Package org.apache.mahout.cf.taste.impl.common

Examples of org.apache.mahout.cf.taste.impl.common.FullRunningAverage

67
68
69
70
71
72
73
74
75
76
77
      PreferenceArray prefs = dataModel.getPreferencesFromUser(key);
      int size = prefs.length();
      if (size == 0) {
        return ZERO;
      }
      RunningAverage average = new FullRunningAverage();
      for (int i = 0; i < size; i++) {
        average.addDatum(prefs.getValue(i));
      }
      return (float) average.getAverage();
    }
View Full Code Here
122
123
124
125
126
127
128
129
130
131
132
133
134
135
    }
    return createFactorization(userVectors, itemVectors);
  }

  double getAveragePreference() throws TasteException {
    RunningAverage average = new FullRunningAverage();
    LongPrimitiveIterator userIDs = dataModel.getUserIDs();
    while (userIDs.hasNext()) {
      for (Preference preference : dataModel.getPreferencesFromUser(userIDs.nextLong())) {
        average.addDatum(preference.getValue());
      }
    }
    return average.getAverage();
  }
View Full Code Here
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
    }
   
    @Override
    public double estimate(Long itemID) throws TasteException {
      DataModel dataModel = getDataModel();
      RunningAverage average = new FullRunningAverage();
      LongPrimitiveIterator it = cluster.iterator();
      while (it.hasNext()) {
        Float pref = dataModel.getPreferenceValue(it.nextLong(), itemID);
        if (pref != null) {
          average.addDatum(pref);
        }
      }
      return average.getAverage();
    }
View Full Code Here
100
101
102
103
104
105
106
107
108
109
110
111
112
113
      userBiases.put(userID, bias);
    }
  }

  private double averageRating() throws TasteException {
    RunningAverage averageRating = new FullRunningAverage();
    LongPrimitiveIterator itemIDs = getDataModel().getItemIDs();
    while (itemIDs.hasNext()) {
      for (Preference pref : getDataModel().getPreferencesForItem(itemIDs.next())) {
        averageRating.addDatum(pref.getValue());
      }
    }
    return averageRating.getAverage();
  }
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
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
   */
  private static Split categoricalSplit(Data data, int attr) {
    FullRunningAverage[] ra = new FullRunningAverage[data.getDataset().nbValues(attr)];
    double[] sk = new double[data.getDataset().nbValues(attr)];
    for (int i = 0; i < ra.length; i++) {
      ra[i] = new FullRunningAverage();
    }
    FullRunningAverage totalRa = new FullRunningAverage();
    double totalSk = 0.0;

    for (int i = 0; i < data.size(); i++) {
      // computes the variance
      Instance instance = data.get(i);
      int value = (int) instance.get(attr);
      double xk = data.getDataset().getLabel(instance);
      if (ra[value].getCount() == 0) {
        ra[value].addDatum(xk);
        sk[value] = 0.0;
      } else {
        double mk = ra[value].getAverage();
        ra[value].addDatum(xk);
        sk[value] += (xk - mk) * (xk - ra[value].getAverage());
      }

      // total variance
      if (i == 0) {
        totalRa.addDatum(xk);
        totalSk = 0.0;
      } else {
        double mk = totalRa.getAverage();
        totalRa.addDatum(xk);
        totalSk += (xk - mk) * (xk - totalRa.getAverage());
      }
    }

    // computes the variance gain
    double ig = totalSk;
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
   * Computes the best split for a NUMERICAL attribute
   */
  private static Split numericalSplit(Data data, int attr) {
    FullRunningAverage[] ra = new FullRunningAverage[2];
    for (int i = 0; i < ra.length; i++) {
      ra[i] = new FullRunningAverage();
    }

    // Instance sort
    Instance[] instances = new Instance[data.size()];
    for (int i = 0; i < data.size(); i++) {
View Full Code Here
169
170
171
172
173
174
175
176
177
178
179
      if (hasMkSk) {
        double mk = Double.parseDouble(tokens[4]);
        double sk = Double.parseDouble(tokens[5]);
        average = new FullRunningAverageAndStdDev(count, diff, mk, sk);
      } else {
        average = new FullRunningAverage(count, diff);
      }
      level1Map.put(itemID2, average);
      averageCount++;
    }
View Full Code Here
89
90
91
92
93
94
95
  // Use RMSE scoring:

  @Override
  protected void reset() {
    average = new FullRunningAverage();
  }
View Full Code Here
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
    this.values = new float[numPrefs];
    this.cachedEstimates = new double[numPrefs];

    index = 0;
    log.info("Loading {} preferences into memory", numPrefs);
    RunningAverage average = new FullRunningAverage();
    for (Preference preference : factorizablePreferences.getPreferences()) {
      userIndexes[index] = userIDMapping.get(preference.getUserID());
      itemIndexes[index] = itemIDMapping.get(preference.getItemID());
      values[index] = preference.getValue();
      cachedEstimates[index] = 0;

      average.addDatum(preference.getValue());

      index++;
      if (index % 1000000 == 0) {
        log.info("Processed {} preferences", index);
      }
    }
    log.info("Processed {} preferences, done.", index);

    double averagePreference = average.getAverage();
    log.info("Average preference value is {}", averagePreference);

    double prefInterval = factorizablePreferences.getMaxPreference() - factorizablePreferences.getMinPreference();
    defaultValue = Math.sqrt((averagePreference - prefInterval * 0.1) / numFeatures);
    interval = prefInterval * 0.1 / numFeatures;
View Full Code Here
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
    info.append(MathHelper.nice(Ak));
    info.append('\n');

    log.info(info.toString());

    RunningAverage avg = new FullRunningAverage();
    for (MatrixSlice slice : preferences) {
      for (Element e : slice.nonZeroes()) {
        if (!Double.isNaN(e.get())) {
          double pref = e.get();
          double estimate = u.viewRow(slice.index()).dot(m.viewRow(e.index()));
          double err = pref - estimate;
          avg.addDatum(err * err);
          log.info("Comparing preference of user [{}] towards item [{}], was [{}] estimate is [{}]",
                   slice.index(), e.index(), pref, estimate);
        }
      }
    }
    double rmse = Math.sqrt(avg.getAverage());
    log.info("RMSE: {}", rmse);

    assertTrue(rmse < 0.2);
  }
View Full Code Here
TOP

Related Classes of org.apache.mahout.cf.taste.impl.common.FullRunningAverage

  • org.apache.mahout.cf.taste.example.kddcup.track1.svd.ParallelArraysSGDFactorizer
  • org.apache.mahout.cf.taste.example.kddcup.track1.svd.Track1SVDRunner
  • org.apache.mahout.cf.taste.example.kddcup.track1.Track1RecommenderEvaluator
  • org.apache.mahout.cf.taste.hadoop.als.FactorizationEvaluator
  • org.apache.mahout.cf.taste.hadoop.als.ParallelALSFactorizationJob$AverageRatingMapper
  • org.apache.mahout.cf.taste.hadoop.als.ParallelALSFactorizationJob$InitializeMReducer
  • org.apache.mahout.cf.taste.hadoop.als.ParallelALSFactorizationJobTest
  • org.apache.mahout.cf.taste.impl.eval.AverageAbsoluteDifferenceRecommenderEvaluator
  • org.apache.mahout.cf.taste.impl.eval.GenericRecommenderIRStatsEvaluator
  • org.apache.mahout.cf.taste.impl.eval.RMSRecommenderEvaluator
    • 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.