[R-sig-Geo] clustering spatial point data

[Luca Morandini]

On 06/03/2011 06:33 AM, Rolf Turner wrote:
>
> I haven't yet seen any replies to your post, so I'll chip in with my
> two cents (or less!) worth.

I'll put down my 2 €cent as well ;)


> One very naive approach would be just to use the Euclidean distance
> between the triples (x_i, y_i, z_i) where x_i and y_i specify the point
> locations and z_i is the numeric mark of the point in question.
>
> Using Euclidean distance is almost surely *not* the right thing to do.

What about the local version of Moran'I auto-correlation index ?

This indicator should be suitable as a metric for detecting the wanted clusters, 
since it takes into account both response variable and spatial distance.

Regards,

Luca Morandini
http://www.lucamorandini.it

-- 
Regards,

Luca Morandini
http://www.lucamorandini.it