Member Junction
    Preparing search index...

    Interface ClusterConfig

    Full configuration for a clustering run.

    Pass an instance to ClusteringService.RunPipeline() or emit from the config panel's RunClustering output.

    interface ClusterConfig {
        Algorithm: ClusterAlgorithm;
        ColorBy?: "cluster" | "entity";
        Dimensions?: 2 | 3;
        DistanceMetric: ClusterDistanceMetric;
        EntityDocumentID: string;
        EntityDocumentIDs?: string[];
        EntityName: string;
        Epsilon: number;
        Filter: string;
        K: number;
        MaxRecords: number;
        MinPoints: number;
    }
    Index

    Properties

    Algorithm ColorBy? Dimensions? DistanceMetric EntityDocumentID EntityDocumentIDs? EntityName Epsilon Filter K MaxRecords MinPoints

    Properties

    Algorithm: ClusterAlgorithm

    Clustering algorithm to use.

    ColorBy?: "cluster" | "entity"

    Legend mode: color points by their assigned cluster (default) or by source entity.

    Dimensions?: 2 | 3

    Projected layout dimensionality: 2 (default, SVG) or 3 (rotatable projection).

    DistanceMetric: ClusterDistanceMetric

    Distance metric used when comparing vectors.

    EntityDocumentID: string

    Specific Entity Document ID to use. When blank, the first active doc for the entity is used.

    EntityDocumentIDs?: string[]

    Multi-entity clustering: one or more Entity Document IDs whose vectors are merged into a single analysis. When two or more are selected this takes precedence over EntityDocumentID/EntityName. All selected documents must use the same embedding model — the server hard-blocks mismatches.

    EntityName: string

    Entity name to cluster (must have Entity Documents with vectors).

    Epsilon: number

    Epsilon neighbourhood radius for DBSCAN (ignored for K-Means).

    Filter: string

    Optional SQL-style filter applied to the Entity Document Runs query.

    K: number

    Number of clusters for K-Means (ignored for DBSCAN).

    MaxRecords: number

    Maximum number of entity records to fetch for clustering.

    MinPoints: number

    Minimum points to form a dense region in DBSCAN (ignored for K-Means).

    Settings

    Member Visibility

    On This Page

    Properties