Examples of BagFactory

org.apache.pig.data.BagFactory
Factory for constructing different types of bags. This class is abstract so that users can override the bag factory if they desire to provide their own that returns their implementation of a bag. If the property pig.data.bag.factory.name is set to a class name and pig.data.bag.factory.jar is set to a URL pointing to a jar that contains the above named class, then getInstance() will create an instance of the named class using the indicated jar. Otherwise, it will create an instance of DefaultBagFactory.

Examples of org.apache.pig.data.BagFactory

            objList.add(new Double(1.0));
            objList.add(new Float(1.0));
            objList.add(new String("World!"));
            Tuple tuple = tupleFactory.newTuple(objList);


            BagFactory bagFactory = BagFactory.getInstance();
            DataBag bag = bagFactory.newDefaultBag();
            bag.add(tuple);


            Map<String, Object> mapInMap = new HashMap<String, Object>();
            mapInMap.put("int", new Integer(10));
            mapInMap.put("float", new Float(10.0));

View Full Code Here

Examples of org.apache.pig.data.BagFactory

  assertEquals("", m.get("k3"), "foo");




        TOP top = new TOP();
        TupleFactory tupleFactory = TupleFactory.getInstance();
        BagFactory bagFactory = DefaultBagFactory.getInstance();
        Tuple inputTuple = tupleFactory.newTuple(3);
        DataBag dBag = bagFactory.newDefaultBag();


        // set N = 10 i.e retain top 10 tuples
        inputTuple.set(0, 10);
        // compare tuples by field number 1
        inputTuple.set(1, 1);
        // set the data bag containing the tuples
        inputTuple.set(2, dBag);


        // generate tuples of the form (group-1, 1), (group-2, 2) ...
        for (long i = 0; i < 100; i++) {
            Tuple nestedTuple = tupleFactory.newTuple(2);
            nestedTuple.set(0, "group-" + i);
            nestedTuple.set(1, i);
            dBag.add(nestedTuple);
        }


        DataBag outBag = top.exec(inputTuple);
        assertEquals(outBag.size(), 10L);
        checkItemsGT(outBag, 1, 89);


        // two initial results
        Tuple init1 = (new TOP.Initial()).exec(inputTuple);
        Tuple init2 = (new TOP.Initial()).exec(inputTuple);
        // two intermediate results


        DataBag intermedBag = bagFactory.newDefaultBag();
        intermedBag.add(init1);
        intermedBag.add(init2);
        Tuple intermedInput = tupleFactory.newTuple(intermedBag);
        Tuple intermedOutput1 = (new TOP.Intermed()).exec(intermedInput);
        Tuple intermedOutput2 = (new TOP.Intermed()).exec(intermedInput);
        checkItemsGT((DataBag)intermedOutput1.get(2), 1, 94);


        // final result
        DataBag finalInputBag = bagFactory.newDefaultBag();
        finalInputBag.add(intermedOutput1);
        finalInputBag.add(intermedOutput2);
        Tuple finalInput = tupleFactory.newTuple(finalInputBag);
        outBag = (new TOP.Final()).exec(finalInput);
        assertEquals(outBag.size(), 10L);

View Full Code Here

Examples of org.apache.pig.data.BagFactory

    }


    @Test
    public void testDIFF() throws Exception {
        // Test it in the case with two bags.
        BagFactory bf = BagFactory.getInstance();
        TupleFactory tf = TupleFactory.getInstance();


        DataBag b1 = bf.newDefaultBag();
        DataBag b2 = bf.newDefaultBag();
        for (int i = 0; i < 10; i++) b1.add(tf.newTuple(new Integer(i)));
        for (int i = 0; i < 10; i += 2) b2.add(tf.newTuple(new Integer(i)));
        Tuple t = tf.newTuple(2);
        t.set(0, b1);
        t.set(1, b2);

View Full Code Here

Examples of org.apache.pig.data.BagFactory

        TupleFactory tf = TupleFactory.getInstance();
        return tf.newTuple( objList );
    }
    
    static DataBag createDataBag() {
        BagFactory bagFactory = BagFactory.getInstance();
        return bagFactory.newDefaultBag();
    }

View Full Code Here

Examples of org.apache.pig.data.BagFactory

    public void testMap() throws Exception {
        String size = "100";
        String numHash = "3";
        String hashFunc = "JENKINS";
        TupleFactory tf = TupleFactory.getInstance();
        BagFactory bf = BagFactory.getInstance();


        Tuple t = tf.newTuple(1);
        t.set(0, 1);
        DataBag b = bf.newDefaultBag();
        b.add(t);
        Tuple input = tf.newTuple(b);
        
        BuildBloom.Initial map =
            new BuildBloom.Initial(hashFunc, "fixed", size, numHash);

View Full Code Here

Examples of org.apache.pig.data.BagFactory

    public void testCombiner() throws Exception {
        String size = "100";
        String numHash = "3";
        String hashFunc = "jenkins";
        TupleFactory tf = TupleFactory.getInstance();
        BagFactory bf = BagFactory.getInstance();


        DataBag combinerBag = bf.newDefaultBag();
        for (int j = 0; j < 3; j++) { // map loop
            Tuple t = tf.newTuple(1);
            t.set(0, 10 + j);
            DataBag mapBag = bf.newDefaultBag();
            mapBag.add(t);
            Tuple input = tf.newTuple(mapBag);
            BuildBloom.Initial map =
                new BuildBloom.Initial(hashFunc, "fixed", size, numHash);
            combinerBag.add(map.exec(input));

View Full Code Here

Examples of org.apache.pig.data.BagFactory

    public void testSingleKey() throws Exception {
        String size = "100";
        String numHash = "3";
        String hashFunc = "MURMUR";
        TupleFactory tf = TupleFactory.getInstance();
        BagFactory bf = BagFactory.getInstance();


        DataBag reducerBag = bf.newDefaultBag();
        for (int i = 0; i < 3; i++) { // combiner loop
            DataBag combinerBag = bf.newDefaultBag();
            for (int j = 0; j < 3; j++) { // map loop
                Tuple t = tf.newTuple(1);
                t.set(0, i * 10 + j);
                DataBag mapBag = bf.newDefaultBag();
                mapBag.add(t);
                Tuple input = tf.newTuple(mapBag);
                BuildBloom.Initial map =
                    new BuildBloom.Initial(hashFunc, "fixed", size, numHash);
                combinerBag.add(map.exec(input));

View Full Code Here

Examples of org.apache.pig.data.BagFactory

    public void testMultiKey() throws Exception {
        String numElements = "10";
        String falsePositive = "0.001";
        String hashFunc = "murmur";
        TupleFactory tf = TupleFactory.getInstance();
        BagFactory bf = BagFactory.getInstance();


        String[][] strs = {
            { "fred", "joe", "bob" },
            { "mary", "sally", "jane" },
            { "fido", "spot", "fluffly" }};


        DataBag reducerBag = bf.newDefaultBag();
        for (int i = 0; i < 3; i++) { // combiner loop
            DataBag combinerBag = bf.newDefaultBag();
            for (int j = 0; j < 3; j++) { // map loop
                Tuple t = tf.newTuple(2);
                t.set(0, i * 10 + j);
                t.set(1, strs[i][j]);
                DataBag mapBag = bf.newDefaultBag();
                mapBag.add(t);
                Tuple input = tf.newTuple(mapBag);
                BuildBloom.Initial map =
                    new BuildBloom.Initial(hashFunc, numElements,
                        falsePositive);

View Full Code Here

Examples of org.apache.pig.data.BagFactory

            objList.add(new Double(1.0));
            objList.add(new Float(1.0));
            objList.add(new String("World!"));
            Tuple tuple = tupleFactory.newTuple(objList);


            BagFactory bagFactory = BagFactory.getInstance();
            DataBag bag = bagFactory.newDefaultBag();
            bag.add(tuple);


            Map<String, Object> mapInMap = new HashMap<String, Object>();
            mapInMap.put("int", new Integer(10));
            mapInMap.put("float", new Float(10.0));

View Full Code Here

Examples of org.apache.pig.data.BagFactory

        TupleFactory tf = TupleFactory.getInstance();
        return tf.newTuple( objList );
    }
    
    static DataBag createDataBag() {
        BagFactory bagFactory = BagFactory.getInstance();
        return bagFactory.newDefaultBag();
    }

View Full Code Here

0 1 2 3 4 5

TOP

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.