I'm using qgis Brighton

I have a dataset of 1500 points. I want to create polygons, where each polygon contains n=3 points from the dataset.

The aim of this is to have clusters of 3 points, which can be used for analysis, qua I can't show information on point level.

I have tried using heatmaps, but I'm not sure if this is the right tool for this analysis.

Here's a picture of my dataset.

enter image description here

Now, I want to create clusters of 3 points like this:

enter image description here (Sorry for the amateur cutting)

I want to this, so I can make analysis of the clusters (with 3 points) e.g. energy demand, house characteristics, socioeconomics and so on.


1 Answer 1


If you just want to create a tesselated patch of polygons formed by 3 vertices using all your 1500 points, you can do so with a Delaunay triangulation. Go to Vector -> Geometry Tools -> Delaunay triangulation. Input will be your point layer, output will be your polygon layer.

If, however, you have a more specific need in the way these polygons should be created, you'll need to clarify the kind of analysis you'll be performing.

  • Thank you for the reply. I tried the Delaunay triangulation, but it's not the way I'm think it should be done. I will edit my question.
    – Kasper
    Feb 10, 2015 at 7:29
  • Do you wish to do this for every possible combination of 3 points, for a random collection of 3 points, or for specific groups of 3 points? In the image you posted, the points between the clusters will themselves form other clusters, or not? Feb 11, 2015 at 12:11
  • I want to create clusters of points that are near each other e.g. like the picture I posted, maybe at a specific distance, though I think this will be to difficult to perform. Note that the example I put up, is only a thought experiment.
    – Kasper
    Feb 11, 2015 at 12:19

Not the answer you're looking for? Browse other questions tagged or ask your own question.