/***********************************************************************************************************************
* Copyright (C) 2010-2013 by the Stratosphere project (http://stratosphere.eu)
*
* Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
* an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
* specific language governing permissions and limitations under the License.
**********************************************************************************************************************/
package eu.stratosphere.test.recordJobs.graph;
import java.util.Iterator;
import eu.stratosphere.api.common.Plan;
import eu.stratosphere.api.common.Program;
import eu.stratosphere.api.common.ProgramDescription;
import eu.stratosphere.api.java.record.operators.DeltaIteration;
import eu.stratosphere.api.java.record.operators.FileDataSink;
import eu.stratosphere.api.java.record.operators.FileDataSource;
import eu.stratosphere.api.java.record.functions.FunctionAnnotation.ConstantFields;
import eu.stratosphere.api.java.record.functions.FunctionAnnotation.ConstantFieldsSecond;
import eu.stratosphere.api.java.record.functions.JoinFunction;
import eu.stratosphere.api.java.record.functions.ReduceFunction;
import eu.stratosphere.api.java.record.io.CsvInputFormat;
import eu.stratosphere.api.java.record.io.CsvOutputFormat;
import eu.stratosphere.api.java.record.operators.JoinOperator;
import eu.stratosphere.api.java.record.operators.ReduceOperator;
import eu.stratosphere.api.java.record.operators.ReduceOperator.Combinable;
import eu.stratosphere.types.DoubleValue;
import eu.stratosphere.types.LongValue;
import eu.stratosphere.types.Record;
import eu.stratosphere.util.Collector;
public class DeltaPageRankWithInitialDeltas implements Program, ProgramDescription {
private static final long serialVersionUID = 1L;
@ConstantFieldsSecond(0)
public static final class RankComparisonMatch extends JoinFunction {
private static final long serialVersionUID = 1L;
private final DoubleValue newRank = new DoubleValue();
@Override
public void join(Record vertexWithDelta, Record vertexWithOldRank, Collector<Record> out) throws Exception {
DoubleValue deltaVal = vertexWithDelta.getField(1, DoubleValue.class);
DoubleValue currentVal = vertexWithOldRank.getField(1, DoubleValue.class);
newRank.setValue(deltaVal.getValue() + currentVal.getValue());
vertexWithOldRank.setField(1, newRank);
out.collect(vertexWithOldRank);
}
}
@Combinable
@ConstantFields(0)
public static final class UpdateRankReduceDelta extends ReduceFunction {
private static final long serialVersionUID = 1L;
private final DoubleValue newRank = new DoubleValue();
@Override
public void reduce(Iterator<Record> records, Collector<Record> out) {
double rankSum = 0.0;
double rank;
Record rec = null;
while (records.hasNext()) {
rec = records.next();
rank = rec.getField(1, DoubleValue.class).getValue();
rankSum += rank;
}
// ignore small deltas
if (Math.abs(rankSum) > 0.00001) {
newRank.setValue(rankSum);
rec.setField(1, newRank);
out.collect(rec);
}
}
}
public class PRDependenciesComputationMatchDelta extends JoinFunction {
private static final long serialVersionUID = 1L;
private final Record result = new Record();
private final DoubleValue partRank = new DoubleValue();
/*
* (srcId, trgId, weight) x (vId, rank) => (trgId, rank / weight)
*/
@Override
public void join(Record vertexWithRank, Record edgeWithWeight, Collector<Record> out) throws Exception {
result.setField(0, edgeWithWeight.getField(1, LongValue.class));
final long outLinks = edgeWithWeight.getField(2, LongValue.class).getValue();
final double rank = vertexWithRank.getField(1, DoubleValue.class).getValue();
partRank.setValue(rank / (double) outLinks);
result.setField(1, partRank);
out.collect(result);
}
}
@SuppressWarnings("unchecked")
@Override
public Plan getPlan(String... args) {
// parse job parameters
final int numSubTasks = (args.length > 0 ? Integer.parseInt(args[0]) : 1);
final String solutionSetInput = (args.length > 1 ? args[1] : "");
final String deltasInput = (args.length > 2 ? args[2] : "");
final String dependencySetInput = (args.length > 3 ? args[3] : "");
final String output = (args.length > 4 ? args[4] : "");
final int maxIterations = (args.length > 5 ? Integer.parseInt(args[5]) : 1);
// create DataSourceContract for the initalSolutionSet
FileDataSource initialSolutionSet = new FileDataSource(new CsvInputFormat(' ', LongValue.class, DoubleValue.class), solutionSetInput, "Initial Solution Set");
// create DataSourceContract for the initalDeltaSet
FileDataSource initialDeltaSet = new FileDataSource(new CsvInputFormat(' ', LongValue.class, DoubleValue.class), deltasInput, "Initial DeltaSet");
// create DataSourceContract for the edges
FileDataSource dependencySet = new FileDataSource(new CsvInputFormat(' ', LongValue.class, LongValue.class, LongValue.class), dependencySetInput, "Dependency Set");
DeltaIteration iteration = new DeltaIteration(0, "Delta PageRank");
iteration.setInitialSolutionSet(initialSolutionSet);
iteration.setInitialWorkset(initialDeltaSet);
iteration.setMaximumNumberOfIterations(maxIterations);
JoinOperator dependenciesMatch = JoinOperator.builder(PRDependenciesComputationMatchDelta.class,
LongValue.class, 0, 0)
.input1(iteration.getWorkset())
.input2(dependencySet)
.name("calculate dependencies")
.build();
ReduceOperator updateRanks = ReduceOperator.builder(UpdateRankReduceDelta.class, LongValue.class, 0)
.input(dependenciesMatch)
.name("update ranks")
.build();
JoinOperator oldRankComparison = JoinOperator.builder(RankComparisonMatch.class, LongValue.class, 0, 0)
.input1(updateRanks)
.input2(iteration.getSolutionSet())
.name("comparison with old ranks")
.build();
iteration.setNextWorkset(updateRanks);
iteration.setSolutionSetDelta(oldRankComparison);
// create DataSinkContract for writing the final ranks
FileDataSink result = new FileDataSink(CsvOutputFormat.class, output, iteration, "Final Ranks");
CsvOutputFormat.configureRecordFormat(result)
.recordDelimiter('\n')
.fieldDelimiter(' ')
.field(LongValue.class, 0)
.field(DoubleValue.class, 1);
// return the PACT plan
Plan plan = new Plan(result, "Delta PageRank");
plan.setDefaultParallelism(numSubTasks);
return plan;
}
@Override
public String getDescription() {
return "Parameters: <numberOfSubTasks> <initialSolutionSet(pageId, rank)> <deltas(pageId, delta)> <dependencySet(srcId, trgId, out_links)> <out> <maxIterations>";
}
}