Package org.apache.mahout.clustering

Examples of org.apache.mahout.clustering.Cluster

122
123
124
125
126
127
128
129
130
131
132
133
134
        Writable key = reader.getKeyClass().asSubclass(Writable.class).newInstance();
        Writable value = reader.getValueClass().asSubclass(Writable.class).newInstance();
        Path path = new Path(output, inPart.getName());
        SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, path, IntWritable.class, VectorWritable.class);
        while (reader.next(key, value)) {
          Cluster cluster = (Cluster) value;
          log.debug("C-" + cluster.getId() + ": " + AbstractCluster.formatVector(cluster.getCenter(), null));
          writer.append(new IntWritable(cluster.getId()), new VectorWritable(cluster.getCenter()));
        }
        writer.close();
      }
    }
  }
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
        Path inPart = part.getPath();
        SequenceFile.Reader reader = new SequenceFile.Reader(fs, inPart, conf);
        Writable key = reader.getKeyClass().asSubclass(Writable.class).newInstance();
        Writable value = reader.getValueClass().asSubclass(Writable.class).newInstance();
        while (reader.next(key, value)) {
          Cluster cluster = (Cluster) value;
          clusters.add(cluster);
          value = reader.getValueClass().asSubclass(Writable.class).newInstance();
        }
        reader.close();
      }
View Full Code Here
134
135
136
137
138
139
140
141
142
143
144
145
146
147
  private void pruneInvalidClusters() {
    if (pruned) {
      return;
    }
    for (Iterator<Cluster> it = clusters.iterator(); it.hasNext();) {
      Cluster cluster = it.next();
      if (invalidCluster(cluster)) {
        log.info("Pruning cluster Id=" + cluster.getId());
        it.remove();
        representativePoints.remove(cluster.getId());
      }
    }
    pruned = true;
  }
View Full Code Here
156
157
158
159
160
161
162
163
164
165
166
167
168
169
    double max = 0;
    double min = Double.MAX_VALUE;
    double sum = 0;
    int count = 0;
    for (int i = 0; i < clusters.size(); i++) {
      Cluster clusterI = clusters.get(i);
      for (int j = i + 1; j < clusters.size(); j++) {
        Cluster clusterJ = clusters.get(j);
        double d = measure.distance(clusterI.getCenter(), clusterJ.getCenter());
        min = Math.min(d, min);
        max = Math.max(d, max);
        sum += d;
        count++;
      }
View Full Code Here
67
68
69
70
71
72
73
74
    try {
      cl = ccl.loadClass(klass);
    } catch (ClassNotFoundException e) {
      log.warn("Error while loading class", e);
    }
    Cluster model = (Cluster) gson.fromJson(modelJson, cl);
    return new DirichletCluster(model, total);
  }
View Full Code Here
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
  }

  /** Reads a typed Model instance from the input stream */
  public static Cluster readModel(DataInput in) throws IOException {
    String modelClassName = in.readUTF();
    Cluster model;
    try {
      model = Class.forName(modelClassName).asSubclass(Cluster.class).newInstance();
    } catch (ClassNotFoundException e) {
      throw new IllegalStateException(e);
    } catch (IllegalAccessException e) {
      throw new IllegalStateException(e);
    } catch (InstantiationException e) {
      throw new IllegalStateException(e);
    }
    model.readFields(in);
    return model;
  }
View Full Code Here
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
      SequenceFile.Reader reader = new SequenceFile.Reader(fs, s.getPath(), conf);
      try {
        Writable key = new Text();
        Writable value = reader.getValueClass().asSubclass(Writable.class).newInstance();
        while (reader.next(key, value)) {
          Cluster cluster = (Cluster) value;
          log.info("Reading Cluster:{} center:{} numPoints:{} radius:{}", new Object[] {
              cluster.getId(),
              AbstractCluster.formatVector(cluster.getCenter(), null),
              cluster.getNumPoints(),
              AbstractCluster.formatVector(cluster.getRadius(), null)
          });
          clusters.add(cluster);
          value = reader.getValueClass().asSubclass(Writable.class).newInstance();
        }
      } finally {
View Full Code Here
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
  }

  @Override
  protected void reduce(Text key, Iterable<VectorWritable> values, Context context) throws IOException, InterruptedException {
    int k = Integer.parseInt(key.toString());
    Cluster model = newModels[k];
    for (VectorWritable value : values) {
      // only observe real points, not the empty placeholders emitted by each mapper
      if (value.get().size() > 0) {
        model.observe(value);
      }
    }
    DirichletCluster cluster = clusterer.updateCluster(model, k);
    context.write(new Text(String.valueOf(k)), cluster);
  }
View Full Code Here
45
46
47
48
49
50
51
52
53
54
55
56
57
  }

  @Override
  public void write(ClusterWritable clusterWritable) throws IOException {
    StringBuilder line = new StringBuilder();
    Cluster cluster = clusterWritable.getValue();
    line.append(cluster.getId());
    List<WeightedVectorWritable> points = getClusterIdToPoints().get(cluster.getId());
    if (points != null) {
      for (WeightedVectorWritable point : points) {
        Vector theVec = point.getVector();
        line.append(',');
        if (theVec instanceof NamedVector) {
View Full Code Here
95
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
   */

  @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<WeightedVectorWritable> 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
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.