Class HDBSCAN

constructor

• new HDBSCAN(
      minClusterSize?: number,
      minPts?: number,
      metric?:
          | "euclid"
          | "manhattan"
          | "chebyshev"
          | ((arg0: number[], arg1: number[]) => number),
  ): HDBSCAN

  Parameters

  • OptionalminClusterSize: number

    Minimum number of clusters to be recognized as a cluster
  • OptionalminPts: number

    Number of neighborhood with core distance
  • Optionalmetric:
        | "euclid"
        | "manhattan"
        | "chebyshev"
        | ((arg0: number[], arg1: number[]) => number)

    Metric name

  Returns HDBSCAN

_d

_lastResult

_metric

_minClusterSize

_minPts

size

• get size(): number

  Number of clusters of last predicted

  Returns number

predict

• predict(datas: number[][]): number[]

  Returns predicted categories.

  Parameters

  • datas: number[][]

    Training data

  Returns number[]

  Predicted values