Apache
/
commons-math
mirror of https://github.com/apache/commons-math.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Remove deprecated stat.clustering package.

This commit is contained in:
tn 2015-02-19 09:25:46 +01:00
parent 745d383af1
commit 2c94388179
12 changed files with 2 additions and 1714 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
findbugs-exclude-filter.xml
View File

@ -312,28 +312,12 @@
<Bug pattern="EQ_UNUSUAL" /> <Bug pattern="EQ_UNUSUAL" />
</Match> </Match>
<Match> <Match>
<Class name="org.apache.commons.math4.stat.clustering.EuclideanIntegerPoint"/> <Class name="org.apache.commons.math4.ml.clustering.DoublePoint"/>
<Method name="&lt;init>" params="int[]" returns="void" />
<Bug pattern="EI_EXPOSE_REP2" />
</Match>
<Match>
<Class name="org.apache.commons.math4.stat.clustering.EuclideanIntegerPoint"/>
<Method name="getPoint" params="" returns="int[]" />
<Bug pattern="EI_EXPOSE_REP" />
</Match>
<Match>
<Or>
<Class name="org.apache.commons.math4.stat.clustering.EuclideanDoublePoint"/>
<Class name="org.apache.commons.math4.ml.clustering.DoublePoint"/>
</Or>
<Method name="&lt;init>" params="double[]" returns="void" /> <Method name="&lt;init>" params="double[]" returns="void" />
<Bug pattern="EI_EXPOSE_REP2" /> <Bug pattern="EI_EXPOSE_REP2" />
</Match> </Match>
<Match> <Match>
<Or> <Class name="org.apache.commons.math4.ml.clustering.DoublePoint"/>
<Class name="org.apache.commons.math4.stat.clustering.EuclideanDoublePoint"/>
<Class name="org.apache.commons.math4.ml.clustering.DoublePoint"/>
</Or>
<Method name="getPoint" params="" returns="double[]" /> <Method name="getPoint" params="" returns="double[]" />
<Bug pattern="EI_EXPOSE_REP" /> <Bug pattern="EI_EXPOSE_REP" />
</Match> </Match>

76
src/main/java/org/apache/commons/math4/stat/clustering/Cluster.java
View File

@ -1,76 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.List;
/**
* Cluster holding a set of {@link Clusterable} points.
* @param <T> the type of points that can be clustered
* @since 2.0
* @deprecated As of 3.2 (to be removed in 4.0),
* use {@link org.apache.commons.math4.ml.clustering.Cluster} instead
*/
@Deprecated
public class Cluster<T extends Clusterable<T>> implements Serializable {
/** Serializable version identifier. */
private static final long serialVersionUID = -3442297081515880464L;
/** The points contained in this cluster. */
private final List<T> points;
/** Center of the cluster. */
private final T center;
/**
* Build a cluster centered at a specified point.
* @param center the point which is to be the center of this cluster
*/
public Cluster(final T center) {
this.center = center;
points = new ArrayList<T>();
}
/**
* Add a point to this cluster.
* @param point point to add
*/
public void addPoint(final T point) {
points.add(point);
}
/**
* Get the points contained in the cluster.
* @return points contained in the cluster
*/
public List<T> getPoints() {
return points;
}
/**
* Get the point chosen to be the center of this cluster.
* @return chosen cluster center
*/
public T getCenter() {
return center;
}
}

48
src/main/java/org/apache/commons/math4/stat/clustering/Clusterable.java
View File

@ -1,48 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.Collection;
/**
* Interface for points that can be clustered together.
* @param <T> the type of point that can be clustered
* @since 2.0
* @deprecated As of 3.2 (to be removed in 4.0),
* use {@link org.apache.commons.math4.ml.clustering.Clusterable} instead
*/
@Deprecated
public interface Clusterable<T> {
/**
* Returns the distance from the given point.
*
* @param p the point to compute the distance from
* @return the distance from the given point
*/
double distanceFrom(T p);
/**
* Returns the centroid of the given Collection of points.
*
* @param p the Collection of points to compute the centroid of
* @return the centroid of the given Collection of Points
*/
T centroidOf(Collection<T> p);
}

226
src/main/java/org/apache/commons/math4/stat/clustering/DBSCANClusterer.java
View File

@ -1,226 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.ArrayList;
import java.util.Collection;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
import org.apache.commons.math4.exception.NotPositiveException;
import org.apache.commons.math4.exception.NullArgumentException;
import org.apache.commons.math4.util.MathUtils;
/**
* DBSCAN (density-based spatial clustering of applications with noise) algorithm.
* <p>
* The DBSCAN algorithm forms clusters based on the idea of density connectivity, i.e.
* a point p is density connected to another point q, if there exists a chain of
* points p<sub>i</sub>, with i = 1 .. n and p<sub>1</sub> = p and p<sub>n</sub> = q,
* such that each pair &lt;p<sub>i</sub>, p<sub>i+1</sub>&gt; is directly density-reachable.
* A point q is directly density-reachable from point p if it is in the &epsilon;-neighborhood
* of this point.
* <p>
* Any point that is not density-reachable from a formed cluster is treated as noise, and
* will thus not be present in the result.
* <p>
* The algorithm requires two parameters:
* <ul>
* <li>eps: the distance that defines the &epsilon;-neighborhood of a point
* <li>minPoints: the minimum number of density-connected points required to form a cluster
* </ul>
* <p>
* <b>Note:</b> as DBSCAN is not a centroid-based clustering algorithm, the resulting
* {@link Cluster} objects will have no defined center, i.e. {@link Cluster#getCenter()} will
* return {@code null}.
*
* @param <T> type of the points to cluster
* @see <a href="http://en.wikipedia.org/wiki/DBSCAN">DBSCAN (wikipedia)</a>
* @see <a href="http://www.dbs.ifi.lmu.de/Publikationen/Papers/KDD-96.final.frame.pdf">
* A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise</a>
* @since 3.1
* @deprecated As of 3.2 (to be removed in 4.0),
* use {@link org.apache.commons.math4.ml.clustering.DBSCANClusterer} instead
*/
@Deprecated
public class DBSCANClusterer<T extends Clusterable<T>> {
/** Maximum radius of the neighborhood to be considered. */
private final double eps;
/** Minimum number of points needed for a cluster. */
private final int minPts;
/** Status of a point during the clustering process. */
private enum PointStatus {
/** The point has is considered to be noise. */
NOISE,
/** The point is already part of a cluster. */
PART_OF_CLUSTER
}
/**
* Creates a new instance of a DBSCANClusterer.
*
* @param eps maximum radius of the neighborhood to be considered
* @param minPts minimum number of points needed for a cluster
* @throws NotPositiveException if {@code eps < 0.0} or {@code minPts < 0}
*/
public DBSCANClusterer(final double eps, final int minPts)
throws NotPositiveException {
if (eps < 0.0d) {
throw new NotPositiveException(eps);
}
if (minPts < 0) {
throw new NotPositiveException(minPts);
}
this.eps = eps;
this.minPts = minPts;
}
/**
* Returns the maximum radius of the neighborhood to be considered.
*
* @return maximum radius of the neighborhood
*/
public double getEps() {
return eps;
}
/**
* Returns the minimum number of points needed for a cluster.
*
* @return minimum number of points needed for a cluster
*/
public int getMinPts() {
return minPts;
}
/**
* Performs DBSCAN cluster analysis.
* <p>
* <b>Note:</b> as DBSCAN is not a centroid-based clustering algorithm, the resulting
* {@link Cluster} objects will have no defined center, i.e. {@link Cluster#getCenter()} will
* return {@code null}.
*
* @param points the points to cluster
* @return the list of clusters
* @throws NullArgumentException if the data points are null
*/
public List<Cluster<T>> cluster(final Collection<T> points) throws NullArgumentException {
// sanity checks
MathUtils.checkNotNull(points);
final List<Cluster<T>> clusters = new ArrayList<Cluster<T>>();
final Map<Clusterable<T>, PointStatus> visited = new HashMap<Clusterable<T>, PointStatus>();
for (final T point : points) {
if (visited.get(point) != null) {
continue;
}
final List<T> neighbors = getNeighbors(point, points);
if (neighbors.size() >= minPts) {
// DBSCAN does not care about center points
final Cluster<T> cluster = new Cluster<T>(null);
clusters.add(expandCluster(cluster, point, neighbors, points, visited));
} else {
visited.put(point, PointStatus.NOISE);
}
}
return clusters;
}
/**
* Expands the cluster to include density-reachable items.
*
* @param cluster Cluster to expand
* @param point Point to add to cluster
* @param neighbors List of neighbors
* @param points the data set
* @param visited the set of already visited points
* @return the expanded cluster
*/
private Cluster<T> expandCluster(final Cluster<T> cluster,
final T point,
final List<T> neighbors,
final Collection<T> points,
final Map<Clusterable<T>, PointStatus> visited) {
cluster.addPoint(point);
visited.put(point, PointStatus.PART_OF_CLUSTER);
List<T> seeds = new ArrayList<T>(neighbors);
int index = 0;
while (index < seeds.size()) {
final T current = seeds.get(index);
PointStatus pStatus = visited.get(current);
// only check non-visited points
if (pStatus == null) {
final List<T> currentNeighbors = getNeighbors(current, points);
if (currentNeighbors.size() >= minPts) {
seeds = merge(seeds, currentNeighbors);
}
}
if (pStatus != PointStatus.PART_OF_CLUSTER) {
visited.put(current, PointStatus.PART_OF_CLUSTER);
cluster.addPoint(current);
}
index++;
}
return cluster;
}
/**
* Returns a list of density-reachable neighbors of a {@code point}.
*
* @param point the point to look for
* @param points possible neighbors
* @return the List of neighbors
*/
private List<T> getNeighbors(final T point, final Collection<T> points) {
final List<T> neighbors = new ArrayList<T>();
for (final T neighbor : points) {
if (point != neighbor && neighbor.distanceFrom(point) <= eps) {
neighbors.add(neighbor);
}
}
return neighbors;
}
/**
* Merges two lists together.
*
* @param one first list
* @param two second list
* @return merged lists
*/
private List<T> merge(final List<T> one, final List<T> two) {
final Set<T> oneSet = new HashSet<T>(one);
for (T item : two) {
if (!oneSet.contains(item)) {
one.add(item);
}
}
return one;
}
}

100
src/main/java/org/apache/commons/math4/stat/clustering/EuclideanDoublePoint.java
View File

@ -1,100 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.io.Serializable;
import java.util.Collection;
import java.util.Arrays;
import org.apache.commons.math4.util.MathArrays;
/**
* A simple implementation of {@link Clusterable} for points with double coordinates.
* @since 3.1
* @deprecated As of 3.2 (to be removed in 4.0),
* use {@link org.apache.commons.math4.ml.clustering.DoublePoint} instead
*/
@Deprecated
public class EuclideanDoublePoint implements Clusterable<EuclideanDoublePoint>, Serializable {
/** Serializable version identifier. */
private static final long serialVersionUID = 8026472786091227632L;
/** Point coordinates. */
private final double[] point;
/**
* Build an instance wrapping an integer array.
* <p>
* The wrapped array is referenced, it is <em>not</em> copied.
*
* @param point the n-dimensional point in integer space
*/
public EuclideanDoublePoint(final double[] point) {
this.point = point;
}
/** {@inheritDoc} */
public EuclideanDoublePoint centroidOf(final Collection<EuclideanDoublePoint> points) {
final double[] centroid = new double[getPoint().length];
for (final EuclideanDoublePoint p : points) {
for (int i = 0; i < centroid.length; i++) {
centroid[i] += p.getPoint()[i];
}
}
for (int i = 0; i < centroid.length; i++) {
centroid[i] /= points.size();
}
return new EuclideanDoublePoint(centroid);
}
/** {@inheritDoc} */
public double distanceFrom(final EuclideanDoublePoint p) {
return MathArrays.distance(point, p.getPoint());
}
/** {@inheritDoc} */
@Override
public boolean equals(final Object other) {
if (!(other instanceof EuclideanDoublePoint)) {
return false;
}
return Arrays.equals(point, ((EuclideanDoublePoint) other).point);
}
/**
* Get the n-dimensional point in integer space.
*
* @return a reference (not a copy!) to the wrapped array
*/
public double[] getPoint() {
return point;
}
/** {@inheritDoc} */
@Override
public int hashCode() {
return Arrays.hashCode(point);
}
/** {@inheritDoc} */
@Override
public String toString() {
return Arrays.toString(point);
}
}

101
src/main/java/org/apache/commons/math4/stat/clustering/EuclideanIntegerPoint.java
View File

@ -1,101 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.io.Serializable;
import java.util.Arrays;
import java.util.Collection;
import org.apache.commons.math4.util.MathArrays;
/**
* A simple implementation of {@link Clusterable} for points with integer coordinates.
* @since 2.0
* @deprecated As of 3.2 (to be removed in 4.0),
* use {@link org.apache.commons.math4.ml.clustering.DoublePoint} instead
*/
@Deprecated
public class EuclideanIntegerPoint implements Clusterable<EuclideanIntegerPoint>, Serializable {
/** Serializable version identifier. */
private static final long serialVersionUID = 3946024775784901369L;
/** Point coordinates. */
private final int[] point;
/**
* Build an instance wrapping an integer array.
* <p>The wrapped array is referenced, it is <em>not</em> copied.</p>
* @param point the n-dimensional point in integer space
*/
public EuclideanIntegerPoint(final int[] point) {
this.point = point;
}
/**
* Get the n-dimensional point in integer space.
* @return a reference (not a copy!) to the wrapped array
*/
public int[] getPoint() {
return point;
}
/** {@inheritDoc} */
public double distanceFrom(final EuclideanIntegerPoint p) {
return MathArrays.distance(point, p.getPoint());
}
/** {@inheritDoc} */
public EuclideanIntegerPoint centroidOf(final Collection<EuclideanIntegerPoint> points) {
int[] centroid = new int[getPoint().length];
for (EuclideanIntegerPoint p : points) {
for (int i = 0; i < centroid.length; i++) {
centroid[i] += p.getPoint()[i];
}
}
for (int i = 0; i < centroid.length; i++) {
centroid[i] /= points.size();
}
return new EuclideanIntegerPoint(centroid);
}
/** {@inheritDoc} */
@Override
public boolean equals(final Object other) {
if (!(other instanceof EuclideanIntegerPoint)) {
return false;
}
return Arrays.equals(point, ((EuclideanIntegerPoint) other).point);
}
/** {@inheritDoc} */
@Override
public int hashCode() {
return Arrays.hashCode(point);
}
/**
* {@inheritDoc}
* @since 2.1
*/
@Override
public String toString() {
return Arrays.toString(point);
}
}

514
src/main/java/org/apache/commons/math4/stat/clustering/KMeansPlusPlusClusterer.java
View File

@ -1,514 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Collections;
import java.util.List;
import java.util.Random;
import org.apache.commons.math4.exception.ConvergenceException;
import org.apache.commons.math4.exception.MathIllegalArgumentException;
import org.apache.commons.math4.exception.NumberIsTooSmallException;
import org.apache.commons.math4.exception.util.LocalizedFormats;
import org.apache.commons.math4.stat.descriptive.moment.Variance;
import org.apache.commons.math4.util.MathUtils;
/**
* Clustering algorithm based on David Arthur and Sergei Vassilvitski k-means++ algorithm.
* @param <T> type of the points to cluster
* @see <a href="http://en.wikipedia.org/wiki/K-means%2B%2B">K-means++ (wikipedia)</a>
* @since 2.0
* @deprecated As of 3.2 (to be removed in 4.0),
* use {@link org.apache.commons.math4.ml.clustering.KMeansPlusPlusClusterer} instead
*/
@Deprecated
public class KMeansPlusPlusClusterer<T extends Clusterable<T>> {
/** Strategies to use for replacing an empty cluster. */
public static enum EmptyClusterStrategy {
/** Split the cluster with largest distance variance. */
LARGEST_VARIANCE,
/** Split the cluster with largest number of points. */
LARGEST_POINTS_NUMBER,
/** Create a cluster around the point farthest from its centroid. */
FARTHEST_POINT,
/** Generate an error. */
ERROR
}
/** Random generator for choosing initial centers. */
private final Random random;
/** Selected strategy for empty clusters. */
private final EmptyClusterStrategy emptyStrategy;
/** Build a clusterer.
* <p>
* The default strategy for handling empty clusters that may appear during
* algorithm iterations is to split the cluster with largest distance variance.
* </p>
* @param random random generator to use for choosing initial centers
*/
public KMeansPlusPlusClusterer(final Random random) {
this(random, EmptyClusterStrategy.LARGEST_VARIANCE);
}
/** Build a clusterer.
* @param random random generator to use for choosing initial centers
* @param emptyStrategy strategy to use for handling empty clusters that
* may appear during algorithm iterations
* @since 2.2
*/
public KMeansPlusPlusClusterer(final Random random, final EmptyClusterStrategy emptyStrategy) {
this.random = random;
this.emptyStrategy = emptyStrategy;
}
/**
* Runs the K-means++ clustering algorithm.
*
* @param points the points to cluster
* @param k the number of clusters to split the data into
* @param numTrials number of trial runs
* @param maxIterationsPerTrial the maximum number of iterations to run the algorithm
* for at each trial run. If negative, no maximum will be used
* @return a list of clusters containing the points
* @throws MathIllegalArgumentException if the data points are null or the number
* of clusters is larger than the number of data points
* @throws ConvergenceException if an empty cluster is encountered and the
* {@link #emptyStrategy} is set to {@code ERROR}
*/
public List<Cluster<T>> cluster(final Collection<T> points, final int k,
int numTrials, int maxIterationsPerTrial)
throws MathIllegalArgumentException, ConvergenceException {
// at first, we have not found any clusters list yet
List<Cluster<T>> best = null;
double bestVarianceSum = Double.POSITIVE_INFINITY;
// do several clustering trials
for (int i = 0; i < numTrials; ++i) {
// compute a clusters list
List<Cluster<T>> clusters = cluster(points, k, maxIterationsPerTrial);
// compute the variance of the current list
double varianceSum = 0.0;
for (final Cluster<T> cluster : clusters) {
if (!cluster.getPoints().isEmpty()) {
// compute the distance variance of the current cluster
final T center = cluster.getCenter();
final Variance stat = new Variance();
for (final T point : cluster.getPoints()) {
stat.increment(point.distanceFrom(center));
}
varianceSum += stat.getResult();
}
}
if (varianceSum <= bestVarianceSum) {
// this one is the best we have found so far, remember it
best = clusters;
bestVarianceSum = varianceSum;
}
}
// return the best clusters list found
return best;
}
/**
* Runs the K-means++ clustering algorithm.
*
* @param points the points to cluster
* @param k the number of clusters to split the data into
* @param maxIterations the maximum number of iterations to run the algorithm
* for. If negative, no maximum will be used
* @return a list of clusters containing the points
* @throws MathIllegalArgumentException if the data points are null or the number
* of clusters is larger than the number of data points
* @throws ConvergenceException if an empty cluster is encountered and the
* {@link #emptyStrategy} is set to {@code ERROR}
*/
public List<Cluster<T>> cluster(final Collection<T> points, final int k,
final int maxIterations)
throws MathIllegalArgumentException, ConvergenceException {
// sanity checks
MathUtils.checkNotNull(points);
// number of clusters has to be smaller or equal the number of data points
if (points.size() < k) {
throw new NumberIsTooSmallException(points.size(), k, false);
}
// create the initial clusters
List<Cluster<T>> clusters = chooseInitialCenters(points, k, random);
// create an array containing the latest assignment of a point to a cluster
// no need to initialize the array, as it will be filled with the first assignment
int[] assignments = new int[points.size()];
assignPointsToClusters(clusters, points, assignments);
// iterate through updating the centers until we're done
final int max = (maxIterations < 0) ? Integer.MAX_VALUE : maxIterations;
for (int count = 0; count < max; count++) {
boolean emptyCluster = false;
List<Cluster<T>> newClusters = new ArrayList<Cluster<T>>();
for (final Cluster<T> cluster : clusters) {
final T newCenter;
if (cluster.getPoints().isEmpty()) {
switch (emptyStrategy) {
case LARGEST_VARIANCE :
newCenter = getPointFromLargestVarianceCluster(clusters);
break;
case LARGEST_POINTS_NUMBER :
newCenter = getPointFromLargestNumberCluster(clusters);
break;
case FARTHEST_POINT :
newCenter = getFarthestPoint(clusters);
break;
default :
throw new ConvergenceException(LocalizedFormats.EMPTY_CLUSTER_IN_K_MEANS);
}
emptyCluster = true;
} else {
newCenter = cluster.getCenter().centroidOf(cluster.getPoints());
}
newClusters.add(new Cluster<T>(newCenter));
}
int changes = assignPointsToClusters(newClusters, points, assignments);
clusters = newClusters;
// if there were no more changes in the point-to-cluster assignment
// and there are no empty clusters left, return the current clusters
if (changes == 0 && !emptyCluster) {
return clusters;
}
}
return clusters;
}
/**
* Adds the given points to the closest {@link Cluster}.
*
* @param <T> type of the points to cluster
* @param clusters the {@link Cluster}s to add the points to
* @param points the points to add to the given {@link Cluster}s
* @param assignments points assignments to clusters
* @return the number of points assigned to different clusters as the iteration before
*/
private static <T extends Clusterable<T>> int
assignPointsToClusters(final List<Cluster<T>> clusters, final Collection<T> points,
final int[] assignments) {
int assignedDifferently = 0;
int pointIndex = 0;
for (final T p : points) {
int clusterIndex = getNearestCluster(clusters, p);
if (clusterIndex != assignments[pointIndex]) {
assignedDifferently++;
}
Cluster<T> cluster = clusters.get(clusterIndex);
cluster.addPoint(p);
assignments[pointIndex++] = clusterIndex;
}
return assignedDifferently;
}
/**
* Use K-means++ to choose the initial centers.
*
* @param <T> type of the points to cluster
* @param points the points to choose the initial centers from
* @param k the number of centers to choose
* @param random random generator to use
* @return the initial centers
*/
private static <T extends Clusterable<T>> List<Cluster<T>>
chooseInitialCenters(final Collection<T> points, final int k, final Random random) {
// Convert to list for indexed access. Make it unmodifiable, since removal of items
// would screw up the logic of this method.
final List<T> pointList = Collections.unmodifiableList(new ArrayList<T> (points));
// The number of points in the list.
final int numPoints = pointList.size();
// Set the corresponding element in this array to indicate when
// elements of pointList are no longer available.
final boolean[] taken = new boolean[numPoints];
// The resulting list of initial centers.
final List<Cluster<T>> resultSet = new ArrayList<Cluster<T>>();
// Choose one center uniformly at random from among the data points.
final int firstPointIndex = random.nextInt(numPoints);
final T firstPoint = pointList.get(firstPointIndex);
resultSet.add(new Cluster<T>(firstPoint));
// Must mark it as taken
taken[firstPointIndex] = true;
// To keep track of the minimum distance squared of elements of
// pointList to elements of resultSet.
final double[] minDistSquared = new double[numPoints];
// Initialize the elements. Since the only point in resultSet is firstPoint,
// this is very easy.
for (int i = 0; i < numPoints; i++) {
if (i != firstPointIndex) { // That point isn't considered
double d = firstPoint.distanceFrom(pointList.get(i));
minDistSquared[i] = d*d;
}
}
while (resultSet.size() < k) {
// Sum up the squared distances for the points in pointList not
// already taken.
double distSqSum = 0.0;
for (int i = 0; i < numPoints; i++) {
if (!taken[i]) {
distSqSum += minDistSquared[i];
}
}
// Add one new data point as a center. Each point x is chosen with
// probability proportional to D(x)2
final double r = random.nextDouble() * distSqSum;
// The index of the next point to be added to the resultSet.
int nextPointIndex = -1;
// Sum through the squared min distances again, stopping when
// sum >= r.
double sum = 0.0;
for (int i = 0; i < numPoints; i++) {
if (!taken[i]) {
sum += minDistSquared[i];
if (sum >= r) {
nextPointIndex = i;
break;
}
}
}
// If it's not set to >= 0, the point wasn't found in the previous
// for loop, probably because distances are extremely small. Just pick
// the last available point.
if (nextPointIndex == -1) {
for (int i = numPoints - 1; i >= 0; i--) {
if (!taken[i]) {
nextPointIndex = i;
break;
}
}
}
// We found one.
if (nextPointIndex >= 0) {
final T p = pointList.get(nextPointIndex);
resultSet.add(new Cluster<T> (p));
// Mark it as taken.
taken[nextPointIndex] = true;
if (resultSet.size() < k) {
// Now update elements of minDistSquared. We only have to compute
// the distance to the new center to do this.
for (int j = 0; j < numPoints; j++) {
// Only have to worry about the points still not taken.
if (!taken[j]) {
double d = p.distanceFrom(pointList.get(j));
double d2 = d * d;
if (d2 < minDistSquared[j]) {
minDistSquared[j] = d2;
}
}
}
}
} else {
// None found --
// Break from the while loop to prevent
// an infinite loop.
break;
}
}
return resultSet;
}
/**
* Get a random point from the {@link Cluster} with the largest distance variance.
*
* @param clusters the {@link Cluster}s to search
* @return a random point from the selected cluster
* @throws ConvergenceException if clusters are all empty
*/
private T getPointFromLargestVarianceCluster(final Collection<Cluster<T>> clusters)
throws ConvergenceException {
double maxVariance = Double.NEGATIVE_INFINITY;
Cluster<T> selected = null;
for (final Cluster<T> cluster : clusters) {
if (!cluster.getPoints().isEmpty()) {
// compute the distance variance of the current cluster
final T center = cluster.getCenter();
final Variance stat = new Variance();
for (final T point : cluster.getPoints()) {
stat.increment(point.distanceFrom(center));
}
final double variance = stat.getResult();
// select the cluster with the largest variance
if (variance > maxVariance) {
maxVariance = variance;
selected = cluster;
}
}
}
// did we find at least one non-empty cluster ?
if (selected == null) {
throw new ConvergenceException(LocalizedFormats.EMPTY_CLUSTER_IN_K_MEANS);
}
// extract a random point from the cluster
final List<T> selectedPoints = selected.getPoints();
return selectedPoints.remove(random.nextInt(selectedPoints.size()));
}
/**
* Get a random point from the {@link Cluster} with the largest number of points
*
* @param clusters the {@link Cluster}s to search
* @return a random point from the selected cluster
* @throws ConvergenceException if clusters are all empty
*/
private T getPointFromLargestNumberCluster(final Collection<Cluster<T>> clusters) throws ConvergenceException {
int maxNumber = 0;
Cluster<T> selected = null;
for (final Cluster<T> cluster : clusters) {
// get the number of points of the current cluster
final int number = cluster.getPoints().size();
// select the cluster with the largest number of points
if (number > maxNumber) {
maxNumber = number;
selected = cluster;
}
}
// did we find at least one non-empty cluster ?
if (selected == null) {
throw new ConvergenceException(LocalizedFormats.EMPTY_CLUSTER_IN_K_MEANS);
}
// extract a random point from the cluster
final List<T> selectedPoints = selected.getPoints();
return selectedPoints.remove(random.nextInt(selectedPoints.size()));
}
/**
* Get the point farthest to its cluster center
*
* @param clusters the {@link Cluster}s to search
* @return point farthest to its cluster center
* @throws ConvergenceException if clusters are all empty
*/
private T getFarthestPoint(final Collection<Cluster<T>> clusters) throws ConvergenceException {
double maxDistance = Double.NEGATIVE_INFINITY;
Cluster<T> selectedCluster = null;
int selectedPoint = -1;
for (final Cluster<T> cluster : clusters) {
// get the farthest point
final T center = cluster.getCenter();
final List<T> points = cluster.getPoints();
for (int i = 0; i < points.size(); ++i) {
final double distance = points.get(i).distanceFrom(center);
if (distance > maxDistance) {
maxDistance = distance;
selectedCluster = cluster;
selectedPoint = i;
}
}
}
// did we find at least one non-empty cluster ?
if (selectedCluster == null) {
throw new ConvergenceException(LocalizedFormats.EMPTY_CLUSTER_IN_K_MEANS);
}
return selectedCluster.getPoints().remove(selectedPoint);
}
/**
* Returns the nearest {@link Cluster} to the given point
*
* @param <T> type of the points to cluster
* @param clusters the {@link Cluster}s to search
* @param point the point to find the nearest {@link Cluster} for
* @return the index of the nearest {@link Cluster} to the given point
*/
private static <T extends Clusterable<T>> int
getNearestCluster(final Collection<Cluster<T>> clusters, final T point) {
double minDistance = Double.MAX_VALUE;
int clusterIndex = 0;
int minCluster = 0;
for (final Cluster<T> c : clusters) {
final double distance = point.distanceFrom(c.getCenter());
if (distance < minDistance) {
minDistance = distance;
minCluster = clusterIndex;
}
clusterIndex++;
}
return minCluster;
}
}

29
src/main/java/org/apache/commons/math4/stat/clustering/package-info.java
View File

@ -1,29 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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.
*/
/**
* <h2>All classes and sub-packages of this package are deprecated.</h2>
* <h3>Please use their replacements, to be found under
* <ul>
* <li>{@link org.apache.commons.math4.ml.clustering}</li>
* </ul>
* </h3>
*
* <p>
* Clustering algorithms.
* </p>
*/
package org.apache.commons.math4.stat.clustering;

195
src/test/java/org/apache/commons/math4/stat/clustering/DBSCANClustererTest.java
View File

@ -1,195 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.Arrays;
import java.util.List;
import org.apache.commons.math4.exception.MathIllegalArgumentException;
import org.apache.commons.math4.exception.NullArgumentException;
import org.apache.commons.math4.stat.clustering.Cluster;
import org.apache.commons.math4.stat.clustering.DBSCANClusterer;
import org.apache.commons.math4.stat.clustering.EuclideanDoublePoint;
import org.apache.commons.math4.stat.clustering.EuclideanIntegerPoint;
import org.junit.Assert;
import org.junit.Test;
@Deprecated
public class DBSCANClustererTest {
@Test
public void testCluster() {
// Test data generated using: http://people.cs.nctu.edu.tw/~rsliang/dbscan/testdatagen.html
final EuclideanDoublePoint[] points = new EuclideanDoublePoint[] {
new EuclideanDoublePoint(new double[] { 83.08303244924173, 58.83387754182331 }),
new EuclideanDoublePoint(new double[] { 45.05445510940626, 23.469642649637535 }),
new EuclideanDoublePoint(new double[] { 14.96417921432294, 69.0264096390456 }),
new EuclideanDoublePoint(new double[] { 73.53189604333602, 34.896145021310076 }),
new EuclideanDoublePoint(new double[] { 73.28498173551634, 33.96860806993209 }),
new EuclideanDoublePoint(new double[] { 73.45828098873608, 33.92584423092194 }),
new EuclideanDoublePoint(new double[] { 73.9657889183145, 35.73191006924026 }),
new EuclideanDoublePoint(new double[] { 74.0074097183533, 36.81735596177168 }),
new EuclideanDoublePoint(new double[] { 73.41247541410848, 34.27314856695011 }),
new EuclideanDoublePoint(new double[] { 73.9156256353017, 36.83206791547127 }),
new EuclideanDoublePoint(new double[] { 74.81499205809087, 37.15682749846019 }),
new EuclideanDoublePoint(new double[] { 74.03144880081527, 37.57399178552441 }),
new EuclideanDoublePoint(new double[] { 74.51870941207744, 38.674258946906775 }),
new EuclideanDoublePoint(new double[] { 74.50754595105536, 35.58903978415765 }),
new EuclideanDoublePoint(new double[] { 74.51322752749547, 36.030572259100154 }),
new EuclideanDoublePoint(new double[] { 59.27900996617973, 46.41091720294207 }),
new EuclideanDoublePoint(new double[] { 59.73744793841615, 46.20015558367595 }),
new EuclideanDoublePoint(new double[] { 58.81134076672606, 45.71150126331486 }),
new EuclideanDoublePoint(new double[] { 58.52225539437495, 47.416083617601544 }),
new EuclideanDoublePoint(new double[] { 58.218626647023484, 47.36228902172297 }),
new EuclideanDoublePoint(new double[] { 60.27139669447206, 46.606106348801404 }),
new EuclideanDoublePoint(new double[] { 60.894962462363765, 46.976924697402865 }),
new EuclideanDoublePoint(new double[] { 62.29048673878424, 47.66970563563518 }),
new EuclideanDoublePoint(new double[] { 61.03857608977705, 46.212924720020965 }),
new EuclideanDoublePoint(new double[] { 60.16916214139201, 45.18193661351688 }),
new EuclideanDoublePoint(new double[] { 59.90036905976012, 47.555364347063005 }),
new EuclideanDoublePoint(new double[] { 62.33003634144552, 47.83941489877179 }),
new EuclideanDoublePoint(new double[] { 57.86035536718555, 47.31117930193432 }),
new EuclideanDoublePoint(new double[] { 58.13715479685925, 48.985960494028404 }),
new EuclideanDoublePoint(new double[] { 56.131923963548616, 46.8508904252667 }),
new EuclideanDoublePoint(new double[] { 55.976329887053, 47.46384037658572 }),
new EuclideanDoublePoint(new double[] { 56.23245975235477, 47.940035191131756 }),
new EuclideanDoublePoint(new double[] { 58.51687048212625, 46.622885352699086 }),
new EuclideanDoublePoint(new double[] { 57.85411081905477, 45.95394361577928 }),
new EuclideanDoublePoint(new double[] { 56.445776311447844, 45.162093662656844 }),
new EuclideanDoublePoint(new double[] { 57.36691949656233, 47.50097194337286 }),
new EuclideanDoublePoint(new double[] { 58.243626387557015, 46.114052729681134 }),
new EuclideanDoublePoint(new double[] { 56.27224595635198, 44.799080066150054 }),
new EuclideanDoublePoint(new double[] { 57.606924816500396, 46.94291057763621 }),
new EuclideanDoublePoint(new double[] { 30.18714230041951, 13.877149710431695 }),
new EuclideanDoublePoint(new double[] { 30.449448810657486, 13.490778346545994 }),
new EuclideanDoublePoint(new double[] { 30.295018390286714, 13.264889000216499 }),
new EuclideanDoublePoint(new double[] { 30.160201832884923, 11.89278262341395 }),
new EuclideanDoublePoint(new double[] { 31.341509791789576, 15.282655921997502 }),
new EuclideanDoublePoint(new double[] { 31.68601630325429, 14.756873246748 }),
new EuclideanDoublePoint(new double[] { 29.325963742565364, 12.097849250072613 }),
new EuclideanDoublePoint(new double[] { 29.54820742388256, 13.613295356975868 }),
new EuclideanDoublePoint(new double[] { 28.79359608888626, 10.36352064087987 }),
new EuclideanDoublePoint(new double[] { 31.01284597092308, 12.788479208014905 }),
new EuclideanDoublePoint(new double[] { 27.58509216737002, 11.47570110601373 }),
new EuclideanDoublePoint(new double[] { 28.593799561727792, 10.780998203903437 }),
new EuclideanDoublePoint(new double[] { 31.356105766724795, 15.080316198524088 }),
new EuclideanDoublePoint(new double[] { 31.25948503636755, 13.674329151166603 }),
new EuclideanDoublePoint(new double[] { 32.31590076372959, 14.95261758659035 }),
new EuclideanDoublePoint(new double[] { 30.460413702763617, 15.88402809202671 }),
new EuclideanDoublePoint(new double[] { 32.56178203062154, 14.586076852632686 }),
new EuclideanDoublePoint(new double[] { 32.76138648530468, 16.239837325178087 }),
new EuclideanDoublePoint(new double[] { 30.1829453331884, 14.709592407103628 }),
new EuclideanDoublePoint(new double[] { 29.55088173528202, 15.0651247180067 }),
new EuclideanDoublePoint(new double[] { 29.004155302187428, 14.089665298582986 }),
new EuclideanDoublePoint(new double[] { 29.339624439831823, 13.29096065578051 }),
new EuclideanDoublePoint(new double[] { 30.997460327576846, 14.551914158277214 }),
new EuclideanDoublePoint(new double[] { 30.66784126125276, 16.269703107886016 })
};
final DBSCANClusterer<EuclideanDoublePoint> transformer =
new DBSCANClusterer<EuclideanDoublePoint>(2.0, 5);
final List<Cluster<EuclideanDoublePoint>> clusters = transformer.cluster(Arrays.asList(points));
final List<EuclideanDoublePoint> clusterOne =
Arrays.asList(points[3], points[4], points[5], points[6], points[7], points[8], points[9], points[10],
points[11], points[12], points[13], points[14]);
final List<EuclideanDoublePoint> clusterTwo =
Arrays.asList(points[15], points[16], points[17], points[18], points[19], points[20], points[21],
points[22], points[23], points[24], points[25], points[26], points[27], points[28],
points[29], points[30], points[31], points[32], points[33], points[34], points[35],
points[36], points[37], points[38]);
final List<EuclideanDoublePoint> clusterThree =
Arrays.asList(points[39], points[40], points[41], points[42], points[43], points[44], points[45],
points[46], points[47], points[48], points[49], points[50], points[51], points[52],
points[53], points[54], points[55], points[56], points[57], points[58], points[59],
points[60], points[61], points[62]);
boolean cluster1Found = false;
boolean cluster2Found = false;
boolean cluster3Found = false;
Assert.assertEquals(3, clusters.size());
for (final Cluster<EuclideanDoublePoint> cluster : clusters) {
if (cluster.getPoints().containsAll(clusterOne)) {
cluster1Found = true;
}
if (cluster.getPoints().containsAll(clusterTwo)) {
cluster2Found = true;
}
if (cluster.getPoints().containsAll(clusterThree)) {
cluster3Found = true;
}
}
Assert.assertTrue(cluster1Found);
Assert.assertTrue(cluster2Found);
Assert.assertTrue(cluster3Found);
}
@Test
public void testSingleLink() {
final EuclideanIntegerPoint[] points = {
new EuclideanIntegerPoint(new int[] {10, 10}), // A
new EuclideanIntegerPoint(new int[] {12, 9}),
new EuclideanIntegerPoint(new int[] {10, 8}),
new EuclideanIntegerPoint(new int[] {8, 8}),
new EuclideanIntegerPoint(new int[] {8, 6}),
new EuclideanIntegerPoint(new int[] {7, 7}),
new EuclideanIntegerPoint(new int[] {5, 6}), // B
new EuclideanIntegerPoint(new int[] {14, 8}), // C
new EuclideanIntegerPoint(new int[] {7, 15}), // N - Noise, should not be present
new EuclideanIntegerPoint(new int[] {17, 8}), // D - single-link connected to C should not be present
};
final DBSCANClusterer<EuclideanIntegerPoint> clusterer = new DBSCANClusterer<EuclideanIntegerPoint>(3, 3);
List<Cluster<EuclideanIntegerPoint>> clusters = clusterer.cluster(Arrays.asList(points));
Assert.assertEquals(1, clusters.size());
final List<EuclideanIntegerPoint> clusterOne =
Arrays.asList(points[0], points[1], points[2], points[3], points[4], points[5], points[6], points[7]);
Assert.assertTrue(clusters.get(0).getPoints().containsAll(clusterOne));
}
@Test
public void testGetEps() {
final DBSCANClusterer<EuclideanDoublePoint> transformer = new DBSCANClusterer<EuclideanDoublePoint>(2.0, 5);
Assert.assertEquals(2.0, transformer.getEps(), 0.0);
}
@Test
public void testGetMinPts() {
final DBSCANClusterer<EuclideanDoublePoint> transformer = new DBSCANClusterer<EuclideanDoublePoint>(2.0, 5);
Assert.assertEquals(5, transformer.getMinPts());
}
@Test(expected = MathIllegalArgumentException.class)
public void testNegativeEps() {
new DBSCANClusterer<EuclideanDoublePoint>(-2.0, 5);
}
@Test(expected = MathIllegalArgumentException.class)
public void testNegativeMinPts() {
new DBSCANClusterer<EuclideanDoublePoint>(2.0, -5);
}
@Test(expected = NullArgumentException.class)
public void testNullDataset() {
DBSCANClusterer<EuclideanDoublePoint> clusterer = new DBSCANClusterer<EuclideanDoublePoint>(2.0, 5);
clusterer.cluster(null);
}
}

64
src/test/java/org/apache/commons/math4/stat/clustering/EuclideanDoublePointTest.java
View File

@ -1,64 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.ArrayList;
import java.util.List;
import org.apache.commons.math4.TestUtils;
import org.apache.commons.math4.stat.clustering.EuclideanDoublePoint;
import org.apache.commons.math4.util.FastMath;
import org.junit.Assert;
import org.junit.Test;
@Deprecated
public class EuclideanDoublePointTest {
@Test
public void testArrayIsReference() {
final double[] array = { -3.0, -2.0, -1.0, 0.0, 1.0 };
Assert.assertArrayEquals(array, new EuclideanDoublePoint(array).getPoint(), 1.0e-15);
}
@Test
public void testDistance() {
final EuclideanDoublePoint e1 = new EuclideanDoublePoint(new double[] { -3.0, -2.0, -1.0, 0.0, 1.0 });
final EuclideanDoublePoint e2 = new EuclideanDoublePoint(new double[] { 1.0, 0.0, -1.0, 1.0, 1.0 });
Assert.assertEquals(FastMath.sqrt(21.0), e1.distanceFrom(e2), 1.0e-15);
Assert.assertEquals(0.0, e1.distanceFrom(e1), 1.0e-15);
Assert.assertEquals(0.0, e2.distanceFrom(e2), 1.0e-15);
}
@Test
public void testCentroid() {
final List<EuclideanDoublePoint> list = new ArrayList<EuclideanDoublePoint>();
list.add(new EuclideanDoublePoint(new double[] { 1.0, 3.0 }));
list.add(new EuclideanDoublePoint(new double[] { 2.0, 2.0 }));
list.add(new EuclideanDoublePoint(new double[] { 3.0, 3.0 }));
list.add(new EuclideanDoublePoint(new double[] { 2.0, 4.0 }));
final EuclideanDoublePoint c = list.get(0).centroidOf(list);
Assert.assertEquals(2.0, c.getPoint()[0], 1.0e-15);
Assert.assertEquals(3.0, c.getPoint()[1], 1.0e-15);
}
@Test
public void testSerial() {
final EuclideanDoublePoint p = new EuclideanDoublePoint(new double[] { -3.0, -2.0, -1.0, 0.0, 1.0 });
Assert.assertEquals(p, TestUtils.serializeAndRecover(p));
}
}

66
src/test/java/org/apache/commons/math4/stat/clustering/EuclideanIntegerPointTest.java
View File

@ -1,66 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.ArrayList;
import java.util.List;
import org.apache.commons.math4.TestUtils;
import org.apache.commons.math4.stat.clustering.EuclideanIntegerPoint;
import org.apache.commons.math4.util.FastMath;
import org.junit.Assert;
import org.junit.Test;
@Deprecated
public class EuclideanIntegerPointTest {
@Test
public void testArrayIsReference() {
int[] array = { -3, -2, -1, 0, 1 };
Assert.assertTrue(array == new EuclideanIntegerPoint(array).getPoint());
}
@Test
public void testDistance() {
EuclideanIntegerPoint e1 = new EuclideanIntegerPoint(new int[] { -3, -2, -1, 0, 1 });
EuclideanIntegerPoint e2 = new EuclideanIntegerPoint(new int[] { 1, 0, -1, 1, 1 });
Assert.assertEquals(FastMath.sqrt(21.0), e1.distanceFrom(e2), 1.0e-15);
Assert.assertEquals(0.0, e1.distanceFrom(e1), 1.0e-15);
Assert.assertEquals(0.0, e2.distanceFrom(e2), 1.0e-15);
}
@Test
public void testCentroid() {
List<EuclideanIntegerPoint> list = new ArrayList<EuclideanIntegerPoint>();
list.add(new EuclideanIntegerPoint(new int[] { 1, 3 }));
list.add(new EuclideanIntegerPoint(new int[] { 2, 2 }));
list.add(new EuclideanIntegerPoint(new int[] { 3, 3 }));
list.add(new EuclideanIntegerPoint(new int[] { 2, 4 }));
EuclideanIntegerPoint c = list.get(0).centroidOf(list);
Assert.assertEquals(2, c.getPoint()[0]);
Assert.assertEquals(3, c.getPoint()[1]);
}
@Test
public void testSerial() {
EuclideanIntegerPoint p = new EuclideanIntegerPoint(new int[] { -3, -2, -1, 0, 1 });
Assert.assertEquals(p, TestUtils.serializeAndRecover(p));
}
}

277
src/test/java/org/apache/commons/math4/stat/clustering/KMeansPlusPlusClustererTest.java
View File

@ -1,277 +0,0 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You 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 org.apache.commons.math4.stat.clustering;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.List;
import java.util.Random;
import org.apache.commons.math4.exception.NumberIsTooSmallException;
import org.apache.commons.math4.stat.clustering.Cluster;
import org.apache.commons.math4.stat.clustering.Clusterable;
import org.apache.commons.math4.stat.clustering.EuclideanIntegerPoint;
import org.apache.commons.math4.stat.clustering.KMeansPlusPlusClusterer;
import org.junit.Assert;
import org.junit.Test;
@Deprecated
public class KMeansPlusPlusClustererTest {
@Test
public void dimension2() {
KMeansPlusPlusClusterer<EuclideanIntegerPoint> transformer =
new KMeansPlusPlusClusterer<EuclideanIntegerPoint>(new Random(1746432956321l));
EuclideanIntegerPoint[] points = new EuclideanIntegerPoint[] {
// first expected cluster
new EuclideanIntegerPoint(new int[] { -15, 3 }),
new EuclideanIntegerPoint(new int[] { -15, 4 }),
new EuclideanIntegerPoint(new int[] { -15, 5 }),
new EuclideanIntegerPoint(new int[] { -14, 3 }),
new EuclideanIntegerPoint(new int[] { -14, 5 }),
new EuclideanIntegerPoint(new int[] { -13, 3 }),
new EuclideanIntegerPoint(new int[] { -13, 4 }),
new EuclideanIntegerPoint(new int[] { -13, 5 }),
// second expected cluster
new EuclideanIntegerPoint(new int[] { -1, 0 }),
new EuclideanIntegerPoint(new int[] { -1, -1 }),
new EuclideanIntegerPoint(new int[] { 0, -1 }),
new EuclideanIntegerPoint(new int[] { 1, -1 }),
new EuclideanIntegerPoint(new int[] { 1, -2 }),
// third expected cluster
new EuclideanIntegerPoint(new int[] { 13, 3 }),
new EuclideanIntegerPoint(new int[] { 13, 4 }),
new EuclideanIntegerPoint(new int[] { 14, 4 }),
new EuclideanIntegerPoint(new int[] { 14, 7 }),
new EuclideanIntegerPoint(new int[] { 16, 5 }),
new EuclideanIntegerPoint(new int[] { 16, 6 }),
new EuclideanIntegerPoint(new int[] { 17, 4 }),
new EuclideanIntegerPoint(new int[] { 17, 7 })
};
List<Cluster<EuclideanIntegerPoint>> clusters =
transformer.cluster(Arrays.asList(points), 3, 5, 10);
Assert.assertEquals(3, clusters.size());
boolean cluster1Found = false;
boolean cluster2Found = false;
boolean cluster3Found = false;
for (Cluster<EuclideanIntegerPoint> cluster : clusters) {
int[] center = cluster.getCenter().getPoint();
if (center[0] < 0) {
cluster1Found = true;
Assert.assertEquals(8, cluster.getPoints().size());
Assert.assertEquals(-14, center[0]);
Assert.assertEquals( 4, center[1]);
} else if (center[1] < 0) {
cluster2Found = true;
Assert.assertEquals(5, cluster.getPoints().size());
Assert.assertEquals( 0, center[0]);
Assert.assertEquals(-1, center[1]);
} else {
cluster3Found = true;
Assert.assertEquals(8, cluster.getPoints().size());
Assert.assertEquals(15, center[0]);
Assert.assertEquals(5, center[1]);
}
}
Assert.assertTrue(cluster1Found);
Assert.assertTrue(cluster2Found);
Assert.assertTrue(cluster3Found);
}
/**
* JIRA: MATH-305
*
* Two points, one cluster, one iteration
*/
@Test
public void testPerformClusterAnalysisDegenerate() {
KMeansPlusPlusClusterer<EuclideanIntegerPoint> transformer = new KMeansPlusPlusClusterer<EuclideanIntegerPoint>(
new Random(1746432956321l));
EuclideanIntegerPoint[] points = new EuclideanIntegerPoint[] {
new EuclideanIntegerPoint(new int[] { 1959, 325100 }),
new EuclideanIntegerPoint(new int[] { 1960, 373200 }), };
List<Cluster<EuclideanIntegerPoint>> clusters = transformer.cluster(Arrays.asList(points), 1, 1);
Assert.assertEquals(1, clusters.size());
Assert.assertEquals(2, (clusters.get(0).getPoints().size()));
EuclideanIntegerPoint pt1 = new EuclideanIntegerPoint(new int[] { 1959, 325100 });
EuclideanIntegerPoint pt2 = new EuclideanIntegerPoint(new int[] { 1960, 373200 });
Assert.assertTrue(clusters.get(0).getPoints().contains(pt1));
Assert.assertTrue(clusters.get(0).getPoints().contains(pt2));
}
@Test
public void testCertainSpace() {
KMeansPlusPlusClusterer.EmptyClusterStrategy[] strategies = {
KMeansPlusPlusClusterer.EmptyClusterStrategy.LARGEST_VARIANCE,
KMeansPlusPlusClusterer.EmptyClusterStrategy.LARGEST_POINTS_NUMBER,
KMeansPlusPlusClusterer.EmptyClusterStrategy.FARTHEST_POINT
};
for (KMeansPlusPlusClusterer.EmptyClusterStrategy strategy : strategies) {
KMeansPlusPlusClusterer<EuclideanIntegerPoint> transformer =
new KMeansPlusPlusClusterer<EuclideanIntegerPoint>(new Random(1746432956321l), strategy);
int numberOfVariables = 27;
// initialise testvalues
int position1 = 1;
int position2 = position1 + numberOfVariables;
int position3 = position2 + numberOfVariables;
int position4 = position3 + numberOfVariables;
// testvalues will be multiplied
int multiplier = 1000000;
EuclideanIntegerPoint[] breakingPoints = new EuclideanIntegerPoint[numberOfVariables];
// define the space which will break the cluster algorithm
for (int i = 0; i < numberOfVariables; i++) {
int points[] = { position1, position2, position3, position4 };
// multiply the values
for (int j = 0; j < points.length; j++) {
points[j] *= multiplier;
}
EuclideanIntegerPoint euclideanIntegerPoint = new EuclideanIntegerPoint(points);
breakingPoints[i] = euclideanIntegerPoint;
position1 += numberOfVariables;
position2 += numberOfVariables;
position3 += numberOfVariables;
position4 += numberOfVariables;
}
for (int n = 2; n < 27; ++n) {
List<Cluster<EuclideanIntegerPoint>> clusters =
transformer.cluster(Arrays.asList(breakingPoints), n, 100);
Assert.assertEquals(n, clusters.size());
int sum = 0;
for (Cluster<EuclideanIntegerPoint> cluster : clusters) {
sum += cluster.getPoints().size();
}
Assert.assertEquals(numberOfVariables, sum);
}
}
}
/**
* A helper class for testSmallDistances(). This class is similar to EuclideanIntegerPoint, but
* it defines a different distanceFrom() method that tends to return distances less than 1.
*/
private class CloseIntegerPoint implements Clusterable<CloseIntegerPoint> {
public CloseIntegerPoint(EuclideanIntegerPoint point) {
euclideanPoint = point;
}
public double distanceFrom(CloseIntegerPoint p) {
return euclideanPoint.distanceFrom(p.euclideanPoint) * 0.001;
}
public CloseIntegerPoint centroidOf(Collection<CloseIntegerPoint> p) {
Collection<EuclideanIntegerPoint> euclideanPoints =
new ArrayList<EuclideanIntegerPoint>();
for (CloseIntegerPoint point : p) {
euclideanPoints.add(point.euclideanPoint);
}
return new CloseIntegerPoint(euclideanPoint.centroidOf(euclideanPoints));
}
@Override
public boolean equals(Object o) {
if (!(o instanceof CloseIntegerPoint)) {
return false;
}
CloseIntegerPoint p = (CloseIntegerPoint) o;
return euclideanPoint.equals(p.euclideanPoint);
}
@Override
public int hashCode() {
return euclideanPoint.hashCode();
}
private EuclideanIntegerPoint euclideanPoint;
}
/**
* Test points that are very close together. See issue MATH-546.
*/
@Test
public void testSmallDistances() {
// Create a bunch of CloseIntegerPoints. Most are identical, but one is different by a
// small distance.
int[] repeatedArray = { 0 };
int[] uniqueArray = { 1 };
CloseIntegerPoint repeatedPoint =
new CloseIntegerPoint(new EuclideanIntegerPoint(repeatedArray));
CloseIntegerPoint uniquePoint =
new CloseIntegerPoint(new EuclideanIntegerPoint(uniqueArray));
Collection<CloseIntegerPoint> points = new ArrayList<CloseIntegerPoint>();
final int NUM_REPEATED_POINTS = 10 * 1000;
for (int i = 0; i < NUM_REPEATED_POINTS; ++i) {
points.add(repeatedPoint);
}
points.add(uniquePoint);
// Ask a KMeansPlusPlusClusterer to run zero iterations (i.e., to simply choose initial
// cluster centers).
final long RANDOM_SEED = 0;
final int NUM_CLUSTERS = 2;
final int NUM_ITERATIONS = 0;
KMeansPlusPlusClusterer<CloseIntegerPoint> clusterer =
new KMeansPlusPlusClusterer<CloseIntegerPoint>(new Random(RANDOM_SEED));
List<Cluster<CloseIntegerPoint>> clusters =
clusterer.cluster(points, NUM_CLUSTERS, NUM_ITERATIONS);
// Check that one of the chosen centers is the unique point.
boolean uniquePointIsCenter = false;
for (Cluster<CloseIntegerPoint> cluster : clusters) {
if (cluster.getCenter().equals(uniquePoint)) {
uniquePointIsCenter = true;
}
}
Assert.assertTrue(uniquePointIsCenter);
}
/**
* 2 variables cannot be clustered into 3 clusters. See issue MATH-436.
*/
@Test(expected=NumberIsTooSmallException.class)
public void testPerformClusterAnalysisToManyClusters() {
KMeansPlusPlusClusterer<EuclideanIntegerPoint> transformer =
new KMeansPlusPlusClusterer<EuclideanIntegerPoint>(
new Random(1746432956321l));
EuclideanIntegerPoint[] points = new EuclideanIntegerPoint[] {
new EuclideanIntegerPoint(new int[] {
1959, 325100
}), new EuclideanIntegerPoint(new int[] {
1960, 373200
})
};
transformer.cluster(Arrays.asList(points), 3, 1);
}
}