Spatial clustering with spdep::skater

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Spatial clustering with spdep::skater

Axel Urbiz
Dear Experts,

My apologies in advance as this is more a statistical question, rather than an R programming question, but hope you can point me in the right direction.

I’m working with spdep::skater to fit clusters to spatial data subject to contiguity constraints. This function fits clusters by edge removal from Minimum Spanning Trees.

In this context, I’d appreciate any pointers on how to tune the number of clusters. What is a sensible criteria to use?

Thanks,
Axel.
_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Spatial clustering with spdep::skater

Elias T Krainski
Hi Axel,

The original paper suggests to look at the SSW decay. You can visualize
it easily as it is returned in the output from skater().

Best,

Elias


On 08/07/18 08:52, Axel Urbiz wrote:

> Dear Experts,
>
> My apologies in advance as this is more a statistical question, rather than an R programming question, but hope you can point me in the right direction.
>
> I’m working with spdep::skater to fit clusters to spatial data subject to contiguity constraints. This function fits clusters by edge removal from Minimum Spanning Trees.
>
> In this context, I’d appreciate any pointers on how to tune the number of clusters. What is a sensible criteria to use?
>
> Thanks,
> Axel.
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo