Examples of SequenceFile

Examples of cascading.scheme.hadoop.SequenceFile

60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79

  @Test
  public void testSimpleCombinerWithMemoryLimit() throws IOException {

    Hfs source = new Hfs(new SequenceFile(new Fields("key", "value")), INPUT_PATH);
    TupleEntryCollector tc = source.openForWrite(CascadingUtil.get().getFlowProcess());
    tc.add(new Tuple("key0", 1));
    tc.add(new Tuple("key0", 2));
    tc.add(new Tuple("key1", 1));
    tc.add(new Tuple("key1", -3));
    tc.add(new Tuple("key0", 10));
    tc.close();

    Tap sink = new Hfs(new SequenceFile(new Fields("key", "sum")), OUTPUT_PATH);

    Pipe pipe = new Pipe("pipe");
    pipe = new Each(pipe, Combiner.function(new SimpleAggregator(), new Fields("key"), new Fields("value"), new Fields("sum"), MemoryBoundLruHashMap.UNLIMITED_ITEM_CAPACITY, 100, new SimpleTupleMemoryUsageEstimator(), new LongMemoryUsageEstimator(), false));

    CascadingUtil.get().getFlowConnector().connect(source, sink, pipe).complete();
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
  }

  @Test
  public void testComplexCombiner() throws IOException {

    Hfs source = new Hfs(new SequenceFile(new Fields("key", "value")), INPUT_PATH);
    TupleEntryCollector tc = source.openForWrite(CascadingUtil.get().getFlowProcess());
    tc.add(new Tuple("k0", 1));
    tc.add(new Tuple("k0", 2));
    tc.add(new Tuple("k1", 1));
    tc.add(new Tuple("k1", 4));
    tc.add(new Tuple("k1", 10));
    tc.close();

    Tap sink = new Hfs(new SequenceFile(new Fields("key", "sum", "num_values", "average")), OUTPUT_PATH);

    Pipe pipe = new Pipe("pipe");
    pipe = Combiner.assembly(pipe,
        new ComplexAggregator(),
        new Fields("key"),
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
  private final String INPUT_PATH = getTestRoot() + "/input";
  private final String OUTPUT_PATH = getTestRoot() + "/output";

  @Test
  public void testMain() throws IOException {
    Hfs source = new Hfs(new SequenceFile(new Fields("key", "value")), INPUT_PATH);
    TupleEntryCollector tc = source.openForWrite(CascadingUtil.get().getFlowProcess());
    tc.add(new Tuple("k0", 1));
    tc.add(new Tuple("k0", 2));
    tc.add(new Tuple("k1", 10));
    tc.add(new Tuple("k1", -2));
    tc.add(new Tuple("k1", -9));
    tc.add(new Tuple("k2", -3));
    tc.add(new Tuple("k2", -6));
    tc.close();

    Tap sink = new Hfs(new SequenceFile(new Fields("key", "sum")), OUTPUT_PATH);
   
    Pipe pipe = new Pipe("pipe");
    pipe = Combiner.assembly(pipe, new SumExactAggregator(1), new Fields("key"), new Fields("value"), new Fields("sum"));

    CascadingUtil.get().getFlowConnector().connect(source, sink, pipe).complete();
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

40
41
42
43
44
45
46
  protected Hfs output;

  @Before
  public void setUp() throws Exception {
    output = new Hfs(new SequenceFile(new Fields("key", "key2", "lhs-value")), getTestRoot() + "/output");
  }
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
  private final String OUTPUT_PATH = getTestRoot() + "/output";

  @Test
  public void testMain() throws IOException {

    Hfs source = new Hfs(new SequenceFile(new Fields("key", "a", "b")), INPUT_PATH);
    TupleEntryCollector tc = source.openForWrite(CascadingUtil.get().getFlowProcess());
    tc.add(new Tuple("k0", 1, 1));
    tc.add(new Tuple("k0", 2, 5));
    tc.add(new Tuple("k1", 1, 7));
    tc.add(new Tuple("k1", -2, 10));
    tc.add(new Tuple("k1", -2, -9));
    tc.close();

    Tap sink = new Hfs(new SequenceFile(new Fields("key", "a_sum", "b_sum", "a_b_sum", "a_b_sum_doubled")), OUTPUT_PATH);

    Pipe pipe = new Pipe("pipe");
    pipe = Combiner.assembly(pipe,
        new MultiExactAggregator(
            new ExactAggregatorDefinition(new Fields("a"), new Fields("a_sum"), new Aggregator()),
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
  private final String SOURCE2 = getTestRoot() + "/mgb_source2";

  @Before
  public void setUp() throws IOException {

    source1 = new Hfs(new SequenceFile(new Fields("key", "num")), SOURCE1);
    TapHelper.writeToTap(source1,
        new Tuple(1, 1),
        new Tuple(1, 3),
        new Tuple(1, 2),
        new Tuple(2, 5),
        new Tuple(3, 3),
        new Tuple(3, 3));

    source2 = new Hfs(new SequenceFile(new Fields("key", "num1", "num2")), SOURCE2);
    TapHelper.writeToTap(source2,
        new Tuple(1, 101, 1),
        new Tuple(5, 5, 2),
        new Tuple(3, 0, 0));
  }
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

73
74
75
76
77
78
79
80
81
82
83
  @Test
  public void testSimple3() throws Exception {

    final String OUTPUT = getTestRoot() + "/mgb_output";

    Hfs sink = new Hfs(new SequenceFile(new Fields("key", "result", " result1", "result2", "result3", "result4", "result5")), OUTPUT);

    Map<String, Tap> sources = new HashMap<String, Tap>();
    sources.put("s1", source1);
    sources.put("s2", source2);
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

105
106
107
108
109
110
111
112
113
114
115
  @Test
  public void testSimple() throws Exception {
    final String OUTPUT = getTestRoot() + "/mgb_output";

    Hfs sink = new Hfs(new SequenceFile(new Fields("key-rename", "result", " result1", "result2", "result3", "result4", "result5")), OUTPUT);

    Map<String, Tap> sources = new HashMap<String, Tap>();
    sources.put("s1", source1);
    sources.put("s2", source2);
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

136
137
138
139
140
141
142
143
144
145
146
  @Test
  public void testSimple2() throws Exception {
    final String OUTPUT = getTestRoot() + "/mgb_output";

    Hfs sink = new Hfs(new SequenceFile(new Fields("key-rename", "result")), OUTPUT);

    Map<String, Tap> sources = new HashMap<String, Tap>();
    sources.put("s1", source1);
    sources.put("s2", source2);
View Full Code Here

Examples of cascading.scheme.hadoop.SequenceFile

157
158
159
160
161
162
163
    verifyRequestsPerDay(sinks.get(def4.getName()));
    verifyRequestsPerAttribute(sinks.get(def5.getName()));
  }

  private Tap getTupleOutputTap(String testname, String name, Fields fields) {
    return new Hfs(new SequenceFile(fields), getTestRoot() + "/multi_combiner_output/" + testname + "/" + name);
  }
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.