An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20110714/e1ae5db0/attachment.pl>
Grouping point clouds into polygons
3 messages · Patrick S, Els Ducheyne, Baptiste Auguie
Hi I think Thiessens polygons are maybe a solution for your probem HTH Els
On 14 Jul 2011, at 08:42, Patrick S wrote:
This is more of a theoretical question than an R-specific question, but does anyone have any suggestions for grouping clouds of points into polygons representing clusters? I have a spatial data set with points for each commercial land use in Atlanta along with the square footage. I would like to cluster these, using K-means probably, and then I would like to draw polygons around them so that all of space is filled. So I don't want alpha hulls, because I don't want gaps between polygons. Rather I want the space between the polygons to be assigned to the adjacent cluster with the strongest "pull" in terms of square footage. Any thoughts? Patrick Sewell [[alternative HTML version deleted]]
_______________________________________________ R-sig-Geo mailing list R-sig-Geo at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Hi, If I understand your question properly, I would suggest the following approach. - Obtain a Voronoi tesselation from the set of points - merge the resulting polygons according to your clustering method Best, baptiste PS: it seems a similar suggestion has just been sent..
On 14 July 2011 18:42, Patrick S <sewellpatrick at gmail.com> wrote:
This is more of a theoretical question than an R-specific question, but does anyone have any suggestions for grouping clouds of points into polygons representing clusters? I have a spatial data set with points for each commercial land use in Atlanta along with the square footage. I would like to cluster these, using K-means probably, and then I would like to draw polygons around them so that all of space is filled. So I don't want alpha hulls, because I don't want gaps between polygons. Rather I want the space between the polygons to be assigned to the adjacent cluster with the strongest "pull" in terms of square footage. Any thoughts? Patrick Sewell ? ? ? ?[[alternative HTML version deleted]]
_______________________________________________ R-sig-Geo mailing list R-sig-Geo at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-geo