hex.genmodel

Interface IClusteringModel

All Known Implementing Classes:

KMeansMojoModel

public interface IClusteringModel

Clustering Model Interface

Method Summary

Methods

Modifier and Type	Method and Description
int	distances(double[] row, double[] distances) Calculates squared distances to all cluster centers.
int	getNumClusters() Returns number of cluster used by this model.
double[]	score0(double[] row, double[] preds)

Method Detail

score0

double[] score0(double[] row,
              double[] preds)

distances

int distances(double[] row,
            double[] distances)

Calculates squared distances to all cluster centers. hex.genmodel.GenModel.KMeans_distances(..) for precise definition of the distance metric. Pass in data in a double[], in a same way as to the score0 function. Cluster distances will be stored into the distances[] array. Function will return the closest cluster. This way the caller can avoid to call score0(..) to retrieve the cluster where the data point belongs. Warning: This function can modify content of row array (same as for score0).

Parameters:

row - input row

distances - vector of distances

Returns:

index of closest cluster

getNumClusters

int getNumClusters()

Returns number of cluster used by this model.

Returns:

number of clusters