com.apporiented.algorithm.clustering.DefaultClusteringAlgorithm Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of hierarchical-clustering Show documentation
Agglomerative hierarchical clustering analysis and visualization implemented in Java
There is a newer version: 1.2.0
Show newest version
/*******************************************************************************
 * Copyright 2013 Lars Behnke
 * 
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 * 
 *   http://www.apache.org/licenses/LICENSE-2.0
 * 
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 ******************************************************************************/

package com.apporiented.algorithm.clustering;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashSet;
import java.util.List;

public class DefaultClusteringAlgorithm implements ClusteringAlgorithm {

	@Override
	public Cluster performClustering(double[][] distances,
	        String[] clusterNames, LinkageStrategy linkageStrategy) {

    checkArguments(distances, clusterNames, linkageStrategy);
    /* Setup model */
    List clusters = createClusters(clusterNames);
    DistanceMap linkages = createLinkages(distances, clusters);

    /* Process */
    HierarchyBuilder builder = new HierarchyBuilder(clusters, linkages);
    while (!builder.isTreeComplete()) {
      builder.agglomerate(linkageStrategy);
    }

    return builder.getRootCluster();
  }

  private void checkArguments(double[][] distances, String[] clusterNames,
      LinkageStrategy linkageStrategy) {
    if (distances == null || distances.length == 0
		        || distances[0].length != distances.length) {
			throw new IllegalArgumentException("Invalid distance matrix");
		}
		if (distances.length != clusterNames.length) {
			throw new IllegalArgumentException("Invalid cluster name array");
		}
		if (linkageStrategy == null) {
			throw new IllegalArgumentException("Undefined linkage strategy");
		}
		int uniqueCount=new HashSet(Arrays.asList(clusterNames)).size();
		if (uniqueCount != clusterNames.length) {
			throw new IllegalArgumentException("Duplicate names");
		}
  }

  @Override
  public Cluster performWeightedClustering(double[][] distances, String[] clusterNames,
      double[] weights, LinkageStrategy linkageStrategy) {

    checkArguments(distances, clusterNames, linkageStrategy);

    if (weights.length != clusterNames.length) {
      throw new IllegalArgumentException("Invalid weights array");
    }

    /* Setup model */
    List clusters = createClusters(clusterNames, weights);
    DistanceMap linkages = createLinkages(distances, clusters);

    /* Process */
    HierarchyBuilder builder = new HierarchyBuilder(clusters, linkages);
    while (!builder.isTreeComplete()) {
      builder.agglomerate(linkageStrategy);
    }

    return builder.getRootCluster();
  }

  private DistanceMap createLinkages(double[][] distances,
	        List clusters) {
        DistanceMap linkages = new DistanceMap();
		for (int col = 0; col < clusters.size(); col++) {
			for (int row = col + 1; row < clusters.size(); row++) {
				ClusterPair link = new ClusterPair();
				Cluster lCluster = clusters.get(col);
				Cluster rCluster = clusters.get(row);
				link.setLinkageDistance(distances[col][row]);
				link.setlCluster(lCluster);
				link.setrCluster(rCluster);
				linkages.add(link);
			}
		}
		return linkages;
	}

	private List createClusters(String[] clusterNames) {
		List clusters = new ArrayList();
        for (String clusterName : clusterNames) {
            Cluster cluster = new Cluster(clusterName);
            clusters.add(cluster);
        }
		return clusters;
	}

  private List createClusters(String[] clusterNames, double[] weights) {
    List clusters = new ArrayList();
    for (int i = 0; i < weights.length; i++) {
      Cluster cluster = new Cluster(clusterNames[i]);
      cluster.setDistance(new Distance(0.0, weights[i]));
      clusters.add(cluster);
    }
    return clusters;
  }

}