Mercurial > hg > de.mpg.mpiwg.itgroup.digilib.plugin

comparison libs/commons-math-2.1/docs/apidocs/src-html/org/apache/commons/math/stat/clustering/KMeansPlusPlusClusterer.html @ 13:cbf34dd4d7e6

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
commons-math-2.1 added
author dwinter
date Tue, 04 Jan 2011 10:02:07 +0100
parents
children
comparison
equal deleted inserted replaced
12:970d26a94fb7 13:cbf34dd4d7e6
1 <HTML>
2 <BODY BGCOLOR="white">
3 <PRE>
4 <FONT color="green">001</FONT> /*<a name="line.1"></a>
5 <FONT color="green">002</FONT> * Licensed to the Apache Software Foundation (ASF) under one or more<a name="line.2"></a>
6 <FONT color="green">003</FONT> * contributor license agreements. See the NOTICE file distributed with<a name="line.3"></a>
7 <FONT color="green">004</FONT> * this work for additional information regarding copyright ownership.<a name="line.4"></a>
8 <FONT color="green">005</FONT> * The ASF licenses this file to You under the Apache License, Version 2.0<a name="line.5"></a>
9 <FONT color="green">006</FONT> * (the "License"); you may not use this file except in compliance with<a name="line.6"></a>
10 <FONT color="green">007</FONT> * the License. You may obtain a copy of the License at<a name="line.7"></a>
11 <FONT color="green">008</FONT> *<a name="line.8"></a>
12 <FONT color="green">009</FONT> * http://www.apache.org/licenses/LICENSE-2.0<a name="line.9"></a>
13 <FONT color="green">010</FONT> *<a name="line.10"></a>
14 <FONT color="green">011</FONT> * Unless required by applicable law or agreed to in writing, software<a name="line.11"></a>
15 <FONT color="green">012</FONT> * distributed under the License is distributed on an "AS IS" BASIS,<a name="line.12"></a>
16 <FONT color="green">013</FONT> * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.<a name="line.13"></a>
17 <FONT color="green">014</FONT> * See the License for the specific language governing permissions and<a name="line.14"></a>
18 <FONT color="green">015</FONT> * limitations under the License.<a name="line.15"></a>
19 <FONT color="green">016</FONT> */<a name="line.16"></a>
20 <FONT color="green">017</FONT> <a name="line.17"></a>
21 <FONT color="green">018</FONT> package org.apache.commons.math.stat.clustering;<a name="line.18"></a>
22 <FONT color="green">019</FONT> <a name="line.19"></a>
23 <FONT color="green">020</FONT> import java.util.ArrayList;<a name="line.20"></a>
24 <FONT color="green">021</FONT> import java.util.Collection;<a name="line.21"></a>
25 <FONT color="green">022</FONT> import java.util.List;<a name="line.22"></a>
26 <FONT color="green">023</FONT> import java.util.Random;<a name="line.23"></a>
27 <FONT color="green">024</FONT> <a name="line.24"></a>
28 <FONT color="green">025</FONT> /**<a name="line.25"></a>
29 <FONT color="green">026</FONT> * Clustering algorithm based on David Arthur and Sergei Vassilvitski k-means++ algorithm.<a name="line.26"></a>
30 <FONT color="green">027</FONT> * @param &lt;T&gt; type of the points to cluster<a name="line.27"></a>
31 <FONT color="green">028</FONT> * @see &lt;a href="http://en.wikipedia.org/wiki/K-means%2B%2B"&gt;K-means++ (wikipedia)&lt;/a&gt;<a name="line.28"></a>
32 <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>
33 <FONT color="green">030</FONT> * @since 2.0<a name="line.30"></a>
34 <FONT color="green">031</FONT> */<a name="line.31"></a>
35 <FONT color="green">032</FONT> public class KMeansPlusPlusClusterer&lt;T extends Clusterable&lt;T&gt;&gt; {<a name="line.32"></a>
36 <FONT color="green">033</FONT> <a name="line.33"></a>
37 <FONT color="green">034</FONT> /** Random generator for choosing initial centers. */<a name="line.34"></a>
38 <FONT color="green">035</FONT> private final Random random;<a name="line.35"></a>
39 <FONT color="green">036</FONT> <a name="line.36"></a>
40 <FONT color="green">037</FONT> /** Build a clusterer.<a name="line.37"></a>
41 <FONT color="green">038</FONT> * @param random random generator to use for choosing initial centers<a name="line.38"></a>
42 <FONT color="green">039</FONT> */<a name="line.39"></a>
43 <FONT color="green">040</FONT> public KMeansPlusPlusClusterer(final Random random) {<a name="line.40"></a>
44 <FONT color="green">041</FONT> this.random = random;<a name="line.41"></a>
45 <FONT color="green">042</FONT> }<a name="line.42"></a>
46 <FONT color="green">043</FONT> <a name="line.43"></a>
47 <FONT color="green">044</FONT> /**<a name="line.44"></a>
48 <FONT color="green">045</FONT> * Runs the K-means++ clustering algorithm.<a name="line.45"></a>
49 <FONT color="green">046</FONT> *<a name="line.46"></a>
50 <FONT color="green">047</FONT> * @param points the points to cluster<a name="line.47"></a>
51 <FONT color="green">048</FONT> * @param k the number of clusters to split the data into<a name="line.48"></a>
52 <FONT color="green">049</FONT> * @param maxIterations the maximum number of iterations to run the algorithm<a name="line.49"></a>
53 <FONT color="green">050</FONT> * for. If negative, no maximum will be used<a name="line.50"></a>
54 <FONT color="green">051</FONT> * @return a list of clusters containing the points<a name="line.51"></a>
55 <FONT color="green">052</FONT> */<a name="line.52"></a>
56 <FONT color="green">053</FONT> public List&lt;Cluster&lt;T&gt;&gt; cluster(final Collection&lt;T&gt; points,<a name="line.53"></a>
57 <FONT color="green">054</FONT> final int k, final int maxIterations) {<a name="line.54"></a>
58 <FONT color="green">055</FONT> // create the initial clusters<a name="line.55"></a>
59 <FONT color="green">056</FONT> List&lt;Cluster&lt;T&gt;&gt; clusters = chooseInitialCenters(points, k, random);<a name="line.56"></a>
60 <FONT color="green">057</FONT> assignPointsToClusters(clusters, points);<a name="line.57"></a>
61 <FONT color="green">058</FONT> <a name="line.58"></a>
62 <FONT color="green">059</FONT> // iterate through updating the centers until we're done<a name="line.59"></a>
63 <FONT color="green">060</FONT> final int max = (maxIterations &lt; 0) ? Integer.MAX_VALUE : maxIterations;<a name="line.60"></a>
64 <FONT color="green">061</FONT> for (int count = 0; count &lt; max; count++) {<a name="line.61"></a>
65 <FONT color="green">062</FONT> boolean clusteringChanged = false;<a name="line.62"></a>
66 <FONT color="green">063</FONT> List&lt;Cluster&lt;T&gt;&gt; newClusters = new ArrayList&lt;Cluster&lt;T&gt;&gt;();<a name="line.63"></a>
67 <FONT color="green">064</FONT> for (final Cluster&lt;T&gt; cluster : clusters) {<a name="line.64"></a>
68 <FONT color="green">065</FONT> final T newCenter = cluster.getCenter().centroidOf(cluster.getPoints());<a name="line.65"></a>
69 <FONT color="green">066</FONT> if (!newCenter.equals(cluster.getCenter())) {<a name="line.66"></a>
70 <FONT color="green">067</FONT> clusteringChanged = true;<a name="line.67"></a>
71 <FONT color="green">068</FONT> }<a name="line.68"></a>
72 <FONT color="green">069</FONT> newClusters.add(new Cluster&lt;T&gt;(newCenter));<a name="line.69"></a>
73 <FONT color="green">070</FONT> }<a name="line.70"></a>
74 <FONT color="green">071</FONT> if (!clusteringChanged) {<a name="line.71"></a>
75 <FONT color="green">072</FONT> return clusters;<a name="line.72"></a>
76 <FONT color="green">073</FONT> }<a name="line.73"></a>
77 <FONT color="green">074</FONT> assignPointsToClusters(newClusters, points);<a name="line.74"></a>
78 <FONT color="green">075</FONT> clusters = newClusters;<a name="line.75"></a>
79 <FONT color="green">076</FONT> }<a name="line.76"></a>
80 <FONT color="green">077</FONT> return clusters;<a name="line.77"></a>
81 <FONT color="green">078</FONT> }<a name="line.78"></a>
82 <FONT color="green">079</FONT> <a name="line.79"></a>
83 <FONT color="green">080</FONT> /**<a name="line.80"></a>
84 <FONT color="green">081</FONT> * Adds the given points to the closest {@link Cluster}.<a name="line.81"></a>
85 <FONT color="green">082</FONT> *<a name="line.82"></a>
86 <FONT color="green">083</FONT> * @param &lt;T&gt; type of the points to cluster<a name="line.83"></a>
87 <FONT color="green">084</FONT> * @param clusters the {@link Cluster}s to add the points to<a name="line.84"></a>
88 <FONT color="green">085</FONT> * @param points the points to add to the given {@link Cluster}s<a name="line.85"></a>
89 <FONT color="green">086</FONT> */<a name="line.86"></a>
90 <FONT color="green">087</FONT> private static &lt;T extends Clusterable&lt;T&gt;&gt; void<a name="line.87"></a>
91 <FONT color="green">088</FONT> assignPointsToClusters(final Collection&lt;Cluster&lt;T&gt;&gt; clusters, final Collection&lt;T&gt; points) {<a name="line.88"></a>
92 <FONT color="green">089</FONT> for (final T p : points) {<a name="line.89"></a>
93 <FONT color="green">090</FONT> Cluster&lt;T&gt; cluster = getNearestCluster(clusters, p);<a name="line.90"></a>
94 <FONT color="green">091</FONT> cluster.addPoint(p);<a name="line.91"></a>
95 <FONT color="green">092</FONT> }<a name="line.92"></a>
96 <FONT color="green">093</FONT> }<a name="line.93"></a>
97 <FONT color="green">094</FONT> <a name="line.94"></a>
98 <FONT color="green">095</FONT> /**<a name="line.95"></a>
99 <FONT color="green">096</FONT> * Use K-means++ to choose the initial centers.<a name="line.96"></a>
100 <FONT color="green">097</FONT> *<a name="line.97"></a>
101 <FONT color="green">098</FONT> * @param &lt;T&gt; type of the points to cluster<a name="line.98"></a>
102 <FONT color="green">099</FONT> * @param points the points to choose the initial centers from<a name="line.99"></a>
103 <FONT color="green">100</FONT> * @param k the number of centers to choose<a name="line.100"></a>
104 <FONT color="green">101</FONT> * @param random random generator to use<a name="line.101"></a>
105 <FONT color="green">102</FONT> * @return the initial centers<a name="line.102"></a>
106 <FONT color="green">103</FONT> */<a name="line.103"></a>
107 <FONT color="green">104</FONT> private static &lt;T extends Clusterable&lt;T&gt;&gt; List&lt;Cluster&lt;T&gt;&gt;<a name="line.104"></a>
108 <FONT color="green">105</FONT> chooseInitialCenters(final Collection&lt;T&gt; points, final int k, final Random random) {<a name="line.105"></a>
109 <FONT color="green">106</FONT> <a name="line.106"></a>
110 <FONT color="green">107</FONT> final List&lt;T&gt; pointSet = new ArrayList&lt;T&gt;(points);<a name="line.107"></a>
111 <FONT color="green">108</FONT> final List&lt;Cluster&lt;T&gt;&gt; resultSet = new ArrayList&lt;Cluster&lt;T&gt;&gt;();<a name="line.108"></a>
112 <FONT color="green">109</FONT> <a name="line.109"></a>
113 <FONT color="green">110</FONT> // Choose one center uniformly at random from among the data points.<a name="line.110"></a>
114 <FONT color="green">111</FONT> final T firstPoint = pointSet.remove(random.nextInt(pointSet.size()));<a name="line.111"></a>
115 <FONT color="green">112</FONT> resultSet.add(new Cluster&lt;T&gt;(firstPoint));<a name="line.112"></a>
116 <FONT color="green">113</FONT> <a name="line.113"></a>
117 <FONT color="green">114</FONT> final double[] dx2 = new double[pointSet.size()];<a name="line.114"></a>
118 <FONT color="green">115</FONT> while (resultSet.size() &lt; k) {<a name="line.115"></a>
119 <FONT color="green">116</FONT> // For each data point x, compute D(x), the distance between x and<a name="line.116"></a>
120 <FONT color="green">117</FONT> // the nearest center that has already been chosen.<a name="line.117"></a>
121 <FONT color="green">118</FONT> int sum = 0;<a name="line.118"></a>
122 <FONT color="green">119</FONT> for (int i = 0; i &lt; pointSet.size(); i++) {<a name="line.119"></a>
123 <FONT color="green">120</FONT> final T p = pointSet.get(i);<a name="line.120"></a>
124 <FONT color="green">121</FONT> final Cluster&lt;T&gt; nearest = getNearestCluster(resultSet, p);<a name="line.121"></a>
125 <FONT color="green">122</FONT> final double d = p.distanceFrom(nearest.getCenter());<a name="line.122"></a>
126 <FONT color="green">123</FONT> sum += d * d;<a name="line.123"></a>
127 <FONT color="green">124</FONT> dx2[i] = sum;<a name="line.124"></a>
128 <FONT color="green">125</FONT> }<a name="line.125"></a>
129 <FONT color="green">126</FONT> <a name="line.126"></a>
130 <FONT color="green">127</FONT> // Add one new data point as a center. Each point x is chosen with<a name="line.127"></a>
131 <FONT color="green">128</FONT> // probability proportional to D(x)2<a name="line.128"></a>
132 <FONT color="green">129</FONT> final double r = random.nextDouble() * sum;<a name="line.129"></a>
133 <FONT color="green">130</FONT> for (int i = 0 ; i &lt; dx2.length; i++) {<a name="line.130"></a>
134 <FONT color="green">131</FONT> if (dx2[i] &gt;= r) {<a name="line.131"></a>
135 <FONT color="green">132</FONT> final T p = pointSet.remove(i);<a name="line.132"></a>
136 <FONT color="green">133</FONT> resultSet.add(new Cluster&lt;T&gt;(p));<a name="line.133"></a>
137 <FONT color="green">134</FONT> break;<a name="line.134"></a>
138 <FONT color="green">135</FONT> }<a name="line.135"></a>
139 <FONT color="green">136</FONT> }<a name="line.136"></a>
140 <FONT color="green">137</FONT> }<a name="line.137"></a>
141 <FONT color="green">138</FONT> <a name="line.138"></a>
142 <FONT color="green">139</FONT> return resultSet;<a name="line.139"></a>
143 <FONT color="green">140</FONT> <a name="line.140"></a>
144 <FONT color="green">141</FONT> }<a name="line.141"></a>
145 <FONT color="green">142</FONT> <a name="line.142"></a>
146 <FONT color="green">143</FONT> /**<a name="line.143"></a>
147 <FONT color="green">144</FONT> * Returns the nearest {@link Cluster} to the given point<a name="line.144"></a>
148 <FONT color="green">145</FONT> *<a name="line.145"></a>
149 <FONT color="green">146</FONT> * @param &lt;T&gt; type of the points to cluster<a name="line.146"></a>
150 <FONT color="green">147</FONT> * @param clusters the {@link Cluster}s to search<a name="line.147"></a>
151 <FONT color="green">148</FONT> * @param point the point to find the nearest {@link Cluster} for<a name="line.148"></a>
152 <FONT color="green">149</FONT> * @return the nearest {@link Cluster} to the given point<a name="line.149"></a>
153 <FONT color="green">150</FONT> */<a name="line.150"></a>
154 <FONT color="green">151</FONT> private static &lt;T extends Clusterable&lt;T&gt;&gt; Cluster&lt;T&gt;<a name="line.151"></a>
155 <FONT color="green">152</FONT> getNearestCluster(final Collection&lt;Cluster&lt;T&gt;&gt; clusters, final T point) {<a name="line.152"></a>
156 <FONT color="green">153</FONT> double minDistance = Double.MAX_VALUE;<a name="line.153"></a>
157 <FONT color="green">154</FONT> Cluster&lt;T&gt; minCluster = null;<a name="line.154"></a>
158 <FONT color="green">155</FONT> for (final Cluster&lt;T&gt; c : clusters) {<a name="line.155"></a>
159 <FONT color="green">156</FONT> final double distance = point.distanceFrom(c.getCenter());<a name="line.156"></a>
160 <FONT color="green">157</FONT> if (distance &lt; minDistance) {<a name="line.157"></a>
161 <FONT color="green">158</FONT> minDistance = distance;<a name="line.158"></a>
162 <FONT color="green">159</FONT> minCluster = c;<a name="line.159"></a>
163 <FONT color="green">160</FONT> }<a name="line.160"></a>
164 <FONT color="green">161</FONT> }<a name="line.161"></a>
165 <FONT color="green">162</FONT> return minCluster;<a name="line.162"></a>
166 <FONT color="green">163</FONT> }<a name="line.163"></a>
167 <FONT color="green">164</FONT> <a name="line.164"></a>
168 <FONT color="green">165</FONT> }<a name="line.165"></a>
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229 </PRE>
230 </BODY>
231 </HTML>