Package org.apache.mahout.math.neighborhood

Examples of org.apache.mahout.math.neighborhood.BruteSearch

90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
   * @param distanceMeasure the distance measure used to compute the distance between two points.
   * @return the minimum distance between the first sampleLimit points
   * @see org.apache.mahout.clustering.streaming.cluster.StreamingKMeans#clusterInternal(Iterable, boolean)
   */
  public static double estimateDistanceCutoff(List<? extends Vector> data, DistanceMeasure distanceMeasure) {
    BruteSearch searcher = new BruteSearch(distanceMeasure);
    searcher.addAll(data);
    double minDistance = Double.POSITIVE_INFINITY;
    for (Vector vector : data) {
      double closest = searcher.searchFirst(vector, true).getWeight();
      if (minDistance > 0 && closest < minDistance) {
        minDistance = closest;
      }
      searcher.add(vector);
    }
    return minDistance;
  }
View Full Code Here
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
   * @param distanceMeasure distance measure to use
   * @return the confusion matrix
   */
  public static Matrix getConfusionMatrix(List<? extends Vector> rowCentroids, List<? extends  Vector> columnCentroids,
                                          Iterable<? extends Vector> datapoints, DistanceMeasure distanceMeasure) {
    Searcher rowSearcher = new BruteSearch(distanceMeasure);
    rowSearcher.addAll(rowCentroids);
    Searcher columnSearcher = new BruteSearch(distanceMeasure);
    columnSearcher.addAll(columnCentroids);

    int numRows = rowCentroids.size();
    int numCols = columnCentroids.size();
    Matrix confusionMatrix = new DenseMatrix(numRows, numCols);

    for (Vector vector : datapoints) {
      WeightedThing<Vector> closestRowCentroid = rowSearcher.search(vector, 1).get(0);
      WeightedThing<Vector> closestColumnCentroid = columnSearcher.search(vector, 1).get(0);
      int row = ((Centroid) closestRowCentroid.getValue()).getIndex();
      int column = ((Centroid) closestColumnCentroid.getValue()).getIndex();
      double vectorWeight;
      if (vector instanceof WeightedVector) {
        vectorWeight = ((WeightedVector) vector).getWeight();
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
        mapDriver.getConfiguration().get(StreamingKMeansDriver.SEARCHER_CLASS_OPTION));
    for (Centroid datapoint : syntheticData.getFirst()) {
      mapDriver.addInput(new IntWritable(0), new VectorWritable(datapoint));
    }
    List<org.apache.hadoop.mrunit.types.Pair<IntWritable,CentroidWritable>> results = mapDriver.run();
    BruteSearch resultSearcher = new BruteSearch(new SquaredEuclideanDistanceMeasure());
    for (org.apache.hadoop.mrunit.types.Pair<IntWritable, CentroidWritable> result : results) {
      resultSearcher.add(result.getSecond().getCentroid());
    }
    System.out.printf("Clustered the data into %d clusters\n", results.size());
    for (Vector mean : syntheticData.getSecond()) {
      WeightedThing<Vector> closest = resultSearcher.search(mean, 1).get(0);
      assertTrue("Weight " + closest.getWeight() + " not less than 0.5", closest.getWeight() < 0.5);
    }
  }
View Full Code Here
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
  private static final int K1 = 100;

  @Test
  public void testClusteringMultipleRuns() {
    for (int i = 1; i <= 10; ++i) {
      BallKMeans clusterer = new BallKMeans(new BruteSearch(new SquaredEuclideanDistanceMeasure()),
          1 << NUM_DIMENSIONS, NUM_ITERATIONS, true, i);
      clusterer.cluster(syntheticData.getFirst());
      double costKMeansPlusPlus = ClusteringUtils.totalClusterCost(syntheticData.getFirst(), clusterer);

      clusterer = new BallKMeans(new BruteSearch(new SquaredEuclideanDistanceMeasure()),
          1 << NUM_DIMENSIONS, NUM_ITERATIONS, false, i);
      clusterer.cluster(syntheticData.getFirst());
      double costKMeansRandom = ClusteringUtils.totalClusterCost(syntheticData.getFirst(), clusterer);

      System.out.printf("%d runs; kmeans++: %f; random: %f\n", i, costKMeansPlusPlus, costKMeansRandom);
View Full Code Here
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
102
103
104
105
106
107
108
109
110
111
112
113
114
115
    }
  }

  @Test
  public void testClustering() {
    UpdatableSearcher searcher = new BruteSearch(new SquaredEuclideanDistanceMeasure());
    BallKMeans clusterer = new BallKMeans(searcher, 1 << NUM_DIMENSIONS, NUM_ITERATIONS);

    long startTime = System.currentTimeMillis();
    clusterer.cluster(syntheticData.getFirst());
    long endTime = System.currentTimeMillis();

    assertEquals("Total weight not preserved", totalWeight(syntheticData.getFirst()), totalWeight(clusterer), 1.0e-9);

    // Verify that each corner of the cube has a centroid very nearby.
    // This is probably FALSE for large-dimensional spaces!
    OnlineSummarizer summarizer = new OnlineSummarizer();
    for (Vector mean : syntheticData.getSecond()) {
      WeightedThing<Vector> v = searcher.search(mean, 1).get(0);
      summarizer.add(v.getWeight());
    }
    assertTrue(String.format("Median weight [%f] too large [>%f]", summarizer.getMedian(),
        DISTRIBUTION_RADIUS), summarizer.getMedian() < DISTRIBUTION_RADIUS);

    double clusterTime = (endTime - startTime) / 1000.0;
    System.out.printf("%s\n%.2f for clustering\n%.1f us per row\n\n",
        searcher.getClass().getName(), clusterTime,
        clusterTime / syntheticData.getFirst().size() * 1.0e6);

    // Verify that the total weight of the centroids near each corner is correct.
    double[] cornerWeights = new double[1 << NUM_DIMENSIONS];
    Searcher trueFinder = new BruteSearch(new EuclideanDistanceMeasure());
    for (Vector trueCluster : syntheticData.getSecond()) {
      trueFinder.add(trueCluster);
    }
    for (Centroid centroid : clusterer) {
      WeightedThing<Vector> closest = trueFinder.search(centroid, 1).get(0);
      cornerWeights[((Centroid)closest.getValue()).getIndex()] += centroid.getWeight();
    }
    int expectedNumPoints = NUM_DATA_POINTS / (1 << NUM_DIMENSIONS);
    for (double v : cornerWeights) {
      System.out.printf("%f ", v);
View Full Code Here
124
125
126
127
128
129
130
131
132
133
134
  public void testInitialization() {
    // Start with super clusterable data.
    List<? extends WeightedVector> data = cubishTestData(0.01);

    // Just do initialization of ball k-means. This should drop a point into each of the clusters.
    BallKMeans r = new BallKMeans(new BruteSearch(new SquaredEuclideanDistanceMeasure()), 6, 20);
    r.cluster(data);

    // Put the centroids into a matrix.
    Matrix x = new DenseMatrix(6, 5);
    int row = 0;
View Full Code Here
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
        searcher.getClass().getName(), clusterTime,
        clusterTime / syntheticData.getFirst().size() * 1.0e6);

    // verify that the total weight of the centroids near each corner is correct
    double[] cornerWeights = new double[1 << NUM_DIMENSIONS];
    Searcher trueFinder = new BruteSearch(new EuclideanDistanceMeasure());
    for (Vector trueCluster : syntheticData.getSecond()) {
      trueFinder.add(trueCluster);
    }
    for (Centroid centroid : clusterer) {
      WeightedThing<Vector> closest = trueFinder.search(centroid, 1).get(0);
      cornerWeights[((Centroid)closest.getValue()).getIndex()] += centroid.getWeight();
    }
    int expectedNumPoints = NUM_DATA_POINTS / (1 << NUM_DIMENSIONS);
    for (double v : cornerWeights) {
      System.out.printf("%f ", v);
View Full Code Here
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
   * @param distanceMeasure the distance measure used to compute the distance between two points.
   * @return the minimum distance between the first sampleLimit points
   * @see org.apache.mahout.clustering.streaming.cluster.StreamingKMeans#clusterInternal(Iterable, boolean)
   */
  public static double estimateDistanceCutoff(List<? extends Vector> data, DistanceMeasure distanceMeasure) {
    BruteSearch searcher = new BruteSearch(distanceMeasure);
    searcher.addAll(data);
    double minDistance = Double.POSITIVE_INFINITY;
    for (Vector vector : data) {
      double closest = searcher.searchFirst(vector, true).getWeight();
      if (minDistance > 0 && closest < minDistance) {
        minDistance = closest;
      }
      searcher.add(vector);
    }
    return minDistance;
  }
View Full Code Here
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
   * @param distanceMeasure distance measure to use
   * @return the confusion matrix
   */
  public static Matrix getConfusionMatrix(List<? extends Vector> rowCentroids, List<? extends  Vector> columnCentroids,
                                          Iterable<? extends Vector> datapoints, DistanceMeasure distanceMeasure) {
    Searcher rowSearcher = new BruteSearch(distanceMeasure);
    rowSearcher.addAll(rowCentroids);
    Searcher columnSearcher = new BruteSearch(distanceMeasure);
    columnSearcher.addAll(columnCentroids);

    int numRows = rowCentroids.size();
    int numCols = columnCentroids.size();
    Matrix confusionMatrix = new DenseMatrix(numRows, numCols);

    for (Vector vector : datapoints) {
      WeightedThing<Vector> closestRowCentroid = rowSearcher.search(vector, 1).get(0);
      WeightedThing<Vector> closestColumnCentroid = columnSearcher.search(vector, 1).get(0);
      int row = ((Centroid) closestRowCentroid.getValue()).getIndex();
      int column = ((Centroid) closestColumnCentroid.getValue()).getIndex();
      double vectorWeight;
      if (vector instanceof WeightedVector) {
        vectorWeight = ((WeightedVector) vector).getWeight();
View Full Code Here
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82

  @Test
  public void testClusteringMultipleRuns() {
    for (int i = 1; i <= 10; ++i) {
      BallKMeans clusterer = new BallKMeans(new BruteSearch(new SquaredEuclideanDistanceMeasure()),
          1 << NUM_DIMENSIONS, NUM_ITERATIONS, true, i);
      clusterer.cluster(syntheticData.getFirst());
      double costKMeansPlusPlus = ClusteringUtils.totalClusterCost(syntheticData.getFirst(), clusterer);

      clusterer = new BallKMeans(new BruteSearch(new SquaredEuclideanDistanceMeasure()),
          1 << NUM_DIMENSIONS, NUM_ITERATIONS, false, i);
      clusterer.cluster(syntheticData.getFirst());
      double costKMeansRandom = ClusteringUtils.totalClusterCost(syntheticData.getFirst(), clusterer);

      System.out.printf("%d runs; kmeans++: %f; random: %f\n", i, costKMeansPlusPlus, costKMeansRandom);
View Full Code Here
TOP

Related Classes of org.apache.mahout.math.neighborhood.BruteSearch

  • org.apache.mahout.clustering.ClusteringUtils
  • org.apache.mahout.clustering.streaming.cluster.BallKMeansTest
  • org.apache.mahout.clustering.streaming.cluster.StreamingKMeansTest
  • org.apache.mahout.clustering.streaming.mapreduce.StreamingKMeansTestMR
  • org.apache.mahout.math.Vector
  • java.util.concurrent.ExecutorService
    • 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.