Search in sources :

Example 1 with ParameterTool

use of org.apache.flink.api.java.utils.ParameterTool 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");
}
Also used : RandomGenerableFactory(org.apache.flink.graph.generator.random.RandomGenerableFactory) DataSet(org.apache.flink.api.java.DataSet) GraphAnalytic(org.apache.flink.graph.GraphAnalytic) JobExecutionResult(org.apache.flink.api.common.JobExecutionResult) NullValue(org.apache.flink.types.NullValue) StringValue(org.apache.flink.types.StringValue) LongValueToUnsignedIntValue(org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue) IntValue(org.apache.flink.types.IntValue) JobExecutionResult(org.apache.flink.api.common.JobExecutionResult) GraphCsvReader(org.apache.flink.graph.GraphCsvReader) RMatGraph(org.apache.flink.graph.generator.RMatGraph) Graph(org.apache.flink.graph.Graph) LongValue(org.apache.flink.types.LongValue) ParameterTool(org.apache.flink.api.java.utils.ParameterTool) LongValueToUnsignedIntValue(org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue) ExecutionEnvironment(org.apache.flink.api.java.ExecutionEnvironment) JDKRandomGeneratorFactory(org.apache.flink.graph.generator.random.JDKRandomGeneratorFactory) TranslateGraphIds(org.apache.flink.graph.asm.translate.TranslateGraphIds) ProgramParametrizationException(org.apache.flink.client.program.ProgramParametrizationException) NumberFormat(java.text.NumberFormat)

Example 2 with ParameterTool

use of org.apache.flink.api.java.utils.ParameterTool in project flink by apache.

the class Graph500 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 directed = parameters.getBoolean("directed");
    if (!parameters.has("simplify")) {
        throw new ProgramParametrizationException(getUsage("must declare '--simplify true' or '--simplify false'"));
    }
    boolean simplify = parameters.getBoolean("simplify");
    // Generate RMat graph
    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;
    boolean clipAndFlip = parameters.getBoolean("clip_and_flip", DEFAULT_CLIP_AND_FLIP);
    Graph<LongValue, NullValue, NullValue> graph = new RMatGraph<>(env, rnd, vertexCount, edgeCount).generate();
    if (directed) {
        if (simplify) {
            graph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<LongValue, NullValue, NullValue>());
        }
    } else {
        if (simplify) {
            graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<LongValue, NullValue, NullValue>(clipAndFlip));
        } else {
            graph = graph.getUndirected();
        }
    }
    DataSet<Tuple2<LongValue, LongValue>> edges = graph.getEdges().project(0, 1);
    // Print, hash, or write RMat graph to disk
    switch(parameters.get("output", "")) {
        case "print":
            System.out.println();
            edges.print();
            break;
        case "hash":
            System.out.println();
            System.out.println(DataSetUtils.checksumHashCode(edges));
            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));
            edges.writeAsCsv(filename, lineDelimiter, fieldDelimiter);
            env.execute("Graph500");
            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");
}
Also used : ParameterTool(org.apache.flink.api.java.utils.ParameterTool) ExecutionEnvironment(org.apache.flink.api.java.ExecutionEnvironment) JDKRandomGeneratorFactory(org.apache.flink.graph.generator.random.JDKRandomGeneratorFactory) JobExecutionResult(org.apache.flink.api.common.JobExecutionResult) NullValue(org.apache.flink.types.NullValue) ProgramParametrizationException(org.apache.flink.client.program.ProgramParametrizationException) Tuple2(org.apache.flink.api.java.tuple.Tuple2) LongValue(org.apache.flink.types.LongValue) JDKRandomGenerator(org.apache.commons.math3.random.JDKRandomGenerator) NumberFormat(java.text.NumberFormat)

Example 3 with ParameterTool

use of org.apache.flink.api.java.utils.ParameterTool 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");
}
Also used : ParameterTool(org.apache.flink.api.java.utils.ParameterTool) LongValueToUnsignedIntValue(org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue) ExecutionEnvironment(org.apache.flink.api.java.ExecutionEnvironment) RandomGenerableFactory(org.apache.flink.graph.generator.random.RandomGenerableFactory) DataSet(org.apache.flink.api.java.DataSet) JDKRandomGeneratorFactory(org.apache.flink.graph.generator.random.JDKRandomGeneratorFactory) NullValue(org.apache.flink.types.NullValue) LongValueToUnsignedIntValue(org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue) IntValue(org.apache.flink.types.IntValue) Simplify(org.apache.flink.graph.asm.simple.directed.Simplify) JobExecutionResult(org.apache.flink.api.common.JobExecutionResult) GraphCsvReader(org.apache.flink.graph.GraphCsvReader) RMatGraph(org.apache.flink.graph.generator.RMatGraph) RMatGraph(org.apache.flink.graph.generator.RMatGraph) Graph(org.apache.flink.graph.Graph) ProgramParametrizationException(org.apache.flink.client.program.ProgramParametrizationException) LongValue(org.apache.flink.types.LongValue) NumberFormat(java.text.NumberFormat)

Example 4 with ParameterTool

use of org.apache.flink.api.java.utils.ParameterTool 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");
}
Also used : ParameterTool(org.apache.flink.api.java.utils.ParameterTool) LongValueToUnsignedIntValue(org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue) ExecutionEnvironment(org.apache.flink.api.java.ExecutionEnvironment) RandomGenerableFactory(org.apache.flink.graph.generator.random.RandomGenerableFactory) DataSet(org.apache.flink.api.java.DataSet) JDKRandomGeneratorFactory(org.apache.flink.graph.generator.random.JDKRandomGeneratorFactory) JobExecutionResult(org.apache.flink.api.common.JobExecutionResult) Result(org.apache.flink.graph.library.similarity.JaccardIndex.Result) NullValue(org.apache.flink.types.NullValue) StringValue(org.apache.flink.types.StringValue) LongValueToUnsignedIntValue(org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue) IntValue(org.apache.flink.types.IntValue) TranslateGraphIds(org.apache.flink.graph.asm.translate.TranslateGraphIds) JobExecutionResult(org.apache.flink.api.common.JobExecutionResult) GraphCsvReader(org.apache.flink.graph.GraphCsvReader) RMatGraph(org.apache.flink.graph.generator.RMatGraph) Graph(org.apache.flink.graph.Graph) ProgramParametrizationException(org.apache.flink.client.program.ProgramParametrizationException) LongValue(org.apache.flink.types.LongValue) NumberFormat(java.text.NumberFormat)

Example 5 with ParameterTool

use of org.apache.flink.api.java.utils.ParameterTool in project flink by apache.

the class ManualExactlyOnceTest method main.

public static void main(String[] args) throws Exception {
    final ParameterTool pt = ParameterTool.fromArgs(args);
    LOG.info("Starting exactly once test");
    final String streamName = "flink-test-" + UUID.randomUUID().toString();
    final String accessKey = pt.getRequired("accessKey");
    final String secretKey = pt.getRequired("secretKey");
    final String region = pt.getRequired("region");
    Properties configProps = new Properties();
    configProps.setProperty(AWSConfigConstants.AWS_ACCESS_KEY_ID, accessKey);
    configProps.setProperty(AWSConfigConstants.AWS_SECRET_ACCESS_KEY, secretKey);
    configProps.setProperty(AWSConfigConstants.AWS_REGION, region);
    AmazonKinesisClient client = AWSUtil.createKinesisClient(configProps);
    // create a stream for the test:
    client.createStream(streamName, 1);
    // wait until stream has been created
    DescribeStreamResult status = client.describeStream(streamName);
    LOG.info("status {}", status);
    while (!status.getStreamDescription().getStreamStatus().equals("ACTIVE")) {
        status = client.describeStream(streamName);
        LOG.info("Status of stream {}", status);
        Thread.sleep(1000);
    }
    final Configuration flinkConfig = new Configuration();
    flinkConfig.setInteger(ConfigConstants.LOCAL_NUMBER_TASK_MANAGER, 1);
    flinkConfig.setInteger(ConfigConstants.TASK_MANAGER_NUM_TASK_SLOTS, 8);
    flinkConfig.setInteger(ConfigConstants.TASK_MANAGER_MEMORY_SIZE_KEY, 16);
    flinkConfig.setString(ConfigConstants.RESTART_STRATEGY_FIXED_DELAY_DELAY, "0 s");
    LocalFlinkMiniCluster flink = new LocalFlinkMiniCluster(flinkConfig, false);
    flink.start();
    final int flinkPort = flink.getLeaderRPCPort();
    try {
        final AtomicReference<Throwable> producerError = new AtomicReference<>();
        Thread producerThread = KinesisEventsGeneratorProducerThread.create(TOTAL_EVENT_COUNT, 2, accessKey, secretKey, region, streamName, producerError, flinkPort, flinkConfig);
        producerThread.start();
        final AtomicReference<Throwable> consumerError = new AtomicReference<>();
        Thread consumerThread = ExactlyOnceValidatingConsumerThread.create(TOTAL_EVENT_COUNT, 200, 2, 500, 500, accessKey, secretKey, region, streamName, consumerError, flinkPort, flinkConfig);
        consumerThread.start();
        boolean deadlinePassed = false;
        // wait at most for two minutes
        long deadline = System.currentTimeMillis() + (1000 * 2 * 60);
        // wait until both producer and consumer finishes, or an unexpected error is thrown
        while ((consumerThread.isAlive() || producerThread.isAlive()) && (producerError.get() == null && consumerError.get() == null)) {
            Thread.sleep(1000);
            if (System.currentTimeMillis() >= deadline) {
                LOG.warn("Deadline passed");
                deadlinePassed = true;
                // enough waiting
                break;
            }
        }
        if (producerThread.isAlive()) {
            producerThread.interrupt();
        }
        if (consumerThread.isAlive()) {
            consumerThread.interrupt();
        }
        if (producerError.get() != null) {
            LOG.info("+++ TEST failed! +++");
            throw new RuntimeException("Producer failed", producerError.get());
        }
        if (consumerError.get() != null) {
            LOG.info("+++ TEST failed! +++");
            throw new RuntimeException("Consumer failed", consumerError.get());
        }
        if (!deadlinePassed) {
            LOG.info("+++ TEST passed! +++");
        } else {
            LOG.info("+++ TEST failed! +++");
        }
    } finally {
        client.deleteStream(streamName);
        client.shutdown();
        // stopping flink
        flink.stop();
    }
}
Also used : ParameterTool(org.apache.flink.api.java.utils.ParameterTool) AmazonKinesisClient(com.amazonaws.services.kinesis.AmazonKinesisClient) Configuration(org.apache.flink.configuration.Configuration) AtomicReference(java.util.concurrent.atomic.AtomicReference) Properties(java.util.Properties) LocalFlinkMiniCluster(org.apache.flink.runtime.minicluster.LocalFlinkMiniCluster) KinesisEventsGeneratorProducerThread(org.apache.flink.streaming.connectors.kinesis.testutils.KinesisEventsGeneratorProducerThread) ExactlyOnceValidatingConsumerThread(org.apache.flink.streaming.connectors.kinesis.testutils.ExactlyOnceValidatingConsumerThread) DescribeStreamResult(com.amazonaws.services.kinesis.model.DescribeStreamResult)

Aggregations

ParameterTool (org.apache.flink.api.java.utils.ParameterTool)43 ExecutionEnvironment (org.apache.flink.api.java.ExecutionEnvironment)19 StreamExecutionEnvironment (org.apache.flink.streaming.api.environment.StreamExecutionEnvironment)19 Tuple2 (org.apache.flink.api.java.tuple.Tuple2)15 JobExecutionResult (org.apache.flink.api.common.JobExecutionResult)7 NumberFormat (java.text.NumberFormat)6 Properties (java.util.Properties)6 ProgramParametrizationException (org.apache.flink.client.program.ProgramParametrizationException)6 JDKRandomGeneratorFactory (org.apache.flink.graph.generator.random.JDKRandomGeneratorFactory)6 LongValue (org.apache.flink.types.LongValue)6 NullValue (org.apache.flink.types.NullValue)6 Graph (org.apache.flink.graph.Graph)5 GraphCsvReader (org.apache.flink.graph.GraphCsvReader)5 LongValueToUnsignedIntValue (org.apache.flink.graph.asm.translate.translators.LongValueToUnsignedIntValue)5 RMatGraph (org.apache.flink.graph.generator.RMatGraph)5 RandomGenerableFactory (org.apache.flink.graph.generator.random.RandomGenerableFactory)5 SimpleStringSchema (org.apache.flink.streaming.util.serialization.SimpleStringSchema)5 IntValue (org.apache.flink.types.IntValue)5 StringValue (org.apache.flink.types.StringValue)4 DataSet (org.apache.flink.api.java.DataSet)3