org.openimaj.ml.clustering.kmeans.SphericalKMeansResult Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of clustering Show documentation
Various clustering algorithm implementations for all primitive types including random, random forest, K-Means (Exact, Hierarchical and Approximate), ...
There is a newer version: 1.3.10
Show newest version
/**
 * Copyright (c) 2011, The University of Southampton and the individual contributors.
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 *   * 	Redistributions of source code must retain the above copyright notice,
 * 	this list of conditions and the following disclaimer.
 *
 *   *	Redistributions in binary form must reproduce the above copyright notice,
 * 	this list of conditions and the following disclaimer in the documentation
 * 	and/or other materials provided with the distribution.
 *
 *   *	Neither the name of the University of Southampton nor the names of its
 * 	contributors may be used to endorse or promote products derived from this
 * 	software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */
package org.openimaj.ml.clustering.kmeans;

import gnu.trove.list.array.TIntArrayList;
import gnu.trove.procedure.TIntProcedure;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;
import java.io.PrintWriter;
import java.util.Arrays;
import java.util.Scanner;

import org.openimaj.feature.DoubleFVComparison;
import org.openimaj.ml.clustering.CentroidsProvider;
import org.openimaj.ml.clustering.Clusters;
import org.openimaj.ml.clustering.SpatialClusterer;
import org.openimaj.ml.clustering.SpatialClusters;
import org.openimaj.ml.clustering.assignment.HardAssigner;
import org.openimaj.ml.clustering.assignment.hard.ExactDoubleAssigner;
import org.openimaj.util.pair.IntDoublePair;

/**
 * The result of a {@link SpatialClusterer} that just produces a flat set of
 * centroids.
 *
 * @author Jonathon Hare ([email protected])
 */
public class SphericalKMeansResult implements SpatialClusters, CentroidsProvider {
	final static String HEADER = Clusters.CLUSTER_HEADER + "SpKM";

	/** The centroids of the clusters */
	public double[][] centroids;

	/** The assignments of the training data to clusters */
	public int[] assignments;

	@Override
	public boolean equals(Object obj) {
		if (!(obj instanceof SphericalKMeansResult))
			return false;

		final SphericalKMeansResult other = (SphericalKMeansResult) obj;
		for (int i = 0; i < this.centroids.length; i++) {
			if (!Arrays.equals(this.centroids[i], other.centroids[i]))
				return false;
		}
		return true;
	}

	@Override
	public String asciiHeader() {
		return "ASCII" + HEADER;
	}

	@Override
	public byte[] binaryHeader() {
		return HEADER.getBytes();
	}

	@Override
	public void readASCII(Scanner br) throws IOException {
		// Read Header
		final int K = Integer.parseInt(br.nextLine().trim());
		final int M = Integer.parseInt(br.nextLine().trim());

		centroids = new double[K][M];
		for (int k = 0; k < K; k++) {
			final String[] parts = br.nextLine().split(",");

			for (int d = 0; d < M; d++) {
				centroids[k][d] = Double.parseDouble(parts[d]);
			}
		}

		final int A = Integer.parseInt(br.nextLine().trim());
		assignments = new int[A];
		for (int a = 0; a < A; a++) {
			assignments[a] = Integer.parseInt(br.nextLine().trim());
		}
	}

	@Override
	public void readBinary(DataInput in) throws IOException {
		final int K = in.readInt();
		final int M = in.readInt();

		centroids = new double[K][M];

		for (int k = 0; k < K; k++) {
			for (int d = 0; d < M; d++) {
				centroids[k][d] = in.readDouble();
			}
		}

		final int A = in.readInt();
		assignments = new int[A];
		for (int a = 0; a < A; a++) {
			assignments[a] = in.readInt();
		}
	}

	@Override
	public void writeASCII(PrintWriter writer) throws IOException {
		writer.println(centroids.length);
		writer.println(centroids[0].length);

		for (int k = 0; k < centroids.length; k++) {
			for (int d = 0; d < centroids[0].length; d++) {
				writer.print(centroids[k][d] + ",");
			}
			writer.println();
		}

		writer.println(assignments.length);
		for (int a = 0; a < assignments.length; a++) {
			writer.println(assignments[a]);
		}
	}

	@Override
	public void writeBinary(DataOutput out) throws IOException {
		out.writeInt(centroids.length);
		out.writeInt(centroids[0].length);

		for (int k = 0; k < centroids.length; k++) {
			for (int d = 0; d < centroids[0].length; d++) {
				out.writeDouble(centroids[k][d]);
			}
		}

		out.writeInt(assignments.length);
		for (int a = 0; a < assignments.length; a++) {
			out.writeInt(assignments[a]);
		}
	}

	@Override
	public String toString() {
		String str = "";
		str += "DoubleCentroidsResult" + "\n";
		str += "No. of Clusters: " + centroids.length + "\n";
		str += "No. of Dimensions: " + centroids[0].length + "\n";
		return str;
	}

	@Override
	public double[][] getCentroids() {
		return this.centroids;
	}

	@Override
	public HardAssigner defaultHardAssigner() {
		return new ExactDoubleAssigner(this, DoubleFVComparison.INNER_PRODUCT);
	}

	@Override
	public int numDimensions() {
		return centroids[0].length;
	}

	@Override
	public int numClusters() {
		return centroids.length;
	}

	/**
	 * Compute the histogram of number of assignments to each cluster
	 *
	 * @return the histogram
	 */
	public int[] getAssignmentHistogram() {
		final int[] hist = new int[centroids.length];

		for (int i = 0; i < assignments.length; i++) {
			hist[assignments[i]]++;
		}

		return hist;
	}

	/**
	 * Filter the cluster centroids be removing those with less than threshold
	 * items
	 *
	 * @param threshold
	 *            minimum number of items
	 * @return the filtered clusters
	 */
	public double[][] filter(int threshold) {
		final int[] hist = getAssignmentHistogram();
		final TIntArrayList toKeep = new TIntArrayList();
		for (int i = 0; i < hist.length; i++) {
			if (hist[i] > threshold) {
				toKeep.add(i);
			}
		}

		final double[][] fcen = new double[toKeep.size()][];
		toKeep.forEach(new TIntProcedure() {
			int i = 0;

			@Override
			public boolean execute(int value) {
				fcen[i++] = centroids[value];
				return true;
			}
		});

		return fcen;
	}
}