Package org.apache.pig.data

Examples of org.apache.pig.data.DefaultDataBag$DefaultDataBagIterator

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
    assertEquals(1.11111, ans.getAtomField(2).numval(),0.0005);
  }
 
  public void testCOR() throws Exception{
    EvalFunc<DataBag> COR = new COR("a","b");
    DataBag dBag = new DefaultDataBag();
    Tuple tup1 = new Tuple(1);
    tup1.setField(0, 1);
    dBag.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 4);
    dBag.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 8);
    dBag.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 4);
    dBag.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 7);
    dBag.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 8);
    dBag.add(tup1);
    DataBag dBag1 = new DefaultDataBag();
    tup1 = new Tuple(1);
    tup1.setField(0, 2);
    dBag1.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 2);
    dBag1.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 3);
    dBag1.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 3);
    dBag1.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 2);
    dBag1.add(tup1);
    tup1 = new Tuple(1);
    tup1.setField(0, 4);
    dBag1.add(tup1);
    Tuple input = new Tuple(2);
    input.setField(0, dBag);
    input.setField(1, dBag1);
    DataBag output = new DefaultDataBag();
    COR.exec(input, output);
    Iterator<Tuple> it = output.iterator();
    Tuple ans = (Tuple) it.next();
    assertEquals(ans.getAtomField(0).toString(),"a");
    assertEquals(ans.getAtomField(1).toString(),"b");
    assertEquals(0.582222509739582, ans.getAtomField(2).numval(),0.0005);
  }
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
  public static DataBag[] testDataBagEvals(EvalFunc<DataBag> eval, Tuple[] tuples) {
   
    List<DataBag> res = new ArrayList<DataBag>();
    try {
      for (Tuple t : tuples) {
        DataBag bag = new DefaultDataBag();
        eval.exec(t, bag);
        System.out.println("Converted: " + t + " to (" + bag + ")");
        res.add(bag);
      }
    } catch (IOException e) {
View Full Code Here
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
152
153
154
155
156
157
158
159
160
    };
   
    ExtractHour eval2 = new ExtractHour();
    testDataAtomEvals(eval2, getTuples(timestamps));

    DataBag bag = new DefaultDataBag();
   
    Tuple t1 = new Tuple();
    t1.appendField(new DataAtom("word"));
    t1.appendField(new DataAtom("02"));
    t1.appendField(new DataAtom(2));
    bag.add(t1);
   
    Tuple t2 = new Tuple();
    t2.appendField(new DataAtom("word"));
    t2.appendField(new DataAtom("05"));
    t2.appendField(new DataAtom(2));
    bag.add(t2);

    Tuple t3 = new Tuple();
    t3.appendField(new DataAtom("word"));
    t3.appendField(new DataAtom("04"));
    t3.appendField(new DataAtom(3));
    bag.add(t3);

    Tuple t4 = new Tuple();
    t4.appendField(new DataAtom("word"));
    t4.appendField(new DataAtom("06"));
    t4.appendField(new DataAtom(4));
    bag.add(t4);

    Tuple[] t = new Tuple[1];
    t[0] = new Tuple();
    t[0].appendField(bag);
View Full Code Here
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
    @SuppressWarnings("unchecked")
    @Test
    public void testBagOfAnnotationMerge() throws IOException {
        TupleFactory tf = TupleFactory.getInstance();
        DefaultDataBag textBag = new DefaultDataBag();
        DefaultDataBag beginBag = new DefaultDataBag();
        DefaultDataBag endBag = new DefaultDataBag();
        DefaultDataBag typeBag = new DefaultDataBag();

        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        beginBag.add(tf.newTupleNoCopy(Arrays.asList(0)));
        endBag.add(tf.newTupleNoCopy(Arrays.asList(10)));
        typeBag.add(tf.newTupleNoCopy(Arrays.asList("person")));

        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        beginBag.add(tf.newTupleNoCopy(Arrays.asList(19)));
        endBag.add(tf.newTupleNoCopy(Arrays.asList(36)));
        typeBag.add(tf.newTupleNoCopy(Arrays.asList("organization")));

        // all bags
        Tuple input = tf.newTupleNoCopy(Arrays.asList(textBag, beginBag,
                endBag, typeBag));
        String merged = merger.exec(input);
View Full Code Here
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
    }

    @Test
    public void testAggregateBagOfText() throws IOException {
        TupleFactory tf = TupleFactory.getInstance();
        DefaultDataBag textBag = new DefaultDataBag();

        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE)));

        // all bags
        Tuple input = tf.newTupleNoCopy(Arrays.asList(textBag));
        String merged = aggregator.exec(input);
        String expected = StringUtils.join(
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
    }

    @Test
    public void testAggregateBagOfTextWithTabs() throws IOException {
        TupleFactory tf = TupleFactory.getInstance();
        DefaultDataBag textBag = new DefaultDataBag();

        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE_WITH_TABS)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE_WITH_TABS)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE_WITH_TABS)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE_WITH_TABS)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE_WITH_TABS)));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(JOHN_SENTENCE_WITH_TABS)));

        // all bags
        Tuple input = tf.newTupleNoCopy(Arrays.asList(textBag));
        String merged = aggregator.exec(input);
        String expected = StringUtils.join(
View Full Code Here
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
    }

    @Test
    public void testConcatTextBag() throws IOException {
        TupleFactory tf = TupleFactory.getInstance();
        DefaultDataBag textBag = new DefaultDataBag();

        textBag.add(tf.newTupleNoCopy(Arrays.asList(" foo1")));
        textBag.add(tf.newTupleNoCopy(Arrays.asList("foo2")));
        textBag.add(tf.newTupleNoCopy(Arrays.asList(" foo3 ")));
        textBag.add(tf.newTupleNoCopy(Arrays.asList("foo1")));
        textBag.add(tf.newTupleNoCopy(Arrays.asList("foo4 ")));

        // all bags
        Tuple input = tf.newTupleNoCopy(Arrays.asList(textBag));
        String merged = concatTextBag.exec(input);
        String expected = "foo1 _ foo2 _ foo3 _ foo4";
View Full Code Here
341
342
343
344
345
346
347
348
349
350
351
352
353
      return null;
    }

    HCatFieldSchema elementSubFieldSchema = hfs.getArrayElementSchema().getFields().get(0);
    if (elementSubFieldSchema.getType() == Type.STRUCT){
      DataBag db = new DefaultDataBag();
      for (Object o : list){
        db.add(transformToTuple((List<Object>)o,elementSubFieldSchema));
      }
      return db;
    } else {
      return  new HCatArrayBag(list);
    }
View Full Code Here
157
158
159
160
161
162
163
164
    return 0;
  }

  @Override
  public void readFields(DataInput arg0) throws IOException {
    convertedBag = new DefaultDataBag();
    convertedBag.readFields(arg0);
  }
View Full Code Here
65
66
67
68
69
70
71
   * Create a PIG Bag object.
   *
   * @return A Pig DataBag object.
   */
  public static DataBag createBag() {
    return new DefaultDataBag();
  }
View Full Code Here
TOP

Related Classes of org.apache.pig.data.DefaultDataBag$DefaultDataBagIterator

  • com.inadco.hbl.piggybank.Dimensions2CuboidKey
  • com.packetloop.packetpig.udf.util.Explode
  • org.apache.cassandra.hadoop.pig.CassandraStorage
  • org.apache.hadoop.zebra.types.TypesUtils
  • org.apache.hcatalog.data.HCatArrayBag
  • org.apache.hcatalog.pig.PigHCatUtil
  • org.apache.hive.hcatalog.pig.PigHCatUtil
  • org.apache.howl.data.HowlArrayBag
  • org.apache.howl.pig.PigHowlUtil
  • org.apache.pig.backend.hadoop.accumulo.TestAbstractAccumuloStorage
    • 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.