QGIS plugin for clustering vector data based on attributes.
- Customizable clustering with 4 methods: two variations of K-Means and two variations of Hierarchical clustering. 3 of them are for known number of clusters, and 1 for unknown
- All clustering methods are available as QGIS Processing tools and via special GUI
- Inpit is any vector layer with one or more numeric attributes
- Output is a new attribute with relevant for object cluster number.
- Different weights for attributes available when using native hierarchical clustring method
- Ability to calculate distances from each object to centroids of each cluster and save them as attributes
- Ability to draw Elbow plot for estimating optimal number of clusters
Elbow plot:
Output attributes (cluster number and distances to cluster centroids):
General plugin documentation is available here: https://ekazakov.me/attribute-based-clustering-qgis-plugin
The plugin is available in the official QGIS Plugin Repository.
You can also install it from ZIP archive.
Eduard Kazakov | ee.kazakov@gmail.com