[GRASS-user] algorithm used in i.cluster

[Markus Neteler]

On Wed, Apr 21, 2010 at 5:29 PM, Georg Kaspar <georg at muenster.de> wrote:
> hi,
>
> a quote from Markus' "Introduction to GRASS GIS Software" (1998):
>
> "[...] GRASS uses the "minimum distance to means" algorithm for deriving the
> clusters [...]

This knowledge is essentially based on
http://grass.osgeo.org/grass64/manuals/html64_user/i.cluster.html

and "The GRASS 4 Image Processing manual"
http://grass.itc.it/gdp/imagery/grass4_image_processing.pdf

and
http://trac.osgeo.org/grass/browser/grass/branches/releasebranch_6_4/imagery/i.cluster
http://download.osgeo.org/grass/grass6_progman/c__exec_8c_source.html
-> /* generate class means */
    00063     I_cluster_means(C);

etc.

> "minimum distance to means" is described in Lillesand and Kiefer (2000), as
> a supervised classification technique which requires a priori information.
> Does the i.cluster module use a modified version or have there been changes
> since 1998?

To my knowledge, no algorithm changes have been taken place.

> the manual page does not contain any further information on how the module
> works...

Luckily it is open source :)

> does anybody know which exact algorithm is used (k-means probably...)?

I'm afraid that you need to compare it to a k-means algorithm. Please let
us know your findings.

Markus