Examples with DataSet org.apache.flink.api.java.DataSet used on opensource projects

Example 1 with DataSet

use of org.apache.flink.api.java.DataSet in project flink by apache.

the class ClusteringCoefficient method main.

public static void main(String[] args) throws Exception {
    // Set up the execution environment
    final ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
    env.getConfig().enableObjectReuse();
    ParameterTool parameters = ParameterTool.fromArgs(args);
    env.getConfig().setGlobalJobParameters(parameters);
    if (!parameters.has("directed")) {
        throw new ProgramParametrizationException(getUsage("must declare execution mode as '--directed true' or '--directed false'"));
    }
    boolean directedAlgorithm = parameters.getBoolean("directed");
    int little_parallelism = parameters.getInt("little_parallelism", PARALLELISM_DEFAULT);
    // global and local clustering coefficient results
    GraphAnalytic gcc;
    GraphAnalytic acc;
    DataSet lcc;
    switch(parameters.get("input", "")) {
        case "csv":
            {
                String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
                String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
                GraphCsvReader reader = Graph.fromCsvReader(parameters.get("input_filename"), env).ignoreCommentsEdges("#").lineDelimiterEdges(lineDelimiter).fieldDelimiterEdges(fieldDelimiter);
                switch(parameters.get("type", "")) {
                    case "integer":
                        {
                            Graph<LongValue, NullValue, NullValue> graph = reader.keyType(LongValue.class);
                            if (directedAlgorithm) {
                                if (parameters.getBoolean("simplify", false)) {
                                    graph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<LongValue, NullValue, NullValue>().setParallelism(little_parallelism));
                                }
                                gcc = graph.run(new org.apache.flink.graph.library.clustering.directed.GlobalClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                acc = graph.run(new org.apache.flink.graph.library.clustering.directed.AverageClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                lcc = graph.run(new org.apache.flink.graph.library.clustering.directed.LocalClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                            } else {
                                if (parameters.getBoolean("simplify", false)) {
                                    graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<LongValue, NullValue, NullValue>(false).setParallelism(little_parallelism));
                                }
                                gcc = graph.run(new org.apache.flink.graph.library.clustering.undirected.GlobalClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                acc = graph.run(new org.apache.flink.graph.library.clustering.undirected.AverageClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                lcc = graph.run(new org.apache.flink.graph.library.clustering.undirected.LocalClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                            }
                        }
                        break;
                    case "string":
                        {
                            Graph<StringValue, NullValue, NullValue> graph = reader.keyType(StringValue.class);
                            if (directedAlgorithm) {
                                if (parameters.getBoolean("simplify", false)) {
                                    graph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<StringValue, NullValue, NullValue>().setParallelism(little_parallelism));
                                }
                                gcc = graph.run(new org.apache.flink.graph.library.clustering.directed.GlobalClusteringCoefficient<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                acc = graph.run(new org.apache.flink.graph.library.clustering.directed.AverageClusteringCoefficient<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                lcc = graph.run(new org.apache.flink.graph.library.clustering.directed.LocalClusteringCoefficient<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                            } else {
                                if (parameters.getBoolean("simplify", false)) {
                                    graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<StringValue, NullValue, NullValue>(false).setParallelism(little_parallelism));
                                }
                                gcc = graph.run(new org.apache.flink.graph.library.clustering.undirected.GlobalClusteringCoefficient<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                acc = graph.run(new org.apache.flink.graph.library.clustering.undirected.AverageClusteringCoefficient<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                                lcc = graph.run(new org.apache.flink.graph.library.clustering.undirected.LocalClusteringCoefficient<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                            }
                        }
                        break;
                    default:
                        throw new ProgramParametrizationException(getUsage("invalid CSV type"));
                }
            }
            break;
        case "rmat":
            {
                int scale = parameters.getInt("scale", DEFAULT_SCALE);
                int edgeFactor = parameters.getInt("edge_factor", DEFAULT_EDGE_FACTOR);
                RandomGenerableFactory<JDKRandomGenerator> rnd = new JDKRandomGeneratorFactory();
                long vertexCount = 1L << scale;
                long edgeCount = vertexCount * edgeFactor;
                Graph<LongValue, NullValue, NullValue> graph = new RMatGraph<>(env, rnd, vertexCount, edgeCount).setParallelism(little_parallelism).generate();
                if (directedAlgorithm) {
                    if (scale > 32) {
                        Graph<LongValue, NullValue, NullValue> newGraph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<LongValue, NullValue, NullValue>().setParallelism(little_parallelism));
                        gcc = newGraph.run(new org.apache.flink.graph.library.clustering.directed.GlobalClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        acc = newGraph.run(new org.apache.flink.graph.library.clustering.directed.AverageClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        lcc = newGraph.run(new org.apache.flink.graph.library.clustering.directed.LocalClusteringCoefficient<LongValue, NullValue, NullValue>().setIncludeZeroDegreeVertices(false).setLittleParallelism(little_parallelism));
                    } else {
                        Graph<IntValue, NullValue, NullValue> newGraph = graph.run(new TranslateGraphIds<LongValue, IntValue, NullValue, NullValue>(new LongValueToUnsignedIntValue()).setParallelism(little_parallelism)).run(new org.apache.flink.graph.asm.simple.directed.Simplify<IntValue, NullValue, NullValue>().setParallelism(little_parallelism));
                        gcc = newGraph.run(new org.apache.flink.graph.library.clustering.directed.GlobalClusteringCoefficient<IntValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        acc = newGraph.run(new org.apache.flink.graph.library.clustering.directed.AverageClusteringCoefficient<IntValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        lcc = newGraph.run(new org.apache.flink.graph.library.clustering.directed.LocalClusteringCoefficient<IntValue, NullValue, NullValue>().setIncludeZeroDegreeVertices(false).setLittleParallelism(little_parallelism));
                    }
                } else {
                    boolean clipAndFlip = parameters.getBoolean("clip_and_flip", DEFAULT_CLIP_AND_FLIP);
                    if (scale > 32) {
                        Graph<LongValue, NullValue, NullValue> newGraph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<LongValue, NullValue, NullValue>(clipAndFlip).setParallelism(little_parallelism));
                        gcc = newGraph.run(new org.apache.flink.graph.library.clustering.undirected.GlobalClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        acc = newGraph.run(new org.apache.flink.graph.library.clustering.undirected.AverageClusteringCoefficient<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        lcc = newGraph.run(new org.apache.flink.graph.library.clustering.undirected.LocalClusteringCoefficient<LongValue, NullValue, NullValue>().setIncludeZeroDegreeVertices(false).setLittleParallelism(little_parallelism));
                    } else {
                        Graph<IntValue, NullValue, NullValue> newGraph = graph.run(new TranslateGraphIds<LongValue, IntValue, NullValue, NullValue>(new LongValueToUnsignedIntValue()).setParallelism(little_parallelism)).run(new org.apache.flink.graph.asm.simple.undirected.Simplify<IntValue, NullValue, NullValue>(clipAndFlip).setParallelism(little_parallelism));
                        gcc = newGraph.run(new org.apache.flink.graph.library.clustering.undirected.GlobalClusteringCoefficient<IntValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        acc = newGraph.run(new org.apache.flink.graph.library.clustering.undirected.AverageClusteringCoefficient<IntValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        lcc = newGraph.run(new org.apache.flink.graph.library.clustering.undirected.LocalClusteringCoefficient<IntValue, NullValue, NullValue>().setIncludeZeroDegreeVertices(false).setLittleParallelism(little_parallelism));
                    }
                }
            }
            break;
        default:
            throw new ProgramParametrizationException(getUsage("invalid input type"));
    }
    switch(parameters.get("output", "")) {
        case "print":
            if (directedAlgorithm) {
                for (Object e : lcc.collect()) {
                    org.apache.flink.graph.library.clustering.directed.LocalClusteringCoefficient.Result result = (org.apache.flink.graph.library.clustering.directed.LocalClusteringCoefficient.Result) e;
                    System.out.println(result.toPrintableString());
                }
            } else {
                for (Object e : lcc.collect()) {
                    org.apache.flink.graph.library.clustering.undirected.LocalClusteringCoefficient.Result result = (org.apache.flink.graph.library.clustering.undirected.LocalClusteringCoefficient.Result) e;
                    System.out.println(result.toPrintableString());
                }
            }
            break;
        case "hash":
            System.out.println(DataSetUtils.checksumHashCode(lcc));
            break;
        case "csv":
            String filename = parameters.get("output_filename");
            String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("output_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
            String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("output_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
            lcc.writeAsCsv(filename, lineDelimiter, fieldDelimiter);
            env.execute("Clustering Coefficient");
            break;
        default:
            throw new ProgramParametrizationException(getUsage("invalid output type"));
    }
    System.out.println(gcc.getResult());
    System.out.println(acc.getResult());
    JobExecutionResult result = env.getLastJobExecutionResult();
    NumberFormat nf = NumberFormat.getInstance();
    System.out.println("Execution runtime: " + nf.format(result.getNetRuntime()) + " ms");
}

Example 2 with DataSet

use of org.apache.flink.api.java.DataSet in project flink by apache.

the class HITS method main.

public static void main(String[] args) throws Exception {
    // Set up the execution environment
    final ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
    env.getConfig().enableObjectReuse();
    ParameterTool parameters = ParameterTool.fromArgs(args);
    env.getConfig().setGlobalJobParameters(parameters);
    int iterations = parameters.getInt("iterations", DEFAULT_ITERATIONS);
    DataSet hits;
    switch(parameters.get("input", "")) {
        case "csv":
            {
                String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
                String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
                GraphCsvReader reader = Graph.fromCsvReader(parameters.getRequired("input_filename"), env).ignoreCommentsEdges("#").lineDelimiterEdges(lineDelimiter).fieldDelimiterEdges(fieldDelimiter);
                switch(parameters.get("type", "")) {
                    case "integer":
                        {
                            hits = reader.keyType(LongValue.class).run(new org.apache.flink.graph.library.link_analysis.HITS<LongValue, NullValue, NullValue>(iterations));
                        }
                        break;
                    case "string":
                        {
                            hits = reader.keyType(StringValue.class).run(new org.apache.flink.graph.library.link_analysis.HITS<StringValue, NullValue, NullValue>(iterations));
                        }
                        break;
                    default:
                        throw new ProgramParametrizationException(getUsage("invalid CSV type"));
                }
            }
            break;
        case "rmat":
            {
                int scale = parameters.getInt("scale", DEFAULT_SCALE);
                int edgeFactor = parameters.getInt("edge_factor", DEFAULT_EDGE_FACTOR);
                RandomGenerableFactory<JDKRandomGenerator> rnd = new JDKRandomGeneratorFactory();
                long vertexCount = 1L << scale;
                long edgeCount = vertexCount * edgeFactor;
                Graph<LongValue, NullValue, NullValue> graph = new RMatGraph<>(env, rnd, vertexCount, edgeCount).generate();
                if (scale > 32) {
                    hits = graph.run(new Simplify<LongValue, NullValue, NullValue>()).run(new org.apache.flink.graph.library.link_analysis.HITS<LongValue, NullValue, NullValue>(iterations));
                } else {
                    hits = graph.run(new TranslateGraphIds<LongValue, IntValue, NullValue, NullValue>(new LongValueToUnsignedIntValue())).run(new Simplify<IntValue, NullValue, NullValue>()).run(new org.apache.flink.graph.library.link_analysis.HITS<IntValue, NullValue, NullValue>(iterations));
                }
            }
            break;
        default:
            throw new ProgramParametrizationException(getUsage("invalid input type"));
    }
    switch(parameters.get("output", "")) {
        case "print":
            System.out.println();
            for (Object e : hits.collect()) {
                System.out.println(((Result) e).toPrintableString());
            }
            break;
        case "hash":
            System.out.println();
            System.out.println(DataSetUtils.checksumHashCode(hits));
            break;
        case "csv":
            String filename = parameters.getRequired("output_filename");
            String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("output_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
            String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("output_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
            hits.writeAsCsv(filename, lineDelimiter, fieldDelimiter);
            env.execute("HITS");
            break;
        default:
            throw new ProgramParametrizationException(getUsage("invalid output type"));
    }
    JobExecutionResult result = env.getLastJobExecutionResult();
    NumberFormat nf = NumberFormat.getInstance();
    System.out.println();
    System.out.println("Execution runtime: " + nf.format(result.getNetRuntime()) + " ms");
}

Example 3 with DataSet

use of org.apache.flink.api.java.DataSet in project flink by apache.

the class JaccardIndex method main.

public static void main(String[] args) throws Exception {
    // Set up the execution environment
    final ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
    env.getConfig().enableObjectReuse();
    ParameterTool parameters = ParameterTool.fromArgs(args);
    env.getConfig().setGlobalJobParameters(parameters);
    int little_parallelism = parameters.getInt("little_parallelism", PARALLELISM_DEFAULT);
    DataSet ji;
    switch(parameters.get("input", "")) {
        case "csv":
            {
                String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
                String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
                GraphCsvReader reader = Graph.fromCsvReader(parameters.getRequired("input_filename"), env).ignoreCommentsEdges("#").lineDelimiterEdges(lineDelimiter).fieldDelimiterEdges(fieldDelimiter);
                switch(parameters.get("type", "")) {
                    case "integer":
                        {
                            Graph<LongValue, NullValue, NullValue> graph = reader.keyType(LongValue.class);
                            if (parameters.getBoolean("simplify", false)) {
                                graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<LongValue, NullValue, NullValue>(false).setParallelism(little_parallelism));
                            }
                            ji = graph.run(new org.apache.flink.graph.library.similarity.JaccardIndex<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        }
                        break;
                    case "string":
                        {
                            Graph<StringValue, NullValue, NullValue> graph = reader.keyType(StringValue.class);
                            if (parameters.getBoolean("simplify", false)) {
                                graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<StringValue, NullValue, NullValue>(false).setParallelism(little_parallelism));
                            }
                            ji = graph.run(new org.apache.flink.graph.library.similarity.JaccardIndex<StringValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                        }
                        break;
                    default:
                        throw new ProgramParametrizationException(getUsage("invalid CSV type"));
                }
            }
            break;
        case "rmat":
            {
                int scale = parameters.getInt("scale", DEFAULT_SCALE);
                int edgeFactor = parameters.getInt("edge_factor", DEFAULT_EDGE_FACTOR);
                RandomGenerableFactory<JDKRandomGenerator> rnd = new JDKRandomGeneratorFactory();
                long vertexCount = 1L << scale;
                long edgeCount = vertexCount * edgeFactor;
                Graph<LongValue, NullValue, NullValue> graph = new RMatGraph<>(env, rnd, vertexCount, edgeCount).setParallelism(little_parallelism).generate();
                boolean clipAndFlip = parameters.getBoolean("clip_and_flip", DEFAULT_CLIP_AND_FLIP);
                if (scale > 32) {
                    ji = graph.run(new Simplify<LongValue, NullValue, NullValue>(clipAndFlip).setParallelism(little_parallelism)).run(new org.apache.flink.graph.library.similarity.JaccardIndex<LongValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                } else {
                    ji = graph.run(new TranslateGraphIds<LongValue, IntValue, NullValue, NullValue>(new LongValueToUnsignedIntValue()).setParallelism(little_parallelism)).run(new Simplify<IntValue, NullValue, NullValue>(clipAndFlip).setParallelism(little_parallelism)).run(new org.apache.flink.graph.library.similarity.JaccardIndex<IntValue, NullValue, NullValue>().setLittleParallelism(little_parallelism));
                }
            }
            break;
        default:
            throw new ProgramParametrizationException(getUsage("invalid input type"));
    }
    switch(parameters.get("output", "")) {
        case "print":
            System.out.println();
            for (Object e : ji.collect()) {
                Result result = (Result) e;
                System.out.println(result.toPrintableString());
            }
            break;
        case "hash":
            System.out.println();
            System.out.println(DataSetUtils.checksumHashCode(ji));
            break;
        case "csv":
            String filename = parameters.getRequired("output_filename");
            String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("output_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
            String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("output_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
            ji.writeAsCsv(filename, lineDelimiter, fieldDelimiter);
            env.execute("Jaccard Index");
            break;
        default:
            throw new ProgramParametrizationException(getUsage("invalid output type"));
    }
    JobExecutionResult result = env.getLastJobExecutionResult();
    NumberFormat nf = NumberFormat.getInstance();
    System.out.println();
    System.out.println("Execution runtime: " + nf.format(result.getNetRuntime()) + " ms");
}

Example 4 with DataSet

use of org.apache.flink.api.java.DataSet in project flink by apache.

the class TPCHQuery10 method main.

// *************************************************************************
//     PROGRAM
// *************************************************************************
public static void main(String[] args) throws Exception {
    if (!parseParameters(args)) {
        return;
    }
    final ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
    // get customer data set: (custkey, name, address, nationkey, acctbal) 
    DataSet<Tuple5<Integer, String, String, Integer, Double>> customers = getCustomerDataSet(env);
    // get orders data set: (orderkey, custkey, orderdate)
    DataSet<Tuple3<Integer, Integer, String>> orders = getOrdersDataSet(env);
    // get lineitem data set: (orderkey, extendedprice, discount, returnflag)
    DataSet<Tuple4<Integer, Double, Double, String>> lineitems = getLineitemDataSet(env);
    // get nation data set: (nationkey, name)
    DataSet<Tuple2<Integer, String>> nations = getNationsDataSet(env);
    // orders filtered by year: (orderkey, custkey)
    DataSet<Tuple2<Integer, Integer>> ordersFilteredByYear = // filter by year
    orders.filter(order -> Integer.parseInt(order.f2.substring(0, 4)) > 1990).project(0, 1);
    // lineitems filtered by flag: (orderkey, extendedprice, discount)
    DataSet<Tuple3<Integer, Double, Double>> lineitemsFilteredByFlag = // filter by flag
    lineitems.filter(lineitem -> lineitem.f3.equals("R")).project(0, 1, 2);
    // join orders with lineitems: (custkey, extendedprice, discount)
    DataSet<Tuple3<Integer, Double, Double>> lineitemsOfCustomerKey = ordersFilteredByYear.joinWithHuge(lineitemsFilteredByFlag).where(0).equalTo(0).projectFirst(1).projectSecond(1, 2);
    // aggregate for revenue: (custkey, revenue)
    DataSet<Tuple2<Integer, Double>> revenueOfCustomerKey = lineitemsOfCustomerKey.map(i -> new Tuple2<>(i.f0, i.f1 * (1 - i.f2))).groupBy(0).sum(1);
    // join customer with nation (custkey, name, address, nationname, acctbal)
    DataSet<Tuple5<Integer, String, String, String, Double>> customerWithNation = customers.joinWithTiny(nations).where(3).equalTo(0).projectFirst(0, 1, 2).projectSecond(1).projectFirst(4);
    // join customer (with nation) with revenue (custkey, name, address, nationname, acctbal, revenue)
    DataSet<Tuple6<Integer, String, String, String, Double, Double>> customerWithRevenue = customerWithNation.join(revenueOfCustomerKey).where(0).equalTo(0).projectFirst(0, 1, 2, 3, 4).projectSecond(1);
    // emit result
    customerWithRevenue.writeAsCsv(outputPath);
    // execute program
    env.execute("TPCH Query 10 Example");
}

Example 5 with DataSet

use of org.apache.flink.api.java.DataSet in project flink by apache.

the class FilterLambda1 method main.

public static void main(String[] args) throws Exception {
    ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
    DataSet<String> input = env.fromElements("Please filter", "the words", "but not this");
    FilterFunction<String> filter = (v) -> WordFilter.filter(v);
    DataSet<String> output = input.filter(filter);
    output.print();
    env.execute();
}