[R-sig-Geo] identifying clusters between sets of points

Sam Veloz sdveloz at ucdavis.edu
Wed Apr 15 23:43:22 CEST 2009


I had thought Ripley's K was only for univariate situations but I just 
found the k12hat function in the splancs package for bivariate 
situations. I will give that a try.
Thanks,
Sam


José Manuel Blanco Moreno wrote:
> I'm not sure that I have correctly understood your problem, but it 
> sounds like the sort of problems that are usually solved through 
> Ripley's K (or some of its derivatives: Condit's Omega, Pair Correlation 
> Function, Circumcircle methods...)
> Good luck.
> José M. Blanco
> 
> Sam Veloz escribió:
>> I have 3 sets of points (A,B,C) with coordinates. I would like to test 
>> whether points from group A are more spatially clustered around points 
>> in group B or are points from group A more spatially clustered around 
>> points in group C. Can anyone suggest a test(or r function) for doing 
>> this?
>> Thanks,
>> Sam
> 
> 

-- 
****************************************************
Sam Veloz
Postdoctoral Researcher
Department of Environmental Science and Policy
University of California, Davis
sdveloz at ucdavis.edu



More information about the R-sig-Geo mailing list