rasterSOM: Self Organizing Maps for rasters
In MariekeDirk/GeoInterpolation: Temperature Interpolation methods

Description Usage Arguments Author(s) Examples

Self organizing maps for raster datasets. The function uses the som function from the kohonen package to calculate the SOMs. The map nodes are typically 3x4 or 6x4. The update radius is based on the input data. The learning rate should decrease with the number of iterations. Richardson(2003) used the following settings:

1000 iterations with learning rate of 0.2
10.000 iterations with a learning rate of 0.1
50.000 iterations with a learning rate of 0.002

1 2	rasterSOM(x, xdim = 3, ydim = 5, rlen = 1000, alpha = 0.2, datum = NULL)

`x`	Raster* format
`xdim`	x-dimension of the map nodes
`ydim`	y-dimension of the map nodes.
`rlen`	number of iterations to fit the SOMs
`alpha`	learning rate
`datum`	optional. Used to calculate the relative frequency of each SOM for each month.