Package edu.uci.jforestsx.input.sparse

Examples of edu.uci.jforestsx.input.sparse.SparseTextFileReader

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
    valueHashMaps = new ArrayList<HashMap<Integer, Integer>>(featureCount);
    for (int f = 0; f < featureCount; f++) {
      valueHashMaps.add(new HashMap<Integer, Integer>());
    }

    SparseTextFileReader reader = new SparseTextFileReader();
    reader.open(textFile);
    SparseTextFileLine line = new SparseTextFileLine();
    HashMap<Integer, Integer> curMap;
    int key;
    instanceCount = 0;
    while (reader.loadNextLine(line)) {
      if (line.meta) {
        continue;
      }
      for (int i = 0; i < line.numPairs; i++) {
        FeatureValuePair pair = line.pairs[i];
        curMap = valueHashMaps.get(pair.featureIndex - 1);
        key = (int) pair.featureValue;
        Integer count = curMap.get(key);
        if (count == null) {
          count = 0;
        }
        count++;
        curMap.put(key, count);
      }
      handle(line);
      instanceCount++;
    }
    reader.close();
    System.out.println("  [Done in: " + timer.getElapsedSeconds() + " seconds.]");
  }
View Full Code Here
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
    }

    targets = new double[instanceCount];
    int[] zeroCount = new int[featureCount];

    SparseTextFileReader reader = new SparseTextFileReader();
    reader.open(textFile);
    SparseTextFileLine line = new SparseTextFileLine();
    int instanceIdx = 0;
    while (reader.loadNextLine(line)) {
      if (line.meta) {
        continue;
      }
      targets[instanceIdx] = line.target;
      for (int i = 0; i < line.numPairs; i++) {
        FeatureValuePair pair = line.pairs[i];
        int fidx = pair.featureIndex - 1;
        int index = Arrays.binarySearch(valueDistributions[fidx], (int) pair.featureValue);
        bins[fidx].set(instanceIdx, index);
        if (index == 0) {
          zeroCount[fidx]++;
        }
      }
      instanceIdx++;
    }
    reader.close();
    System.out.println("  [Done in: " + timer.getElapsedSeconds() + " seconds.]");
  }
View Full Code Here
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
*/

public class Sparse2SvmConvertor {

  public static void convert(String inputFile, String outputFile) throws Exception {
    SparseTextFileReader reader = new SparseTextFileReader();
    reader.open(inputFile);
    SparseTextFileLine line = new SparseTextFileLine();
    int count = 0;
    PrintStream output = new PrintStream(new File(outputFile));
    StringBuilder sb = new StringBuilder();
    int featureCount = 0;
    while (reader.loadNextLine(line)) {
      if (line.meta) {
        MetaData metaData = MetaLineParser.parse(line.content);
        if (metaData instanceof FeatureMetaData) {
          featureCount++;
        }
      } else {
        sb.setLength(0);
        sb.append(line.target);
        for (int i = 0; i < line.numPairs; i++) {
          FeatureValuePair pair = line.pairs[i];
          sb.append(" " + pair.featureIndex + ":" + pair.featureValue);
        }       
        sb.append("\n");
        output.print(sb.toString());
        count++;
        if (count % 10000 == 0) {
          System.out.println("\t Processed: " + count);
        }
      }
    }
    reader.close();
  }
View Full Code Here
43
44
45
46
47
48
49
50
51
52
53
  private int featureCount;
  private StringBuilder sb;
  private int targetColIdx;
 
  public void init(String inputFile, String outputFile, boolean dumpAsSparse) throws Exception {
    reader = new SparseTextFileReader();
    reader.open(inputFile);
    this.dumpAsSparse = dumpAsSparse;
   
    output = new PrintStream(new File(outputFile));
    output.println("@RELATION myrel");
View Full Code Here
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
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
    FeatureAnalyzer featureAnalyzer = new FeatureAnalyzer();
    featureAnalyzer.loadFeaturesFromFile(featuresStatFile);

    StringBuilder sb = new StringBuilder();
    double value;
    SparseTextFileReader reader = new SparseTextFileReader();
    SparseTextFileLine line = new SparseTextFileLine();
    reader.open(inputFilename);
    int intValue;
    try {
      PrintStream output = new PrintStream(new File(outputFilename));
      int count = 0;
      while (reader.loadNextLine(line)) {
        if (line.meta) {
          output.println(line.content);
          continue;
        }
        sb.setLength(0);
        sb.append(line.target);
        if (line.qid != null) {
          sb.append(" qid:" + line.qid);
        }


        // System.out.println("Converting line with qid:" + line.qid); // SISTA
        for (int i = 0; i < line.numPairs; i++) {
          FeatureValuePair pair = line.pairs[i];
          // System.out.println("\t" + pair.featureIndex + ":" + pair.featureValue); // SISTA
          value = pair.featureValue;
          int idx = pair.featureIndex - 1;
          if (featureAnalyzer.onLogScale[idx]) {
            value = (Math.log(value - featureAnalyzer.min[idx] + 1) * featureAnalyzer.factor[idx]);
          } else {
            value = (value - featureAnalyzer.min[idx]) * featureAnalyzer.factor[idx];
          }
          intValue = (int) Math.round(value);
          // System.out.println("\t\tdiscrete value: " + intValue); // SISTA
          if (intValue != 0) {
            sb.append(" " + pair.featureIndex + ":" + intValue);
          }
        }
        output.println(sb.toString());
        count++;
        if (count % 10000 == 0) {
          System.out.println(count);
        }
      }
      output.close();
    } catch (Exception e) {
      e.printStackTrace();
    }
    reader.close();
  }
View Full Code Here
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
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
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
    }
  }

  public void processFile(String inputFile) {
    System.out.println("Processing: " + inputFile);
    SparseTextFileReader reader = new SparseTextFileReader();
    reader.open(inputFile);
    SparseTextFileLine line = new SparseTextFileLine();
    FeatureStatistics stat;
    int count = 0;
    int maxFeatureIndex = 0;
    while (reader.loadNextLine(line)) {
      if (line.meta) {
        MetaData metaData = MetaLineParser.parse(line.content);
        if (metaData instanceof FeatureMetaData) {
          fid2name.put(((FeatureMetaData) metaData).id, ((FeatureMetaData) metaData).name);
        }
      } else {
        int prevIdx = 0;
        for (int i = 0; i < line.numPairs; i++) {
          FeatureValuePair pair = line.pairs[i];
          if (pair.featureIndex != (prevIdx + 1)) {
            for (int f = prevIdx + 1; f < pair.featureIndex; f++) {
              stat = fid2statistics.get(f);
              if (stat != null) {
                if (stat.maxValue < 0) {
                  stat.maxValue = 0;
                }
                if (stat.minValue > 0) {
                  stat.minValue = 0;
                }
              }
            }
          }
          stat = fid2statistics.get(pair.featureIndex);
          if (stat == null) {
            stat = new FeatureStatistics();
            fid2statistics.put(pair.featureIndex, stat);
            if (count > 0) {
              stat.minValue = 0;
              stat.maxValue = 0;
            }
            if (pair.featureIndex > maxFeatureIndex) {
              maxFeatureIndex = pair.featureIndex;
            }
          }
          if (Double.isInfinite(pair.featureValue)) {
            System.out.println(count + "\t" + pair.featureValue);
          }
          if (pair.featureValue > stat.maxValue) {
            stat.maxValue = pair.featureValue;
          }
          if (pair.featureValue < stat.minValue) {
            stat.minValue = pair.featureValue;
          }         
          prevIdx = pair.featureIndex;
        }
        if (prevIdx < maxFeatureIndex) {
          for (int f = prevIdx + 1; f <= maxFeatureIndex; f++) {
            stat = fid2statistics.get(f);
            if (stat != null) {
              if (stat.maxValue < 0) {
                stat.maxValue = 0;
              }
              if (stat.minValue > 0) {
                stat.minValue = 0;
              }
            }
          }
        }
        count++;
        if (count % 100000 == 0) {
          System.out.println("\t Processed: " + count);
          dumpStatistics(System.out);
        }
      }
    }
    reader.close();
    loadStatistics();   
  }
View Full Code Here
TOP

Related Classes of edu.uci.jforestsx.input.sparse.SparseTextFileReader

  • edu.uci.jforestsx.input.BinaryFileGenerator
  • edu.uci.jforestsx.input.DiscreteSparseTextFileGenerator
  • edu.uci.jforestsx.input.FeatureAnalyzer
  • edu.uci.jforestsx.tools.Sparse2ArffConvertor
  • edu.uci.jforestsx.tools.Sparse2SvmConvertor
  • java.io.File
  • java.io.FileReader
  • java.io.BufferedReader
  • java.util.StringTokenizer
    • 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.