Question

Cluster points based on location (K-MEANS method)

  • 2 March 2016
  • 5 replies
  • 28 views
F
Badge

Hi,

I am trying to do the following:

I have a total number of n adresses (points) spatially divided over a given area.

I want to group them into x clusters using y members per cluster. This has to be as spatially optimal as possible.

As far as I know, it should be something in the lines of a k-means algorithm.

Actually, the Neighbourhood Aggregator comes close, but I want to set a fixed number of members per cluster, not a minimal.

Anyone got any ideas? I already tried working with the cluster modeller, to no avail.

Thanks in advance!

Flip

5 replies

erik_jan
Rank
Userlevel 2
Badge +12

Hi Flip,

Have you tried creating a pointcloud from the points.

Then use the PointCloudThinner to generate the points in the center of your clusters (thin by Y).

Then use the NeighborFinder to find all points, not in the pointcloud nearest to the selected points.

Erik Jan

mark2atsafe
Rank
Userlevel 4
Badge +25

Or... as @erik_jan says, convert to a point cloud then use the PointCloudDensityTiler. That will (I think) create features with an equal number of points, which you could then turn back into plain points.

F
Badge

Thanks Erik Jan & Mark, I'll give that a try!

B

Hi @flipvandervalk, did the above solution work for you?

helmoet
Badge +2

Hi @flipvandervalk @erik_jan @Mark2AtSafe @bieahart, I posted an idea and a custom transformer to FME Hub. Might be interesting for you? See https://knowledge.safe.com/idea/59941/k-means-point-clustering-using-fme.html?

 

Reply


Community Stats

28,634
Posts
109,179
Replies
36,513
Members
Terms & ConditionsCookie settings
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.