Mercurial > hg > de.mpg.mpiwg.itgroup.digilib.plugin
view libs/commons-math-2.1/docs/apidocs/src-html/org/apache/commons/math/stat/clustering/KMeansPlusPlusClusterer.html @ 13:cbf34dd4d7e6
commons-math-2.1 added
author | dwinter |
---|---|
date | Tue, 04 Jan 2011 10:02:07 +0100 |
parents | |
children |
line wrap: on
line source
<HTML> <BODY BGCOLOR="white"> <PRE> <FONT color="green">001</FONT> /*<a name="line.1"></a> <FONT color="green">002</FONT> * Licensed to the Apache Software Foundation (ASF) under one or more<a name="line.2"></a> <FONT color="green">003</FONT> * contributor license agreements. See the NOTICE file distributed with<a name="line.3"></a> <FONT color="green">004</FONT> * this work for additional information regarding copyright ownership.<a name="line.4"></a> <FONT color="green">005</FONT> * The ASF licenses this file to You under the Apache License, Version 2.0<a name="line.5"></a> <FONT color="green">006</FONT> * (the "License"); you may not use this file except in compliance with<a name="line.6"></a> <FONT color="green">007</FONT> * the License. You may obtain a copy of the License at<a name="line.7"></a> <FONT color="green">008</FONT> *<a name="line.8"></a> <FONT color="green">009</FONT> * http://www.apache.org/licenses/LICENSE-2.0<a name="line.9"></a> <FONT color="green">010</FONT> *<a name="line.10"></a> <FONT color="green">011</FONT> * Unless required by applicable law or agreed to in writing, software<a name="line.11"></a> <FONT color="green">012</FONT> * distributed under the License is distributed on an "AS IS" BASIS,<a name="line.12"></a> <FONT color="green">013</FONT> * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.<a name="line.13"></a> <FONT color="green">014</FONT> * See the License for the specific language governing permissions and<a name="line.14"></a> <FONT color="green">015</FONT> * limitations under the License.<a name="line.15"></a> <FONT color="green">016</FONT> */<a name="line.16"></a> <FONT color="green">017</FONT> <a name="line.17"></a> <FONT color="green">018</FONT> package org.apache.commons.math.stat.clustering;<a name="line.18"></a> <FONT color="green">019</FONT> <a name="line.19"></a> <FONT color="green">020</FONT> import java.util.ArrayList;<a name="line.20"></a> <FONT color="green">021</FONT> import java.util.Collection;<a name="line.21"></a> <FONT color="green">022</FONT> import java.util.List;<a name="line.22"></a> <FONT color="green">023</FONT> import java.util.Random;<a name="line.23"></a> <FONT color="green">024</FONT> <a name="line.24"></a> <FONT color="green">025</FONT> /**<a name="line.25"></a> <FONT color="green">026</FONT> * Clustering algorithm based on David Arthur and Sergei Vassilvitski k-means++ algorithm.<a name="line.26"></a> <FONT color="green">027</FONT> * @param <T> type of the points to cluster<a name="line.27"></a> <FONT color="green">028</FONT> * @see <a href="http://en.wikipedia.org/wiki/K-means%2B%2B">K-means++ (wikipedia)</a><a name="line.28"></a> <FONT color="green">029</FONT> * @version $Revision: 811685 $ $Date: 2009-09-05 13:36:48 -0400 (Sat, 05 Sep 2009) $<a name="line.29"></a> <FONT color="green">030</FONT> * @since 2.0<a name="line.30"></a> <FONT color="green">031</FONT> */<a name="line.31"></a> <FONT color="green">032</FONT> public class KMeansPlusPlusClusterer<T extends Clusterable<T>> {<a name="line.32"></a> <FONT color="green">033</FONT> <a name="line.33"></a> <FONT color="green">034</FONT> /** Random generator for choosing initial centers. */<a name="line.34"></a> <FONT color="green">035</FONT> private final Random random;<a name="line.35"></a> <FONT color="green">036</FONT> <a name="line.36"></a> <FONT color="green">037</FONT> /** Build a clusterer.<a name="line.37"></a> <FONT color="green">038</FONT> * @param random random generator to use for choosing initial centers<a name="line.38"></a> <FONT color="green">039</FONT> */<a name="line.39"></a> <FONT color="green">040</FONT> public KMeansPlusPlusClusterer(final Random random) {<a name="line.40"></a> <FONT color="green">041</FONT> this.random = random;<a name="line.41"></a> <FONT color="green">042</FONT> }<a name="line.42"></a> <FONT color="green">043</FONT> <a name="line.43"></a> <FONT color="green">044</FONT> /**<a name="line.44"></a> <FONT color="green">045</FONT> * Runs the K-means++ clustering algorithm.<a name="line.45"></a> <FONT color="green">046</FONT> *<a name="line.46"></a> <FONT color="green">047</FONT> * @param points the points to cluster<a name="line.47"></a> <FONT color="green">048</FONT> * @param k the number of clusters to split the data into<a name="line.48"></a> <FONT color="green">049</FONT> * @param maxIterations the maximum number of iterations to run the algorithm<a name="line.49"></a> <FONT color="green">050</FONT> * for. If negative, no maximum will be used<a name="line.50"></a> <FONT color="green">051</FONT> * @return a list of clusters containing the points<a name="line.51"></a> <FONT color="green">052</FONT> */<a name="line.52"></a> <FONT color="green">053</FONT> public List<Cluster<T>> cluster(final Collection<T> points,<a name="line.53"></a> <FONT color="green">054</FONT> final int k, final int maxIterations) {<a name="line.54"></a> <FONT color="green">055</FONT> // create the initial clusters<a name="line.55"></a> <FONT color="green">056</FONT> List<Cluster<T>> clusters = chooseInitialCenters(points, k, random);<a name="line.56"></a> <FONT color="green">057</FONT> assignPointsToClusters(clusters, points);<a name="line.57"></a> <FONT color="green">058</FONT> <a name="line.58"></a> <FONT color="green">059</FONT> // iterate through updating the centers until we're done<a name="line.59"></a> <FONT color="green">060</FONT> final int max = (maxIterations < 0) ? Integer.MAX_VALUE : maxIterations;<a name="line.60"></a> <FONT color="green">061</FONT> for (int count = 0; count < max; count++) {<a name="line.61"></a> <FONT color="green">062</FONT> boolean clusteringChanged = false;<a name="line.62"></a> <FONT color="green">063</FONT> List<Cluster<T>> newClusters = new ArrayList<Cluster<T>>();<a name="line.63"></a> <FONT color="green">064</FONT> for (final Cluster<T> cluster : clusters) {<a name="line.64"></a> <FONT color="green">065</FONT> final T newCenter = cluster.getCenter().centroidOf(cluster.getPoints());<a name="line.65"></a> <FONT color="green">066</FONT> if (!newCenter.equals(cluster.getCenter())) {<a name="line.66"></a> <FONT color="green">067</FONT> clusteringChanged = true;<a name="line.67"></a> <FONT color="green">068</FONT> }<a name="line.68"></a> <FONT color="green">069</FONT> newClusters.add(new Cluster<T>(newCenter));<a name="line.69"></a> <FONT color="green">070</FONT> }<a name="line.70"></a> <FONT color="green">071</FONT> if (!clusteringChanged) {<a name="line.71"></a> <FONT color="green">072</FONT> return clusters;<a name="line.72"></a> <FONT color="green">073</FONT> }<a name="line.73"></a> <FONT color="green">074</FONT> assignPointsToClusters(newClusters, points);<a name="line.74"></a> <FONT color="green">075</FONT> clusters = newClusters;<a name="line.75"></a> <FONT color="green">076</FONT> }<a name="line.76"></a> <FONT color="green">077</FONT> return clusters;<a name="line.77"></a> <FONT color="green">078</FONT> }<a name="line.78"></a> <FONT color="green">079</FONT> <a name="line.79"></a> <FONT color="green">080</FONT> /**<a name="line.80"></a> <FONT color="green">081</FONT> * Adds the given points to the closest {@link Cluster}.<a name="line.81"></a> <FONT color="green">082</FONT> *<a name="line.82"></a> <FONT color="green">083</FONT> * @param <T> type of the points to cluster<a name="line.83"></a> <FONT color="green">084</FONT> * @param clusters the {@link Cluster}s to add the points to<a name="line.84"></a> <FONT color="green">085</FONT> * @param points the points to add to the given {@link Cluster}s<a name="line.85"></a> <FONT color="green">086</FONT> */<a name="line.86"></a> <FONT color="green">087</FONT> private static <T extends Clusterable<T>> void<a name="line.87"></a> <FONT color="green">088</FONT> assignPointsToClusters(final Collection<Cluster<T>> clusters, final Collection<T> points) {<a name="line.88"></a> <FONT color="green">089</FONT> for (final T p : points) {<a name="line.89"></a> <FONT color="green">090</FONT> Cluster<T> cluster = getNearestCluster(clusters, p);<a name="line.90"></a> <FONT color="green">091</FONT> cluster.addPoint(p);<a name="line.91"></a> <FONT color="green">092</FONT> }<a name="line.92"></a> <FONT color="green">093</FONT> }<a name="line.93"></a> <FONT color="green">094</FONT> <a name="line.94"></a> <FONT color="green">095</FONT> /**<a name="line.95"></a> <FONT color="green">096</FONT> * Use K-means++ to choose the initial centers.<a name="line.96"></a> <FONT color="green">097</FONT> *<a name="line.97"></a> <FONT color="green">098</FONT> * @param <T> type of the points to cluster<a name="line.98"></a> <FONT color="green">099</FONT> * @param points the points to choose the initial centers from<a name="line.99"></a> <FONT color="green">100</FONT> * @param k the number of centers to choose<a name="line.100"></a> <FONT color="green">101</FONT> * @param random random generator to use<a name="line.101"></a> <FONT color="green">102</FONT> * @return the initial centers<a name="line.102"></a> <FONT color="green">103</FONT> */<a name="line.103"></a> <FONT color="green">104</FONT> private static <T extends Clusterable<T>> List<Cluster<T>><a name="line.104"></a> <FONT color="green">105</FONT> chooseInitialCenters(final Collection<T> points, final int k, final Random random) {<a name="line.105"></a> <FONT color="green">106</FONT> <a name="line.106"></a> <FONT color="green">107</FONT> final List<T> pointSet = new ArrayList<T>(points);<a name="line.107"></a> <FONT color="green">108</FONT> final List<Cluster<T>> resultSet = new ArrayList<Cluster<T>>();<a name="line.108"></a> <FONT color="green">109</FONT> <a name="line.109"></a> <FONT color="green">110</FONT> // Choose one center uniformly at random from among the data points.<a name="line.110"></a> <FONT color="green">111</FONT> final T firstPoint = pointSet.remove(random.nextInt(pointSet.size()));<a name="line.111"></a> <FONT color="green">112</FONT> resultSet.add(new Cluster<T>(firstPoint));<a name="line.112"></a> <FONT color="green">113</FONT> <a name="line.113"></a> <FONT color="green">114</FONT> final double[] dx2 = new double[pointSet.size()];<a name="line.114"></a> <FONT color="green">115</FONT> while (resultSet.size() < k) {<a name="line.115"></a> <FONT color="green">116</FONT> // For each data point x, compute D(x), the distance between x and<a name="line.116"></a> <FONT color="green">117</FONT> // the nearest center that has already been chosen.<a name="line.117"></a> <FONT color="green">118</FONT> int sum = 0;<a name="line.118"></a> <FONT color="green">119</FONT> for (int i = 0; i < pointSet.size(); i++) {<a name="line.119"></a> <FONT color="green">120</FONT> final T p = pointSet.get(i);<a name="line.120"></a> <FONT color="green">121</FONT> final Cluster<T> nearest = getNearestCluster(resultSet, p);<a name="line.121"></a> <FONT color="green">122</FONT> final double d = p.distanceFrom(nearest.getCenter());<a name="line.122"></a> <FONT color="green">123</FONT> sum += d * d;<a name="line.123"></a> <FONT color="green">124</FONT> dx2[i] = sum;<a name="line.124"></a> <FONT color="green">125</FONT> }<a name="line.125"></a> <FONT color="green">126</FONT> <a name="line.126"></a> <FONT color="green">127</FONT> // Add one new data point as a center. Each point x is chosen with<a name="line.127"></a> <FONT color="green">128</FONT> // probability proportional to D(x)2<a name="line.128"></a> <FONT color="green">129</FONT> final double r = random.nextDouble() * sum;<a name="line.129"></a> <FONT color="green">130</FONT> for (int i = 0 ; i < dx2.length; i++) {<a name="line.130"></a> <FONT color="green">131</FONT> if (dx2[i] >= r) {<a name="line.131"></a> <FONT color="green">132</FONT> final T p = pointSet.remove(i);<a name="line.132"></a> <FONT color="green">133</FONT> resultSet.add(new Cluster<T>(p));<a name="line.133"></a> <FONT color="green">134</FONT> break;<a name="line.134"></a> <FONT color="green">135</FONT> }<a name="line.135"></a> <FONT color="green">136</FONT> }<a name="line.136"></a> <FONT color="green">137</FONT> }<a name="line.137"></a> <FONT color="green">138</FONT> <a name="line.138"></a> <FONT color="green">139</FONT> return resultSet;<a name="line.139"></a> <FONT color="green">140</FONT> <a name="line.140"></a> <FONT color="green">141</FONT> }<a name="line.141"></a> <FONT color="green">142</FONT> <a name="line.142"></a> <FONT color="green">143</FONT> /**<a name="line.143"></a> <FONT color="green">144</FONT> * Returns the nearest {@link Cluster} to the given point<a name="line.144"></a> <FONT color="green">145</FONT> *<a name="line.145"></a> <FONT color="green">146</FONT> * @param <T> type of the points to cluster<a name="line.146"></a> <FONT color="green">147</FONT> * @param clusters the {@link Cluster}s to search<a name="line.147"></a> <FONT color="green">148</FONT> * @param point the point to find the nearest {@link Cluster} for<a name="line.148"></a> <FONT color="green">149</FONT> * @return the nearest {@link Cluster} to the given point<a name="line.149"></a> <FONT color="green">150</FONT> */<a name="line.150"></a> <FONT color="green">151</FONT> private static <T extends Clusterable<T>> Cluster<T><a name="line.151"></a> <FONT color="green">152</FONT> getNearestCluster(final Collection<Cluster<T>> clusters, final T point) {<a name="line.152"></a> <FONT color="green">153</FONT> double minDistance = Double.MAX_VALUE;<a name="line.153"></a> <FONT color="green">154</FONT> Cluster<T> minCluster = null;<a name="line.154"></a> <FONT color="green">155</FONT> for (final Cluster<T> c : clusters) {<a name="line.155"></a> <FONT color="green">156</FONT> final double distance = point.distanceFrom(c.getCenter());<a name="line.156"></a> <FONT color="green">157</FONT> if (distance < minDistance) {<a name="line.157"></a> <FONT color="green">158</FONT> minDistance = distance;<a name="line.158"></a> <FONT color="green">159</FONT> minCluster = c;<a name="line.159"></a> <FONT color="green">160</FONT> }<a name="line.160"></a> <FONT color="green">161</FONT> }<a name="line.161"></a> <FONT color="green">162</FONT> return minCluster;<a name="line.162"></a> <FONT color="green">163</FONT> }<a name="line.163"></a> <FONT color="green">164</FONT> <a name="line.164"></a> <FONT color="green">165</FONT> }<a name="line.165"></a> </PRE> </BODY> </HTML>