FALSE) Return the median of the MDS matrix values.
...	Additional arguments passed to raster::focal

Value

A raster class object or raster written to disk

References

Quinn, G.P., & M.J. Keough (2002) Experimental design and data analysis for biologists. Cambridge University Press. Ch. 18. Multidimensional scaling and cluster analysis.

Author

Jeffrey S. Evans <jeffrey_evans@tnc.org>

Examples

# \donttest{ 
 library(raster)
 r <- raster(system.file("external/rlogo.grd", package="raster")) 
   r <- r / cellStats(r, "max")
 
 diss <- raster.mds(r)
 diss.med <- raster.mds(r, window.median = TRUE)

opar <- par(no.readonly=TRUE)
   par(mfrow=c(2,2))
   plot(r)
     title("R logo band-1")
   plot( focal(r, w = matrix(1, nrow=5, ncol=5), fun = var) )
     title("Variance")     
     plot(diss)
       title("MDS")
     plot(diss.med)
       title("Median MDS")
par(opar)
# }