DBSCANConfig
VariantDBSCAN density-based clustering. See setup_DBSCAN.
Properties
epsnumberdefault0.5> 0Radius of neighborhood.
min_pointsintegerdefault5≥ 1Minimum number of points in a neighborhood to form a cluster.
border_pointsbooleandefaulttrueAssign border points to clusters.
searchstringdefault"kdtree"Nearest neighbor search strategy.
one of
"kdtree""linear""dist"bucket_sizeintegerdefault100≥ 1Size of buckets for the k-d tree search.
split_rulestringdefault"SUGGEST"Rule for splitting the k-d tree.
one of
"SUGGEST""STD""MIDPT""FAIR""SL_MIDPT""SL_FAIR"approxbooleandefaultfalseUse approximate nearest neighbor search.
weightsnumber[] | null≥ 1 itemsOptional weights for data points. null = unweighted.
Relationships
Used by