Package org.apache.mahout.math.Vector

Examples of org.apache.mahout.math.Vector.Element

71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
      // case one
      Vector updateVector = dataPoint.times(1 / this.promotionStep);
      log.info("Winnow update positive: {}", updateVector);
      Iterator<Element> iter = updateVector.iterateNonZero();
      while (iter.hasNext()) {
        Element element = iter.next();
        model.timesDelta(element.index(), element.get());
      }
    } else {
      // case two
      Vector updateVector = dataPoint.times(1 / this.promotionStep);
      log.info("Winnow update negative: {}", updateVector);
      Iterator<Element> iter = updateVector.iterateNonZero();
      while (iter.hasNext()) {
        Element element = iter.next();
        model.timesDelta(element.index(), element.get());
      }
    }
    log.info(model.toString());
  }
View Full Code Here
73
74
75
76
77
78
79
80
81
82
  }

  public SequentialAccessSparseVector(Vector other) {
    this(other.getName(), other.size(), other.getNumNondefaultElements());
    Iterator<Element> it = other.iterateNonZero();
    Element e;
    while(it.hasNext() && (e = it.next()) != null) {
      set(e.index(), e.get());
    }
  }
View Full Code Here
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
    double result = 0;
    if (x instanceof SequentialAccessSparseVector) {
      // For sparse SeqAccVectors. do dot product without lookup in a linear fashion
      Iterator<Element> myIter = iterateNonZero();
      Iterator<Element> otherIter = x.iterateNonZero();
      Element myCurrent = null;
      Element otherCurrent = null;
      while (myIter.hasNext() && otherIter.hasNext()) {
        if (myCurrent == null) myCurrent = myIter.next();
        if (otherCurrent == null) otherCurrent = otherIter.next();
       
        int myIndex = myCurrent.index();
        int otherIndex = otherCurrent.index();
       
        if (myIndex < otherIndex) {
          // due to the sparseness skipping occurs more hence checked before equality
          myCurrent = null;
        } else if (myIndex > otherIndex){
          otherCurrent = null;
        } else { // both are equal
          result += myCurrent.get() * otherCurrent.get();
          myCurrent = null;
          otherCurrent = null;
        }
      }
      return result;
    } else { // seq.rand. seq.dense
      Iterator<Element> iter = iterateNonZero();
      while (iter.hasNext()) {
        Element element = iter.next();
        result += element.get() * x.getQuick(element.index());
      }
      return result;
    }
  }
View Full Code Here
48
49
50
51
52
53
54
55
56
57
                  Reporter reporter) throws IOException {
    Vector vector = value.get();
    Iterator<Element> it = vector.iterateNonZero();
   
    while (it.hasNext()) {
      Element e = it.next();
      output.collect(new IntWritable(e.index()), ONE);
    }
    output.collect(TOTAL_COUNT, ONE);
  }
View Full Code Here
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
    Vector value = values.next().get();
    Iterator<Element> it = value.iterateNonZero();
    Vector vector = new RandomAccessSparseVector(key.toString(), (int) featureCount, value
        .getNumNondefaultElements());
    while (it.hasNext()) {
      Element e = it.next();
      if (!dictionary.containsKey(e.index())) {
        continue;
      }
      long df = dictionary.get(e.index());
      if (df / vectorCount > maxDfPercent) {
        continue;
      }
      if (df < minDf) {
        df = minDf;
      }
      vector.setQuick(e.index(), tfidf.calculate((int) e.get(), (int) df, (int) featureCount,
        (int) vectorCount));
    }
    if (sequentialAccess) {
      vector = new SequentialAccessSparseVector(vector);
    }
View Full Code Here
43
44
45
46
47
48
49
50
51
  protected void map(WritableComparable<?> key, VectorWritable value,
      Context context) throws IOException, InterruptedException {
    Vector probabilities = classifier.classify(value.get());
    Vector selections = policy.select(probabilities);
    for (Iterator<Element> it = selections.iterateNonZero(); it.hasNext();) {
      Element el = it.next();
      classifier.train(el.index(), value.get(), el.get());
    }
  }
View Full Code Here
68
69
70
71
72
73
74
75
76
77
      for (int j = 0; j < n; j++) {
        accumDots(j, aRow.getQuick(j), yRow);
      }
    } else {
      for (Iterator<Element> iter = aRow.iterateNonZero(); iter.hasNext();) {
        Element el = iter.next();
        accumDots(el.index(), el.get(), yRow);
      }
    }
  }
View Full Code Here
90
91
92
93
94
95
96
97
98
99
      for (int j = 0; j < n; j++) {
        accumDots(j, aRow.getQuick(j), yRowOut);
      }
    } else {
      for (Iterator<Element> iter = aRow.iterateNonZero(); iter.hasNext();) {
        Element el = iter.next();
        accumDots(el.index(), el.get(), yRowOut);
      }
    }
  }
View Full Code Here
38
39
40
41
42
43
44
45
46
47
 
  public double getScoreForLabelInstance(int label, Vector instance) {
    double result = 0.0;
    Iterator<Element> it = instance.iterateNonZero();
    while (it.hasNext()) {
      Element e = it.next();
      result +=  getScoreForLabelFeature(label, e.index());
    }
    return result;
  }
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
62
    Vector vector = value.get();
    int label = key.get();
    double sigmaK = labelSum.get(label);
    Iterator<Element> it = vector.iterateNonZero();
    while (it.hasNext()) {
      Element e = it.next();
      double numerator = featureSum.get(e.index()) - e.get() + alphaI;
      double denominator = totalSum - sigmaK + vocabCount;
      double weight = Math.log(numerator / denominator);
      perLabelThetaNormalizer.set(label, perLabelThetaNormalizer.get(label) + weight);
    }
  }
View Full Code Here
TOP

Related Classes of org.apache.mahout.math.Vector.Element

  • com.digitalpebble.behemoth.mahout.util.Mahout2LibSVM
  • com.mozilla.grouperfish.mahout.clustering.display.kmeans.WordCloud
  • com.mozilla.grouperfish.mahout.clustering.display.lda.DisplayLDABase
  • com.mozilla.grouperfish.mahout.clustering.display.lda.OriginalText
  • com.twitter.elephantbird.pig.mahout.VectorWritableConverter
  • org.apache.mahout.classifier.discriminative.WinnowTrainer
  • org.apache.mahout.classifier.naivebayes.AbstractNaiveBayesClassifier
  • org.apache.mahout.classifier.naivebayes.NaiveBayesTestBase
  • org.apache.mahout.classifier.naivebayes.trainer.NaiveBayesThetaComplementaryMapper
  • org.apache.mahout.clustering.CIMapper
    • 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.