Porting clustering and summarize about qgis-edge-bundling HOT 2 OPEN

Hi Lucas,

Thank you for the kind words!

The documentation of the built-in K-Means clustering tool in QGIS 3.4 states

If input geometries are lines or polygons, the clustering is based on the centroid of the feature.

The clustering we implemented in https://github.com/dts-ait/qgis-edge-bundling/blob/master/cluster_lines_kmeans.py uses start and end points instead and lines are clustered irrespective of direction.

I'm not aware of any current plans to port the old script but if you're interested in looking into it, that would be a very welcome contribution!

from qgis-edge-bundling.

Hi Anita,

Thank you for your quick reply.

Ah yes, I see.. clustering on centroid seems to be quite suboptimal for my use-case. Perhaps I can do k-means clustering twice manually on (1) start-points and (2) end-points, and build new clusters from the permutations. That way the clusters would even take direction into consideration 🤔

I have to see if I can find the time to spend on porting, but I must say I am not familiar with the PyQGIS API.

Thanks again!

from qgis-edge-bundling.