Search in sources :

Example 6 with ReduceOperatorBase

use of org.apache.flink.api.common.operators.base.ReduceOperatorBase in project flink by apache.

the class ReduceTranslationTests method translateNonGroupedReduce.

@Test
public void translateNonGroupedReduce() {
    try {
        final int parallelism = 8;
        ExecutionEnvironment env = ExecutionEnvironment.createLocalEnvironment(parallelism);
        DataSet<Tuple3<Double, StringValue, LongValue>> initialData = getSourceDataSet(env);
        initialData.reduce(new RichReduceFunction<Tuple3<Double, StringValue, LongValue>>() {

            public Tuple3<Double, StringValue, LongValue> reduce(Tuple3<Double, StringValue, LongValue> value1, Tuple3<Double, StringValue, LongValue> value2) {
                return value1;
            }
        }).output(new DiscardingOutputFormat<Tuple3<Double, StringValue, LongValue>>());
        Plan p = env.createProgramPlan();
        GenericDataSinkBase<?> sink = p.getDataSinks().iterator().next();
        ReduceOperatorBase<?, ?> reducer = (ReduceOperatorBase<?, ?>) sink.getInput();
        // check types
        assertEquals(initialData.getType(), reducer.getOperatorInfo().getInputType());
        assertEquals(initialData.getType(), reducer.getOperatorInfo().getOutputType());
        // check keys
        assertTrue(reducer.getKeyColumns(0) == null || reducer.getKeyColumns(0).length == 0);
        // parallelism was not configured on the operator
        assertTrue(reducer.getParallelism() == 1 || reducer.getParallelism() == ExecutionConfig.PARALLELISM_DEFAULT);
        assertTrue(reducer.getInput() instanceof GenericDataSourceBase<?, ?>);
    } catch (Exception e) {
        System.err.println(e.getMessage());
        e.printStackTrace();
        fail("Test caused an error: " + e.getMessage());
    }
}
Also used : ExecutionEnvironment(org.apache.flink.api.java.ExecutionEnvironment) ReduceOperatorBase(org.apache.flink.api.common.operators.base.ReduceOperatorBase) Plan(org.apache.flink.api.common.Plan) RichReduceFunction(org.apache.flink.api.common.functions.RichReduceFunction) Tuple3(org.apache.flink.api.java.tuple.Tuple3) LongValue(org.apache.flink.types.LongValue) StringValue(org.apache.flink.types.StringValue) Test(org.junit.Test)

Example 7 with ReduceOperatorBase

use of org.apache.flink.api.common.operators.base.ReduceOperatorBase in project flink by apache.

the class DistinctOperator method translateToDataFlow.

@Override
protected org.apache.flink.api.common.operators.SingleInputOperator<?, T, ?> translateToDataFlow(Operator<T> input) {
    final ReduceFunction<T> function = new DistinctFunction<>();
    String name = getName() != null ? getName() : "Distinct at " + distinctLocationName;
    if (keys instanceof Keys.ExpressionKeys) {
        int[] logicalKeyPositions = keys.computeLogicalKeyPositions();
        UnaryOperatorInformation<T, T> operatorInfo = new UnaryOperatorInformation<>(getInputType(), getResultType());
        ReduceOperatorBase<T, ReduceFunction<T>> po = new ReduceOperatorBase<>(function, operatorInfo, logicalKeyPositions, name);
        po.setCombineHint(hint);
        po.setInput(input);
        po.setParallelism(getParallelism());
        // make sure that distinct preserves the partitioning for the fields on which they operate
        if (getType().isTupleType()) {
            SingleInputSemanticProperties sProps = new SingleInputSemanticProperties();
            for (int field : keys.computeLogicalKeyPositions()) {
                sProps.addForwardedField(field, field);
            }
            po.setSemanticProperties(sProps);
        }
        return po;
    } else if (keys instanceof SelectorFunctionKeys) {
        @SuppressWarnings("unchecked") SelectorFunctionKeys<T, ?> selectorKeys = (SelectorFunctionKeys<T, ?>) keys;
        org.apache.flink.api.common.operators.SingleInputOperator<?, T, ?> po = translateSelectorFunctionDistinct(selectorKeys, function, getResultType(), name, input, parallelism, hint);
        return po;
    } else {
        throw new UnsupportedOperationException("Unrecognized key type.");
    }
}
Also used : SelectorFunctionKeys(org.apache.flink.api.common.operators.Keys.SelectorFunctionKeys) ReduceOperatorBase(org.apache.flink.api.common.operators.base.ReduceOperatorBase) ReduceFunction(org.apache.flink.api.common.functions.ReduceFunction) CombineHint(org.apache.flink.api.common.operators.base.ReduceOperatorBase.CombineHint) UnaryOperatorInformation(org.apache.flink.api.common.operators.UnaryOperatorInformation) SingleInputSemanticProperties(org.apache.flink.api.common.operators.SingleInputSemanticProperties)

Example 8 with ReduceOperatorBase

use of org.apache.flink.api.common.operators.base.ReduceOperatorBase in project flink by apache.

the class ReduceTranslationTests method translateGroupedReduceNoMapper.

@Test
public void translateGroupedReduceNoMapper() {
    try {
        final int parallelism = 8;
        ExecutionEnvironment env = ExecutionEnvironment.createLocalEnvironment(parallelism);
        DataSet<Tuple3<Double, StringValue, LongValue>> initialData = getSourceDataSet(env);
        initialData.groupBy(2).reduce(new RichReduceFunction<Tuple3<Double, StringValue, LongValue>>() {

            public Tuple3<Double, StringValue, LongValue> reduce(Tuple3<Double, StringValue, LongValue> value1, Tuple3<Double, StringValue, LongValue> value2) {
                return value1;
            }
        }).output(new DiscardingOutputFormat<Tuple3<Double, StringValue, LongValue>>());
        Plan p = env.createProgramPlan();
        GenericDataSinkBase<?> sink = p.getDataSinks().iterator().next();
        ReduceOperatorBase<?, ?> reducer = (ReduceOperatorBase<?, ?>) sink.getInput();
        // check types
        assertEquals(initialData.getType(), reducer.getOperatorInfo().getInputType());
        assertEquals(initialData.getType(), reducer.getOperatorInfo().getOutputType());
        // parallelism was not configured on the operator
        assertTrue(reducer.getParallelism() == parallelism || reducer.getParallelism() == ExecutionConfig.PARALLELISM_DEFAULT);
        // check keys
        assertArrayEquals(new int[] { 2 }, reducer.getKeyColumns(0));
        assertTrue(reducer.getInput() instanceof GenericDataSourceBase<?, ?>);
    } catch (Exception e) {
        System.err.println(e.getMessage());
        e.printStackTrace();
        fail("Test caused an error: " + e.getMessage());
    }
}
Also used : ExecutionEnvironment(org.apache.flink.api.java.ExecutionEnvironment) ReduceOperatorBase(org.apache.flink.api.common.operators.base.ReduceOperatorBase) Plan(org.apache.flink.api.common.Plan) RichReduceFunction(org.apache.flink.api.common.functions.RichReduceFunction) Tuple3(org.apache.flink.api.java.tuple.Tuple3) LongValue(org.apache.flink.types.LongValue) StringValue(org.apache.flink.types.StringValue) Test(org.junit.Test)

Example 9 with ReduceOperatorBase

use of org.apache.flink.api.common.operators.base.ReduceOperatorBase in project flink by apache.

the class GraphCreatingVisitor method preVisit.

@SuppressWarnings("deprecation")
@Override
public boolean preVisit(Operator<?> c) {
    // check if we have been here before
    if (this.con2node.containsKey(c)) {
        return false;
    }
    final OptimizerNode n;
    // create a node for the operator (or sink or source) if we have not been here before
    if (c instanceof GenericDataSinkBase) {
        DataSinkNode dsn = new DataSinkNode((GenericDataSinkBase<?>) c);
        this.sinks.add(dsn);
        n = dsn;
    } else if (c instanceof GenericDataSourceBase) {
        n = new DataSourceNode((GenericDataSourceBase<?, ?>) c);
    } else if (c instanceof MapOperatorBase) {
        n = new MapNode((MapOperatorBase<?, ?, ?>) c);
    } else if (c instanceof MapPartitionOperatorBase) {
        n = new MapPartitionNode((MapPartitionOperatorBase<?, ?, ?>) c);
    } else if (c instanceof FlatMapOperatorBase) {
        n = new FlatMapNode((FlatMapOperatorBase<?, ?, ?>) c);
    } else if (c instanceof FilterOperatorBase) {
        n = new FilterNode((FilterOperatorBase<?, ?>) c);
    } else if (c instanceof ReduceOperatorBase) {
        n = new ReduceNode((ReduceOperatorBase<?, ?>) c);
    } else if (c instanceof GroupCombineOperatorBase) {
        n = new GroupCombineNode((GroupCombineOperatorBase<?, ?, ?>) c);
    } else if (c instanceof GroupReduceOperatorBase) {
        n = new GroupReduceNode((GroupReduceOperatorBase<?, ?, ?>) c);
    } else if (c instanceof InnerJoinOperatorBase) {
        n = new JoinNode((InnerJoinOperatorBase<?, ?, ?, ?>) c);
    } else if (c instanceof OuterJoinOperatorBase) {
        n = new OuterJoinNode((OuterJoinOperatorBase<?, ?, ?, ?>) c);
    } else if (c instanceof CoGroupOperatorBase) {
        n = new CoGroupNode((CoGroupOperatorBase<?, ?, ?, ?>) c);
    } else if (c instanceof CoGroupRawOperatorBase) {
        n = new CoGroupRawNode((CoGroupRawOperatorBase<?, ?, ?, ?>) c);
    } else if (c instanceof CrossOperatorBase) {
        n = new CrossNode((CrossOperatorBase<?, ?, ?, ?>) c);
    } else if (c instanceof BulkIterationBase) {
        n = new BulkIterationNode((BulkIterationBase<?>) c);
    } else if (c instanceof DeltaIterationBase) {
        n = new WorksetIterationNode((DeltaIterationBase<?, ?>) c);
    } else if (c instanceof Union) {
        n = new BinaryUnionNode((Union<?>) c);
    } else if (c instanceof PartitionOperatorBase) {
        n = new PartitionNode((PartitionOperatorBase<?>) c);
    } else if (c instanceof SortPartitionOperatorBase) {
        n = new SortPartitionNode((SortPartitionOperatorBase<?>) c);
    } else if (c instanceof BulkIterationBase.PartialSolutionPlaceHolder) {
        if (this.parent == null) {
            throw new InvalidProgramException("It is currently not supported to create data sinks inside iterations.");
        }
        final BulkIterationBase.PartialSolutionPlaceHolder<?> holder = (BulkIterationBase.PartialSolutionPlaceHolder<?>) c;
        final BulkIterationBase<?> enclosingIteration = holder.getContainingBulkIteration();
        final BulkIterationNode containingIterationNode = (BulkIterationNode) this.parent.con2node.get(enclosingIteration);
        // catch this for the recursive translation of step functions
        BulkPartialSolutionNode p = new BulkPartialSolutionNode(holder, containingIterationNode);
        p.setParallelism(containingIterationNode.getParallelism());
        n = p;
    } else if (c instanceof DeltaIterationBase.WorksetPlaceHolder) {
        if (this.parent == null) {
            throw new InvalidProgramException("It is currently not supported to create data sinks inside iterations.");
        }
        final DeltaIterationBase.WorksetPlaceHolder<?> holder = (DeltaIterationBase.WorksetPlaceHolder<?>) c;
        final DeltaIterationBase<?, ?> enclosingIteration = holder.getContainingWorksetIteration();
        final WorksetIterationNode containingIterationNode = (WorksetIterationNode) this.parent.con2node.get(enclosingIteration);
        // catch this for the recursive translation of step functions
        WorksetNode p = new WorksetNode(holder, containingIterationNode);
        p.setParallelism(containingIterationNode.getParallelism());
        n = p;
    } else if (c instanceof DeltaIterationBase.SolutionSetPlaceHolder) {
        if (this.parent == null) {
            throw new InvalidProgramException("It is currently not supported to create data sinks inside iterations.");
        }
        final DeltaIterationBase.SolutionSetPlaceHolder<?> holder = (DeltaIterationBase.SolutionSetPlaceHolder<?>) c;
        final DeltaIterationBase<?, ?> enclosingIteration = holder.getContainingWorksetIteration();
        final WorksetIterationNode containingIterationNode = (WorksetIterationNode) this.parent.con2node.get(enclosingIteration);
        // catch this for the recursive translation of step functions
        SolutionSetNode p = new SolutionSetNode(holder, containingIterationNode);
        p.setParallelism(containingIterationNode.getParallelism());
        n = p;
    } else {
        throw new IllegalArgumentException("Unknown operator type: " + c);
    }
    this.con2node.put(c, n);
    // key-less reducer (all-reduce)
    if (n.getParallelism() < 1) {
        // set the parallelism
        int par = c.getParallelism();
        if (par > 0) {
            if (this.forceParallelism && par != this.defaultParallelism) {
                par = this.defaultParallelism;
                Optimizer.LOG.warn("The parallelism of nested dataflows (such as step functions in iterations) is " + "currently fixed to the parallelism of the surrounding operator (the iteration).");
            }
        } else {
            par = this.defaultParallelism;
        }
        n.setParallelism(par);
    }
    return true;
}
Also used : FlatMapOperatorBase(org.apache.flink.api.common.operators.base.FlatMapOperatorBase) GroupReduceOperatorBase(org.apache.flink.api.common.operators.base.GroupReduceOperatorBase) ReduceOperatorBase(org.apache.flink.api.common.operators.base.ReduceOperatorBase) FilterNode(org.apache.flink.optimizer.dag.FilterNode) CrossOperatorBase(org.apache.flink.api.common.operators.base.CrossOperatorBase) CoGroupOperatorBase(org.apache.flink.api.common.operators.base.CoGroupOperatorBase) BulkPartialSolutionNode(org.apache.flink.optimizer.dag.BulkPartialSolutionNode) DeltaIterationBase(org.apache.flink.api.common.operators.base.DeltaIterationBase) MapPartitionNode(org.apache.flink.optimizer.dag.MapPartitionNode) SortPartitionOperatorBase(org.apache.flink.api.common.operators.base.SortPartitionOperatorBase) MapPartitionOperatorBase(org.apache.flink.api.common.operators.base.MapPartitionOperatorBase) DataSinkNode(org.apache.flink.optimizer.dag.DataSinkNode) JoinNode(org.apache.flink.optimizer.dag.JoinNode) OuterJoinNode(org.apache.flink.optimizer.dag.OuterJoinNode) OuterJoinOperatorBase(org.apache.flink.api.common.operators.base.OuterJoinOperatorBase) FlatMapNode(org.apache.flink.optimizer.dag.FlatMapNode) GroupReduceOperatorBase(org.apache.flink.api.common.operators.base.GroupReduceOperatorBase) OuterJoinNode(org.apache.flink.optimizer.dag.OuterJoinNode) BulkIterationBase(org.apache.flink.api.common.operators.base.BulkIterationBase) CoGroupRawOperatorBase(org.apache.flink.api.common.operators.base.CoGroupRawOperatorBase) GroupReduceNode(org.apache.flink.optimizer.dag.GroupReduceNode) GenericDataSinkBase(org.apache.flink.api.common.operators.GenericDataSinkBase) WorksetNode(org.apache.flink.optimizer.dag.WorksetNode) CoGroupNode(org.apache.flink.optimizer.dag.CoGroupNode) FlatMapNode(org.apache.flink.optimizer.dag.FlatMapNode) MapNode(org.apache.flink.optimizer.dag.MapNode) GroupCombineNode(org.apache.flink.optimizer.dag.GroupCombineNode) Union(org.apache.flink.api.common.operators.Union) FlatMapOperatorBase(org.apache.flink.api.common.operators.base.FlatMapOperatorBase) MapOperatorBase(org.apache.flink.api.common.operators.base.MapOperatorBase) SolutionSetNode(org.apache.flink.optimizer.dag.SolutionSetNode) OptimizerNode(org.apache.flink.optimizer.dag.OptimizerNode) WorksetIterationNode(org.apache.flink.optimizer.dag.WorksetIterationNode) InvalidProgramException(org.apache.flink.api.common.InvalidProgramException) InnerJoinOperatorBase(org.apache.flink.api.common.operators.base.InnerJoinOperatorBase) DataSourceNode(org.apache.flink.optimizer.dag.DataSourceNode) SortPartitionNode(org.apache.flink.optimizer.dag.SortPartitionNode) MapPartitionOperatorBase(org.apache.flink.api.common.operators.base.MapPartitionOperatorBase) PartitionOperatorBase(org.apache.flink.api.common.operators.base.PartitionOperatorBase) SortPartitionOperatorBase(org.apache.flink.api.common.operators.base.SortPartitionOperatorBase) CoGroupRawNode(org.apache.flink.optimizer.dag.CoGroupRawNode) BinaryUnionNode(org.apache.flink.optimizer.dag.BinaryUnionNode) MapPartitionNode(org.apache.flink.optimizer.dag.MapPartitionNode) PartitionNode(org.apache.flink.optimizer.dag.PartitionNode) SortPartitionNode(org.apache.flink.optimizer.dag.SortPartitionNode) FilterOperatorBase(org.apache.flink.api.common.operators.base.FilterOperatorBase) GroupCombineOperatorBase(org.apache.flink.api.common.operators.base.GroupCombineOperatorBase) BulkIterationNode(org.apache.flink.optimizer.dag.BulkIterationNode) ReduceNode(org.apache.flink.optimizer.dag.ReduceNode) GroupReduceNode(org.apache.flink.optimizer.dag.GroupReduceNode) GenericDataSourceBase(org.apache.flink.api.common.operators.GenericDataSourceBase) CrossNode(org.apache.flink.optimizer.dag.CrossNode)

Aggregations

ReduceOperatorBase (org.apache.flink.api.common.operators.base.ReduceOperatorBase)9 Plan (org.apache.flink.api.common.Plan)7 ExecutionEnvironment (org.apache.flink.api.java.ExecutionEnvironment)7 Test (org.junit.Test)6 Tuple3 (org.apache.flink.api.java.tuple.Tuple3)5 RichReduceFunction (org.apache.flink.api.common.functions.RichReduceFunction)2 MapOperatorBase (org.apache.flink.api.common.operators.base.MapOperatorBase)2 InvalidProgramException (org.apache.flink.api.common.InvalidProgramException)1 ReduceFunction (org.apache.flink.api.common.functions.ReduceFunction)1 GenericDataSinkBase (org.apache.flink.api.common.operators.GenericDataSinkBase)1 GenericDataSourceBase (org.apache.flink.api.common.operators.GenericDataSourceBase)1 SelectorFunctionKeys (org.apache.flink.api.common.operators.Keys.SelectorFunctionKeys)1 SingleInputSemanticProperties (org.apache.flink.api.common.operators.SingleInputSemanticProperties)1 UnaryOperatorInformation (org.apache.flink.api.common.operators.UnaryOperatorInformation)1 Union (org.apache.flink.api.common.operators.Union)1 BulkIterationBase (org.apache.flink.api.common.operators.base.BulkIterationBase)1 CoGroupOperatorBase (org.apache.flink.api.common.operators.base.CoGroupOperatorBase)1 CoGroupRawOperatorBase (org.apache.flink.api.common.operators.base.CoGroupRawOperatorBase)1 CrossOperatorBase (org.apache.flink.api.common.operators.base.CrossOperatorBase)1 DeltaIterationBase (org.apache.flink.api.common.operators.base.DeltaIterationBase)1