The DBSCAN filter performs Density-Based Spatial Clustering of Applications with Noise (DBSCAN) [Ester1996] and labels each point with its associated cluster ID. Points that do not belong to a cluster are given a Cluster ID of -1. The remaining clusters are labeled as integers starting from 0.

Default Embedded Stage

This stage is enabled by default

New in version 2.1.





The minimum cluster size min_points should be greater than or equal to the number of dimensions (e.g., X, Y, and Z) plus one. As a rule of thumb, two times the number of dimensions is often used. [Default: 6]


The epsilon parameter can be estimated from a k-distance graph (for k = min_points minus one). eps defines the Euclidean distance that will be used when searching for neighbors. [Default: 1.0]


Comma-separated string indicating dimensions to use for clustering. [Default: X,Y,Z]