Package org.apache.mahout.clustering

Examples of org.apache.mahout.clustering.Cluster

96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
   */

  @Override
  public void write(ClusterWritable clusterWritable) throws IOException {
    StringBuilder line = new StringBuilder();
    Cluster cluster = clusterWritable.getValue();
    Color rgb = getColor(cluster.getId());

    String topTerms = "";
    if (dictionary != null) {
      topTerms = getTopTerms(cluster.getCenter(), dictionary, numTopFeatures);
    }
    String clusterLabel = String.valueOf(cluster.getId()) + '_' + topTerms;
    //do some positioning so that items are visible and grouped together
    //TODO: put in a real layout algorithm
    float x = lastX + 1000;
    float y = lastY;
    if (x > (1000 + posStep)) {
      y = lastY + 1000;
      x = 0;
    }

    line.append(createNode(clusterLabel, rgb, x, y));
    List<WeightedPropertyVectorWritable> points = clusterIdToPoints.get(cluster.getId());
    if (points != null) {
      for (WeightedVectorWritable point : points) {
        Vector theVec = point.getVector();
        double distance = 1;
        if (measure != null) {
          //scale the distance
          distance = measure.distance(cluster.getCenter().getLengthSquared(), cluster.getCenter(), theVec) * 500;
        }
        String vecStr;
        int angle = random.nextInt(360); //pick an angle at random and then scale along that angle
        double angleRads = Math.toRadians(angle);

View Full Code Here
163
164
165
166
167
168
169
170
171
172
173
    Path finalClustersPath = finalClustersPath(conf, clusterOutputPath);
    Iterator<?> it = new SequenceFileDirValueIterator<Writable>(finalClustersPath, PathType.LIST,
        PathFilters.partFilter(), null, false, conf);
    while (it.hasNext()) {
      ClusterWritable next = (ClusterWritable) it.next();
      Cluster cluster = next.getValue();
      cluster.configure(conf);
      clusterModels.add(cluster);
    }
    return clusterModels;
  }
View Full Code Here
249
250
251
252
253
254
255
256
257
258
259
260
261
262
  }

  private static void write(List<Cluster> clusterModels, SequenceFile.Writer writer,
      WeightedPropertyVectorWritable weightedPropertyVectorWritable,
      int maxValueIndex) throws IOException {
    Cluster cluster = clusterModels.get(maxValueIndex);

    DistanceMeasureCluster distanceMeasureCluster = (DistanceMeasureCluster) cluster;
    DistanceMeasure distanceMeasure = distanceMeasureCluster.getMeasure();
    double distance = distanceMeasure.distance(cluster.getCenter(), weightedPropertyVectorWritable.getVector());

    weightedPropertyVectorWritable.getProperties().put(new Text("distance"), new Text(Double.toString(distance)));
    writer.append(new IntWritable(cluster.getId()), weightedPropertyVectorWritable);
  }
View Full Code Here
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
    }
  }
 
  private void write(VectorWritable vw, Context context, int clusterIndex, double weight)
    throws IOException, InterruptedException {
    Cluster cluster = clusterModels.get(clusterIndex);
    clusterId.set(cluster.getId());

    DistanceMeasureCluster distanceMeasureCluster = (DistanceMeasureCluster) cluster;
    DistanceMeasure distanceMeasure = distanceMeasureCluster.getMeasure();
    double distance = distanceMeasure.distance(cluster.getCenter(), vw.get());

    Map<Text, Text> props = Maps.newHashMap();
    props.put(new Text("distance"), new Text(Double.toString(distance)));
    context.write(clusterId, new WeightedPropertyVectorWritable(weight, vw.get(), props));
  }
View Full Code Here
140
141
142
143
144
145
146
147
148
149
150
    Iterator<?> it = new SequenceFileDirValueIterator<Writable>(
        clusterFiles[0].getPath(), PathType.LIST, PathFilters.partFilter(),
        null, false, conf);
    while (it.hasNext()) {
      ClusterWritable next = (ClusterWritable) it.next();
      Cluster cluster = next.getValue();
      cluster.configure(conf);
      clusters.add(cluster);
    }
    return clusters;
  }
View Full Code Here
332
333
334
335
336
337
338
339
340
341
342
343
344
345
  protected static List<Cluster> readClustersWritable(Path clustersIn) {
    List<Cluster> clusters = Lists.newArrayList();
    Configuration conf = new Configuration();
    for (ClusterWritable value : new SequenceFileDirValueIterable<ClusterWritable>(clustersIn, PathType.LIST,
        PathFilters.logsCRCFilter(), conf)) {
      Cluster cluster = value.getValue();
      log.info(
          "Reading Cluster:{} center:{} numPoints:{} radius:{}",
          cluster.getId(), AbstractCluster.formatVector(cluster.getCenter(), null),
          cluster.getNumObservations(), AbstractCluster.formatVector(cluster.getRadius(), null));
      clusters.add(cluster);
    }
    return clusters;
  }
View Full Code Here
47
48
49
50
51
52
53
54
55
56
57
58
    this.subString = subString;
  }
 
  @Override
  public void write(ClusterWritable clusterWritable) throws IOException {
    Cluster cluster = clusterWritable.getValue();
    String fmtStr = cluster.asFormatString(dictionary);
    Writer writer = getWriter();
    if (subString > 0 && fmtStr.length() > subString) {
      writer.write(':');
      writer.write(fmtStr, 0, Math.min(subString, fmtStr.length()));
    } else {
View Full Code Here
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
    return buf.toString();
  }
 
  private void printClusters(List<Cluster> models, String[] docs) {
    for (int m = 0; m < models.size(); m++) {
      Cluster model = models.get(m);
      long total = model.getTotalObservations();
      if (total == 0) {
        continue;
      }
      System.out.println();
      System.out.println("Model[" + m + "] had " + total + " observations");
      System.out.println("pdf           document");
      MapElement[] map = new MapElement[sampleData.size()];
      // sort the samples by pdf
      double maxPdf = Double.MIN_NORMAL;
      for (int i = 0; i < sampleData.size(); i++) {
        VectorWritable sample = new VectorWritable(sampleData.get(i));
        double pdf = Math.abs(model.pdf(sample));
        maxPdf = Math.max(maxPdf, pdf);
        map[i] = new MapElement(pdf, docs[i]);
      }
      Arrays.sort(map);
      for (int i = 0; i < map.length; i++) {
View Full Code Here
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
    int clusterCount = 0;
    Collection<Integer> set = new HashSet<Integer>();
    for (ClusterWritable clusterWritable :
         new SequenceFileValueIterable<ClusterWritable>(new Path(output, "part-randomSeed"), true, conf)) {
      clusterCount++;
      Cluster cluster = clusterWritable.getValue();
      int id = cluster.getId();
      assertTrue(set.add(id)); // validate unique id's
     
      Vector v = cluster.getCenter();
      assertVectorEquals(RAW[id], v); // validate values match
    }

    assertEquals(4, clusterCount); // validate sample count
  }
View Full Code Here
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
   * @return The list of clusters found by the clustering.
   * @throws IOException
   */
  private static List<Cluster> populateClusterModels(Path clusterOutputPath, Configuration conf) throws IOException {
    List<Cluster> clusterModels = new ArrayList<Cluster>();
    Cluster cluster = null;
    Path finalClustersPath = finalClustersPath(conf, clusterOutputPath);
    Iterator<?> it = new SequenceFileDirValueIterator<Writable>(finalClustersPath, PathType.LIST,
        PathFilters.partFilter(), null, false, conf);
    while (it.hasNext()) {
      ClusterWritable next = (ClusterWritable) it.next();
      cluster = (Cluster) next.getValue();
      cluster.configure(conf);
      clusterModels.add(cluster);
    }
    return clusterModels;
  }
View Full Code Here
TOP

Related Classes of org.apache.mahout.clustering.Cluster

  • com.ikanow.infinit.e.processing.custom.utils.HadoopUtils
  • mia.clustering.ch10.InterClusterDistances
  • org.apache.mahout.clustering.cdbw.CDbwEvaluator
  • org.apache.mahout.clustering.classify.ClusterClassificationDriver
  • org.apache.mahout.clustering.classify.ClusterClassificationMapper
  • org.apache.mahout.clustering.classify.ClusterClassifier
  • org.apache.mahout.clustering.dirichlet.DirichletCluster
  • org.apache.mahout.clustering.dirichlet.DirichletReducer
  • org.apache.mahout.clustering.dirichlet.JsonClusterAdapter
  • org.apache.mahout.clustering.dirichlet.TestL1ModelClustering
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.