views:

1106

answers:

4

Hi Everyone,

I am simply looking for a good tutorial that will walk me through how to create a SOM in R. I am reading Kohonen and Kaski's paper on using the maps to identify the structure of Welfare, and want to try the technique my self.

I think many of the examples in R have a lot to be desired. I have looked on the web and didn't find anything that I thought made the process easy to understand.

Any help you can provide will be greatly appreciated.

A: 

Did you look at the CRAN package SOM ?

Dirk Eddelbuettel
I did, but looking at the documentation with a few lines of code and no context is a little tough.
Btibert3
+3  A: 

A few pointers:

Most of the SOM related packages are from the Chemometrics and Computational Physics area, but you also have a look at the cluster view on CRAN.

Edit: Incidentally, I'm not aware of any tutorials, but one good way to find examples is to google with ext:r. There are several "som" scripts that show up that way, including this example which shows how to use it with GGobi.

Shane
rcs
+3  A: 

Shane already mentioned the kohonen package. An introduction and tutorial can be found in the Journal of Statistical Software: Self- and Super-organizing Maps in R: The kohonen Package

rcs
I saw this too, and it is the closest that I have found to being a good tutorial. Maybe it is wishful thinking, but I am hoping I can find a tutorial that is geared towards someone being introduced to the technique as well as how to perform the task in R, step-by-step.
Btibert3
A: 

GeoSpatial Data Mining course from edugi has pretty good materials.

radek