Examples of org.apache.mahout.clustering.WeightedVectorWritable

org.apache.mahout.clustering.WeightedVectorWritable

        clusterId = clusters.get(i).getId();
        clusterPdf = pdf;
      }
    }
    // System.out.println("cluster-" + clusterId + ": " + ClusterBase.formatVector(point, null));
    writer.append(new IntWritable(clusterId), new WeightedVectorWritable(clusterPdf, point));
  }

View Full Code Here

            s.getPath(), conf)) {
          MeanShiftCanopy canopy = record.getSecond();
          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

        clusterId = i;
        clusterPdf = pdf;
      }
    }
    //System.out.println(clusterId + ": " + ClusterBase.formatVector(vector.get(), null));
    context.write(new IntWritable(clusterId), new WeightedVectorWritable(clusterPdf, point.get()));
  }

View Full Code Here

    throws IOException, InterruptedException {
    for (int i = 0; i < clusters.size(); i++) {
      double pdf = pi.get(i);
      if (pdf > threshold && clusters.get(i).getTotalCount() > 0) {
        //System.out.println(i + ": " + ClusterBase.formatVector(vector.get(), null));
        context.write(new IntWritable(i), new WeightedVectorWritable(pdf, point.get()));
      }
    }
  }

View Full Code Here

    throws IOException {
    for (int i = 0; i < clusters.size(); i++) {
      double pdf = pi.get(i);
      if (pdf > threshold && clusters.get(i).getTotalCount() > 0) {
        //System.out.println(i + ": " + ClusterBase.formatVector(vector.get(), null));
        writer.append(new IntWritable(i), new WeightedVectorWritable(pdf, vector.get()));
      }
    }
  }

View Full Code Here

        maxPdf = pdf;
        clusterId = i;
      }
    }
    //System.out.println(i + ": " + ClusterBase.formatVector(vector.get(), null));
    writer.append(new IntWritable(clusterId), new WeightedVectorWritable(maxPdf, vector.get()));
  }

View Full Code Here

    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


            List<WeightedVectorWritable> points = clusterIdToPoints.get(cluster.getId());
            if (points != null) {
              writer.write("\tWeight:  Point:\n\t");
              for (Iterator<WeightedVectorWritable> iterator = points.iterator(); iterator.hasNext();) {
                WeightedVectorWritable point = iterator.next();
                writer.write(String.valueOf(point.getWeight()));
                writer.write(": ");
                writer.write(AbstractCluster.formatVector(point.getVector(), dictionary));
                if (iterator.hasNext()) {
                  writer.write("\n\t");
                }
              }
              writer.write('\n');

View Full Code Here

    for (FileStatus file : children) {
      Path path = file.getPath();
      SequenceFile.Reader reader = new SequenceFile.Reader(fs, path, conf);
      try {
        IntWritable key = reader.getKeyClass().asSubclass(IntWritable.class).newInstance();
        WeightedVectorWritable value = reader.getValueClass().asSubclass(WeightedVectorWritable.class).newInstance();
        while (reader.next(key, value)) {
          // value is the cluster id as an int, key is the name/id of the
          // vector, but that doesn't matter because we only care about printing
          // it
          //String clusterId = value.toString();

View Full Code Here

                              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

0 1 2 3 4 5 6 7

TOP

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

com.mozilla.grouperfish.mahout.clustering.display.kmeans.DisplayKMeansBase

com.mozilla.grouperfish.mahout.clustering.display.kmeans.OriginalText

com.mozilla.grouperfish.mahout.clustering.display.kmeans.WordCloud

com.mozilla.grouperfish.transforms.coclustering.display.WriteCoClusteringOutput

com.mozilla.grouperfish.transforms.coclustering.pig.storage.KMeansOutputLoader

mia.clustering.ch07.SimpleKMeansClustering

mia.clustering.ch09.NewsFuzzyKMeansClustering

mia.clustering.ch09.NewsKMeansClustering

org.apache.mahout.clustering.canopy.CanopyClusterer

org.apache.mahout.clustering.canopy.CanopyDriver

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.