Package org.apache.mahout.clustering.classify

Examples of org.apache.mahout.clustering.classify.WeightedVectorWritable

489
490
491
492
493
494
495
496
497
498
499
        for (Pair<Writable, ClusterWritable> record
            : new SequenceFileIterable<Writable, ClusterWritable>(s.getPath(), conf)) {
          ClusterWritable clusterWritable = record.getSecond();
          MeanShiftCanopy canopy = (MeanShiftCanopy) clusterWritable.getValue();
          MeanShiftCanopy closest = MeanShiftCanopyClusterer.findCoveringCanopy(canopy, clusters);
          writer.append(new IntWritable(closest.getId()), new WeightedVectorWritable(1, canopy.getCenter()));
        }
      } finally {
        Closeables.close(writer, false);
      }
    }
View Full Code Here
47
48
49
50
51
52
53
54
55
56
    int vectorId = canopy.getId();
    for (MeanShiftCanopy msc : canopies) {
      for (int containedId : msc.getBoundPoints().toList()) {
        if (vectorId == containedId) {
          context.write(new IntWritable(msc.getId()),
                         new WeightedVectorWritable(1, canopy.getCenter()));
        }
      }
    }
  }
View Full Code Here
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
    Map<Integer,List<WeightedVectorWritable>> clusterIdToPoints = getClusterIdToPoints();
    List<WeightedVectorWritable> points = clusterIdToPoints.get(clusterWritable.getValue().getId());
    if (points != null) {
      writer.write("\tWeight : [props - optional]:  Point:\n\t");
      for (Iterator<WeightedVectorWritable> iterator = points.iterator(); iterator.hasNext();) {
        WeightedVectorWritable point = iterator.next();
        writer.write(String.valueOf(point.getWeight()));
        if (point instanceof WeightedPropertyVectorWritable) {
          WeightedPropertyVectorWritable tmp = (WeightedPropertyVectorWritable) point;
          Map<Text,Text> map = tmp.getProperties();
          // map can be null since empty maps when written are returned as null
          writer.write(" : [");
          if (map != null) {
            for (Map.Entry<Text,Text> entry : map.entrySet()) {
              writer.write(entry.getKey().toString());
              writer.write("=");
              writer.write(entry.getValue().toString());
            }
          }
          writer.write("]");
        }
       
        writer.write(": ");
       
        writer.write(AbstractCluster.formatVector(point.getVector(), dictionary));
        if (iterator.hasNext()) {
          writer.write("\n\t");
        }
      }
      writer.write('\n');
View Full Code Here
45
46
47
48
49
50
51
52
53
54
55
56
57
  @Override
  protected void reduce(IntWritable key, Iterable<WeightedVectorWritable> values, Context context)
    throws IOException, InterruptedException {
    // find the most distant point
    WeightedVectorWritable mdp = null;
    for (WeightedVectorWritable dpw : values) {
      if (mdp == null || mdp.getWeight() < dpw.getWeight()) {
        mdp = new WeightedVectorWritable(dpw.getWeight(), dpw.getVector());
      }
    }
    context.write(new IntWritable(key.get()), new VectorWritable(mdp.getVector()));
  }
View Full Code Here
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
                              WeightedVectorWritable point,
                              DistanceMeasure measure,
                              Map<Integer, List<VectorWritable>> representativePoints,
                              Map<Integer, WeightedVectorWritable> mostDistantPoints) {
    int key = clusterId.get();
    WeightedVectorWritable currentMDP = mostDistantPoints.get(key);

    List<VectorWritable> repPoints = representativePoints.get(key);
    double totalDistance = 0.0;
    if (repPoints != null) {
      for (VectorWritable refPoint : repPoints) {
        totalDistance += measure.distance(refPoint.get(), point.getVector());
      }
    }
    if (currentMDP == null || currentMDP.getWeight() < totalDistance) {
      mostDistantPoints.put(key, new WeightedVectorWritable(totalDistance, point.getVector().clone()));
    }
  }
View Full Code Here
493
494
495
496
497
498
499
500
501
502
503
          ClusterWritable clusterWritable = record.getSecond();
      MeanShiftCanopy canopy = (MeanShiftCanopy) clusterWritable.getValue();
          MeanShiftCanopy closest = MeanShiftCanopyClusterer
              .findCoveringCanopy(canopy, clusters);
          writer.append(new IntWritable(closest.getId()),
              new WeightedVectorWritable(1, canopy.getCenter()));
        }
      } finally {
        Closeables.closeQuietly(writer);
      }
    }
View Full Code Here
46
47
48
49
50
51
52
53
54
55
    int vectorId = canopy.getId();
    for (MeanShiftCanopy msc : canopies) {
      for (int containedId : msc.getBoundPoints().toList()) {
        if (vectorId == containedId) {
          context.write(new IntWritable(msc.getId()),
                         new WeightedVectorWritable(1, canopy.getCenter()));
        }
      }
    }
  }
View Full Code Here
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
                              WeightedVectorWritable point,
                              DistanceMeasure measure,
                              Map<Integer, List<VectorWritable>> representativePoints,
                              Map<Integer, WeightedVectorWritable> mostDistantPoints) {
    int key = clusterId.get();
    WeightedVectorWritable currentMDP = mostDistantPoints.get(key);

    List<VectorWritable> repPoints = representativePoints.get(key);
    double totalDistance = 0.0;
    for (VectorWritable refPoint : repPoints) {
      totalDistance += measure.distance(refPoint.get(), point.getVector());
    }
    if (currentMDP == null || currentMDP.getWeight() < totalDistance) {
      mostDistantPoints.put(key, new WeightedVectorWritable(totalDistance, point.getVector().clone()));
    }
  }
View Full Code Here
TOP

Related Classes of org.apache.mahout.clustering.classify.WeightedVectorWritable

  • org.apache.mahout.clustering.evaluation.RepresentativePointsMapper
  • org.apache.mahout.clustering.evaluation.RepresentativePointsReducer
  • org.apache.mahout.clustering.meanshift.MeanShiftCanopyClusterMapper
  • org.apache.mahout.clustering.meanshift.MeanShiftCanopyDriver
  • org.apache.mahout.math.Vector
  • org.apache.mahout.utils.clustering.ClusterDumperWriter
    • 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.