Package org.apache.mahout.math

Examples of org.apache.mahout.math.DenseVector.assign()

65
66
67
68
69
70
71
72
73
74
75
    transitionMatrix.assign(pseudoCount);
    emissionMatrix.assign(pseudoCount);
    // given no prior knowledge, we have to assume that all initial hidden
    // states are equally likely
    DenseVector initialProbabilities = new DenseVector(nrOfHiddenStates);
    initialProbabilities.assign(1.0 / (double) nrOfHiddenStates);

    // now loop over the sequences to count the number of transitions
    countTransitions(transitionMatrix, emissionMatrix, observedSequence,
        hiddenSequence);
View Full Code Here
148
149
150
151
152
153
154
155
156
157
158
    DenseVector initialProbabilities = new DenseVector(nrOfHiddenStates);

    // assign pseudo count to avoid zero probabilities
    transitionMatrix.assign(pseudoCount);
    emissionMatrix.assign(pseudoCount);
    initialProbabilities.assign(pseudoCount);

    // now loop over the sequences to count the number of transitions
    Iterator<int[]> hiddenSequenceIt = hiddenSequences.iterator();
    Iterator<int[]> observedSequenceIt = observedSequences.iterator();
    while (hiddenSequenceIt.hasNext() && observedSequenceIt.hasNext()) {
View Full Code Here
212
213
214
215
216
217
218
219
220
221
222
      }
      // converged!
      double eigenValue = state.getStatusProgress().get(state.getStatusProgress().size() - 1).getEigenValue();
      // it's actually more efficient to do this to normalize than to call currentEigen = currentEigen.normalize(),
      // because the latter does a clone, which isn't necessary here.
      currentEigen.assign(new TimesFunction(), 1 / currentEigen.norm(2));
      eigens.assignRow(i, currentEigen);
      eigenValues.add(eigenValue);
      state.setCurrentEigenValues(eigenValues);
      log.info("Found eigenvector {}, eigenvalue: {}", i, eigenValue);
View Full Code Here
281
282
283
284
285
286
287
288
289
290
291
    sealed = false;
  }

  private void regularizeAll() {
    Vector all = new DenseVector(beta.numCols());
    all.assign(1);
    regularize(all);
  }

  @Override
  public void close() {
View Full Code Here
86
87
88
89
90
91
92
93
94
95
96
    double docTotal = wordCounts.zSum();
    int docLength = wordCounts.size(); // cardinality of document vectors
   
    // initialize variational approximation to p(z|doc)
    Vector gamma = new DenseVector(state.getNumTopics());
    gamma.assign(state.getTopicSmoothing() + docTotal / state.getNumTopics());
    Vector nextGamma = new DenseVector(state.getNumTopics());
    createPhiMatrix(docLength);
   
    Vector digammaGamma = digammaGamma(gamma);
   
View Full Code Here
99
100
101
102
103
104
105
106
107
108
109
    int iteration = 0;
   
    boolean converged = false;
    double oldLL = 1.0;
    while (!converged && iteration < MAX_ITER) {
      nextGamma.assign(state.getTopicSmoothing()); // nG := alpha, for all topics
     
      int mapping = 0;
      for (Iterator<Vector.Element> iter = wordCounts.iterateNonZero(); iter.hasNext();) {
        Vector.Element e = iter.next();
        int word = e.index();
View Full Code Here
233
234
235
236
237
238
239
240
241
242
243
    return phi;
  }
 
  private static Vector digamma(Vector v) {
    Vector digammaGamma = new DenseVector(v.size());
    digammaGamma.assign(v, new DoubleDoubleFunction() {
      @Override
      public double apply(double unused, double g) {
        return digamma(g);
      }
    });
View Full Code Here
138
139
140
141
142
143
144
145
146
147
148
    }
  }

  private static Vector randomVector(final Random gen, int n) {
    Vector x = new DenseVector(n);
    x.assign(new DoubleFunction() {
      @Override
      public double apply(double v) {
        return gen.nextGaussian();
      }
    });
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
    } else {
      int i = 0;
      for (Cluster model : models) {
        pdfs.set(i++, model.pdf(new VectorWritable(instance)));
      }
      return pdfs.assign(new TimesFunction(), 1.0 / pdfs.zSum());
    }
  }
 
  @Override
  public double classifyScalar(Vector instance) {
View Full Code Here
172
173
174
175
176
177
178
179
180
181
182
    // target variable of 0, the last 30 a target of 1.  The remaining columns are are random noise.
    input = readCsv("sgd.csv");

    // regenerate the target variable
    Vector target = new DenseVector(60);
    target.assign(0);
    target.viewPart(30, 30).assign(1);
    return target;
  }

  private static void train(Matrix input, Vector target, OnlineLearner lr) {
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. 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.