Version 1.20
SELECT * FROM KMeans ( ON { table | view | (query) } AS InputTable [ ON { table | view | (query) } AS CentroidsTable ] OUT TABLE OutputTable (output_table) [ OUT TABLE ClusterAssignmentTable (cluster_assignement_table) ] USING [ IDColumn ('id_column') ] [ TargetColumns ({ 'target_column' | target_column_range [,...]) ] [ UnpackColumns ({'true'|'t'|'yes'|'y'|'1'|'false'|'f'|'no'|'n'|'0'}) ] [ InitialSeeds ('starting_cluster' [,...]) ] [ NumClusters (number_of_means) [ Seed (seed) ] ] [ StopThreshold (threshold) ] [ MaxIterNum (max_iterations) ] ) AS alias;
You must specify only one of NumClusters, InitialSeeds, or CentroidsTable. If you specify more than one, the function gives top priority to InitialSeeds, then to NumClusters, and then to CentroidsTable.