Examples of StringRecordIterator

Examples of org.apache.mahout.common.StringRecordIterator

189
190
191
192
193
194
195
196
197
198
199
200
    SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, path, Text.class, TopKStringPatterns.class);
   
    FPGrowth<String> fp = new FPGrowth<String>();
    Set<String> features = new HashSet<String>();
   
    fp.generateTopKFrequentPatterns(new StringRecordIterator(new FileLineIterable(new File(input), encoding,
        false), pattern), fp.generateFList(new StringRecordIterator(new FileLineIterable(new File(input),
        encoding, false), pattern), minSupport), minSupport, maxHeapSize, features,
      new StringOutputConverter(new SequenceFileOutputCollector<Text,TopKStringPatterns>(writer)),
      new ContextStatusUpdater(null));
    writer.close();
   
View Full Code Here

Examples of org.apache.mahout.common.StringRecordIterator

190
191
192
193
194
195
196
197
198
199
200
201
202
   
    FPGrowth<String> fp = new FPGrowth<String>();
    Set<String> features = new HashSet<String>();
   
    fp.generateTopKFrequentPatterns(
        new StringRecordIterator(new FileLineIterable(new File(input), encoding, false), pattern),
        fp.generateFList(
            new StringRecordIterator(new FileLineIterable(new File(input), encoding, false), pattern),
            minSupport),
        minSupport,
        maxHeapSize,
        features,
        new StringOutputConverter(new SequenceFileOutputCollector<Text,TopKStringPatterns>(writer)),
View Full Code Here

Examples of org.apache.mahout.common.StringRecordIterator

218
219
220
221
222
223
224
225
226
227
228
229
230
        Text.class, TopKStringPatterns.class);

    FPGrowth<String> fp = new FPGrowth<String>();
    Set<String> features = new HashSet<String>();

    fp.generateTopKFrequentPatterns(new StringRecordIterator(
        new FileLineIterable(new File(input), encoding, false), pattern), fp
        .generateFList(new StringRecordIterator(new FileLineIterable(new File(
            input), encoding, false), pattern), minSupport), minSupport,
        maxHeapSize, features, new StringOutputConvertor(
            new SequenceFileOutputCollector<Text, TopKStringPatterns>(writer)));
    writer.close();
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

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
      FPGrowth<String> fp1 = new org.apache.mahout.fpm.pfpgrowth.fpgrowth.FPGrowth<String>();

    final Map<Set<String>,Long> results1 = Maps.newHashMap();
   
    fp1.generateTopKFrequentPatterns(
      new StringRecordIterator(new FileLineIterable(Resources.getResource(inputFilename).openStream()), "\\s+"),

      fp1.generateFList(new StringRecordIterator(new FileLineIterable(Resources.getResource(inputFilename)
           .openStream()), "\\s+"), minSupport), minSupport, 100000,
      returnableFeatures,
      new OutputCollector<String,List<Pair<List<String>,Long>>>() {
       
        @Override
        public void collect(String key, List<Pair<List<String>,Long>> value) {
         
          for (Pair<List<String>,Long> v : value) {
            List<String> l = v.getFirst();
            results1.put(new HashSet<String>(l), v.getSecond());
            log.info("found pat ["+v.getSecond()+"]: "+ v.getFirst());
          }
        }
       
      }, new StatusUpdater() {
       
        @Override
        public void update(String status) {}
      });

    FPGrowthObj<String> fp2 = new FPGrowthObj<String>();
    final Map<Set<String>,Long> initialResults2 = Maps.newHashMap();
    fp2.generateTopKFrequentPatterns(
      new StringRecordIterator(new FileLineIterable(Resources.getResource(inputFilename).openStream()), "\\s+"),

      fp2.generateFList(new StringRecordIterator(new FileLineIterable(Resources.getResource(inputFilename)
           .openStream()), "\\s+"), minSupport), minSupport, 100000,
      new HashSet<String>(),
      new OutputCollector<String,List<Pair<List<String>,Long>>>() {
       
        @Override
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
    File outputDir2 = getTestTempDir("frequentpatterns2");
    paramsImpl2.set(PFPGrowth.OUTPUT, outputDir2.getAbsolutePath());

    Writer writer = Files.newWriter(input, Charsets.UTF_8);
    try {
      StringRecordIterator it = new StringRecordIterator(new FileLineIterable(Resources.getResource(
        "retail.dat").openStream()), "\\s+");
      Collection<List<String>> transactions = Lists.newArrayList();
     
      while (it.hasNext()) {
        Pair<List<String>,Long> next = it.next();
        transactions.add(next.getFirst());
      }
     
      for (List<String> transaction : transactions) {
        String sep = "";
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

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
69
70
71
  @Test
  public void testSpecificCaseFromRetailDataMinSup500() throws IOException {
    FPGrowthObj<String> fp = new FPGrowthObj<String>();
   
    StringRecordIterator it = new StringRecordIterator(new FileLineIterable(Resources.getResource(
      "retail.dat").openStream()), "\\s+");
    int pattern_41_36_39 = 0;
    while (it.hasNext()) {
      Pair<List<String>,Long> next = it.next();
      List<String> items = next.getFirst();
      if (items.contains("41") && items.contains("36") && items.contains("39")) {
        pattern_41_36_39++;
      }
    }
   
    final Map<Set<String>,Long> results = Maps.newHashMap();
   
    Set<String> returnableFeatures = new HashSet<String>();
    returnableFeatures.add("41");
    returnableFeatures.add("36");
    returnableFeatures.add("39");
   
    fp.generateTopKFrequentPatterns(
      new StringRecordIterator(new FileLineIterable(Resources.getResource("retail.dat").openStream()), "\\s+"),

      fp.generateFList(new StringRecordIterator(new FileLineIterable(Resources.getResource("retail.dat")
          .openStream()), "\\s+"), 500), 500, 1000, returnableFeatures,
      new OutputCollector<String,List<Pair<List<String>,Long>>>() {
       
        @Override
        public void collect(String key, List<Pair<List<String>,Long>> value) {
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
    File input = new File(inputDir, "test.txt");
    params.set(PFPGrowth.INPUT, input.getAbsolutePath());
    params.set(PFPGrowth.OUTPUT, outputDir.getAbsolutePath());
    Writer writer = Files.newWriter(input, Charsets.UTF_8);
    try {
      StringRecordIterator it = new StringRecordIterator(new FileLineIterable(Resources.getResource(
        "retail.dat").openStream()), "\\s+");
      Collection<List<String>> transactions = Lists.newArrayList();
     
      while (it.hasNext()) {
        Pair<List<String>,Long> next = it.next();
        transactions.add(next.getFirst());
      }
     
      for (List<String> transaction : transactions) {
        String sep = "";
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
  /**
   * Test Parallel FPGrowth on retail data using top-level runPFPGrowth() method
   */
  @Test
  public void testRetailDataMinSup100() throws Exception {
    StringRecordIterator it = new StringRecordIterator(new FileLineIterable(Resources.getResource(
      "retail_results_with_min_sup_100.dat").openStream()), "\\s+");
    Map<Set<String>,Long> expectedResults = Maps.newHashMap();
    while (it.hasNext()) {
      Pair<List<String>,Long> next = it.next();
      List<String> items = Lists.newArrayList(next.getFirst());
      String supportString = items.remove(items.size() - 1);
      Long support = Long.parseLong(supportString.substring(1, supportString.length() - 1));
      expectedResults.put(new HashSet<String>(items), support);
    }
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
  /**
   * Test Parallel FPG on retail data, running various stages individually
   */
  @Test
  public void testRetailDataMinSup100InSteps() throws Exception {
    StringRecordIterator it = new StringRecordIterator(new FileLineIterable(Resources.getResource(
      "retail_results_with_min_sup_100.dat").openStream()), "\\s+");   
    Map<Set<String>,Long> expectedResults = Maps.newHashMap();
    while (it.hasNext()) {
      Pair<List<String>,Long> next = it.next();
      List<String> items = Lists.newArrayList(next.getFirst());
      String supportString = items.remove(items.size() - 1);
      Long support = Long.parseLong(supportString.substring(1, supportString.length() - 1));
      expectedResults.put(new HashSet<String>(items), support);
    }
View Full Code Here

Examples of org.apache.mahout.common.iterator.StringRecordIterator

68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
    File input = new File(inputDir, "synth_test.txt");
    params.set(PFPGrowth.INPUT, input.getAbsolutePath());
    params.set(PFPGrowth.OUTPUT, outputDir.getAbsolutePath());
    Writer writer = Files.newWriter(input, Charsets.UTF_8);
    try {
      StringRecordIterator it = new StringRecordIterator(new FileLineIterable(Resources.getResource(
        "FPGsynth.dat").openStream()), "\\s+");
      Collection<List<String>> transactions = Lists.newArrayList();
     
      while (it.hasNext()) {
        Pair<List<String>,Long> next = it.next();
        transactions.add(next.getFirst());
      }
     
      for (List<String> transaction : transactions) {
        String sep = "";
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.