Package ivory.lsh.data

Examples of ivory.lsh.data.MinhashSignature

172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
  @Test
  public void testSubSignature() {
    for (int i = 0; i < 100; i++) {

      MinhashSignature s = getRandomSignature();

      System.out.println(s);

      MinhashSignature slide = s.getSubSignature(0, D / 2);
      MinhashSignature slide2 = s.getSubSignature(D / 2 + 1, D - 1);
      System.out.println(slide + "," + slide2);

      assertTrue(s.toString().equals(slide.toString() + "," + slide2.toString()));
    }
    System.out.println("done");
  }
View Full Code Here
22
23
24
25
26
27
28
29
30
31
32
33
  private static final String TMP_FILENAME1 = "tmp1.out";

  private static int D = 20, vocabSize = 1000;

  private MinhashSignature getRandomSignature() {
    MinhashSignature s = new MinhashSignature(D);
    for (int i = 0; i < D; i++) {
      int elt = (int) (Math.random() * vocabSize);
      s.add(elt);
    }
    return s;
  }
View Full Code Here
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
  @Test
  public void testPermute() throws IOException {
    PermutationByBit p = new PermutationByBit(D);

    MinhashSignature s = getRandomSignature();
    System.out.println(s);

    int loopcnt = 0;
    MinhashSignature permutedS = new MinhashSignature(D);
    while (loopcnt++ < 10) {
      ArrayListOfIntsWritable a = p.nextPermutation();
      s.perm(a, permutedS);
      for (int i = 0; i < s.size(); i++) {
        assertTrue(permutedS.containsTerm(s.get(i)));
      }
      assertTrue(permutedS.size() == s.size());
      System.out.println(permutedS);
    }
  }
View Full Code Here
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
    }
  }

  @Test
  public void testReadWrite() throws IOException {
    MinhashSignature s = getRandomSignature();
    MinhashSignature s2 = getRandomSignature();

    FileSystem fs;
    SequenceFile.Writer w;
    Configuration conf = new Configuration();

    try {
      fs = FileSystem.get(conf);
      w = SequenceFile.createWriter(fs, conf, new Path(TMP_FILENAME1),
          IntWritable.class, MinhashSignature.class);
      w.append(new IntWritable(1), s);
      w.append(new IntWritable(2), s2);
      w.close();
    } catch (IOException e) {
      e.printStackTrace();
    }

    List<PairOfWritables<WritableComparable<IntWritable>, Writable>> listOfKeysPairs =
        SequenceFileUtils.readFile(new Path(TMP_FILENAME1));
    FileSystem.get(conf).delete(new Path(TMP_FILENAME1), true);

    MinhashSignature read1 = (MinhashSignature) listOfKeysPairs.get(0).getRightElement();
    MinhashSignature read2 = (MinhashSignature) listOfKeysPairs.get(1).getRightElement();

    assertTrue(read1.toString().equals(s.toString()));
    assertTrue(read2.toString().equals(s2.toString()));

    System.out.println(read1.toString());
    System.out.println(read2.toString());
  }
View Full Code Here
90
91
92
93
94
95
96
97
98
99
100
    FileSystem fs;
    SequenceFile.Writer w;
    Configuration conf = new Configuration();

    try {
      MinhashSignature s = getRandomSignature();
      fs = FileSystem.get(conf);
      w = SequenceFile.createWriter(fs, conf, new Path("test2"),
          IntWritable.class, MinhashSignature.class);
      for (int i = 0; i < 1000000; i++) {
        w.append(new IntWritable(1), s);
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
  }

  @Test
  public void testBasic() {
    MinhashSignature s = new MinhashSignature(D);

    s.add(1);
    s.add(2);
    assertTrue(s.get(0) == 1);
    assertTrue(s.get(1) == 2);

    s.set(0, 3);
    assertTrue(s.get(0) == 3);
  }
View Full Code Here
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
    assertTrue(s.get(0) == 3);
  }

  @Test
  public void testHammingDistance() {
    MinhashSignature s1 = new MinhashSignature(D);
    MinhashSignature s2 = new MinhashSignature(D);
    s1.add(1);
    s1.add(2);
    s1.add(3);
    s1.add(4);
    s1.add(5);

    s2.add(3);
    s2.add(2);
    s2.add(5);
    s2.add(4);
    s2.add(1);

    assertTrue(s1.hammingDistance(s2) == 3);
    assertTrue(s1.hammingDistance(s2, 2) == 3);
    assertTrue(s1.hammingDistance(s2, 5) == 3);
View Full Code Here
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
  }

  @Test
  public void testCompare() {
    MinhashSignature s1 = new MinhashSignature(D);
    MinhashSignature s2 = new MinhashSignature(D);

    s1.add(1);
    s1.add(2);
    s1.add(3);
    s1.add(4);
    s1.add(5);

    s2.add(3);
    s2.add(2);
    s2.add(5);
    s2.add(4);
    s2.add(1);

    assertTrue(s1.compareTo(s2) + "", s1.compareTo(s2) < 0);
    assertTrue(s2.compareTo(s1) + "", s2.compareTo(s1) > 0);
    assertTrue(s1.compareTo(s1) + "", s1.compareTo(s1) == 0);
    assertTrue(s2.compareTo(s2) + "", s2.compareTo(s2) == 0);

  }
View Full Code Here
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
  @Test
  public void testSubSignature() {
    for (int i = 0; i < 100; i++) {

      MinhashSignature s = getRandomSignature();

      System.out.println(s);

      MinhashSignature slide = s.getSubSignature(0, D / 2);
      MinhashSignature slide2 = s.getSubSignature(D / 2 + 1, D - 1);
      System.out.println(slide + "," + slide2);

      assertTrue(s.toString().equals(slide.toString() + "," + slide2.toString()));
    }
    System.out.println("done");
  }
View Full Code Here
38
39
40
41
42
43
44
45
46
47
48
    }
    return s;
  }

  private MinhashSignature getRandomMinhashSignature() {
    MinhashSignature s = new MinhashSignature(D);
    for (int i = 0; i < s.size(); i++) {
      s.add((int) (Math.random() * 39000));
    }
    return s;
  }
View Full Code Here
TOP

Related Classes of ivory.lsh.data.MinhashSignature

  • edu.umd.cloud9.io.array.ArrayListOfIntsWritable
  • ivory.lsh.MinhashSignatureTest
  • ivory.lsh.projection.ComputeSignaturesMinhash$MyMapper
  • ivory.lsh.SignatureTest
    • 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.