Example usage for org.apache.mahout.math CardinalityException CardinalityException

List of usage examples for org.apache.mahout.math CardinalityException CardinalityException

  1. HOME
  2. Java
  3. org.apache.mahout
  4. org.apache.mahout.math.*
  5. CardinalityException
  6. CardinalityException

Introduction

In this page you can find the example usage for org.apache.mahout.math CardinalityException CardinalityException.

Prototype

public CardinalityException(int expected, int cardinality)

Source Link

Usage

From source file:at.illecker.hadoop.rootbeer.examples.matrixmultiplication.cpu.MatrixMultiplicationCpu.java

License:Apache License

@Override
public int run(String[] strings) throws Exception {
    addOption("numRowsA", "nra", "Number of rows of the first input matrix", true);
    addOption("numColsA", "nca", "Number of columns of the first input matrix", true);
    addOption("numRowsB", "nrb", "Number of rows of the second input matrix", true);
    addOption("numColsB", "ncb", "Number of columns of the second input matrix", true);
    addOption("debug", "db", "Enable debugging (true|false)", false);

    Map<String, List<String>> argMap = parseArguments(strings);
    if (argMap == null) {
        return -1;
    }//from   www. j av a  2  s.  co  m

    int numRowsA = Integer.parseInt(getOption("numRowsA"));
    int numColsA = Integer.parseInt(getOption("numColsA"));
    int numRowsB = Integer.parseInt(getOption("numRowsB"));
    int numColsB = Integer.parseInt(getOption("numColsB"));
    boolean isDebugging = Boolean.parseBoolean(getOption("debug"));

    LOG.info("numRowsA: " + numRowsA);
    LOG.info("numColsA: " + numColsA);
    LOG.info("numRowsB: " + numRowsB);
    LOG.info("numColsB: " + numColsB);
    LOG.info("isDebugging: " + isDebugging);
    LOG.info("outputPath: " + OUTPUT_DIR);

    if (numColsA != numRowsB) {
        throw new CardinalityException(numColsA, numRowsB);
    }

    Configuration conf = new Configuration(getConf());

    // Create random DistributedRowMatrix
    // use constant seeds to get reproducable results
    // Matrix A is stored transposed
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_TRANSPOSED_PATH, true);
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsB, numColsB, new Random(1337L),
            MATRIX_B_PATH, false);

    // Load DistributedRowMatrix a and b
    DistributedRowMatrix aTransposed = new DistributedRowMatrix(MATRIX_A_TRANSPOSED_PATH, OUTPUT_DIR, numRowsA,
            numColsA);
    aTransposed.setConf(conf);

    DistributedRowMatrix b = new DistributedRowMatrix(MATRIX_B_PATH, OUTPUT_DIR, numRowsB, numColsB);
    b.setConf(conf);

    // MatrixMultiply all within a new MapReduce job
    long startTime = System.currentTimeMillis();
    DistributedRowMatrix c = aTransposed.multiplyMapReduce(b, MATRIX_C_PATH, false, true, 0, isDebugging);
    System.out.println("MatrixMultiplicationCpu using Hadoop finished in "
            + (System.currentTimeMillis() - startTime) / 1000.0 + " seconds");

    // Verification
    // Overwrite matrix A, NOT transposed for verification check
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_PATH, false);
    DistributedRowMatrix a = new DistributedRowMatrix(MATRIX_A_PATH, OUTPUT_DIR, numRowsA, numColsA);
    a.setConf(conf);

    DistributedRowMatrix d = a.multiplyJava(b, MATRIX_D_PATH);
    if (c.verify(d)) {
        System.out.println("Verify PASSED!");
    } else {
        System.out.println("Verify FAILED!");
    }

    if (isDebugging) {
        System.out.println("Matrix A:");
        a.printDistributedRowMatrix();
        System.out.println("Matrix A transposed:");
        aTransposed.printDistributedRowMatrix();
        System.out.println("Matrix B:");
        b.printDistributedRowMatrix();
        System.out.println("Matrix C:");
        c.printDistributedRowMatrix();
        System.out.println("Matrix D:");
        d.printDistributedRowMatrix();

        printOutput(conf);
    }
    return 0;
}

From source file:at.illecker.hadoop.rootbeer.examples.matrixmultiplication.DistributedRowMatrix.java

License:Apache License

/**
 * This implements matrix this.transpose().times(other)
 * /* w  w  w .  jav a 2  s . co m*/
 * @param other a DistributedRowMatrix
 * @param outPath path to write result to
 * @return a DistributedRowMatrix containing the product
 */
public DistributedRowMatrix times(DistributedRowMatrix other, Path outPath) throws IOException {
    if (numRows != other.numRows()) {
        throw new CardinalityException(numRows, other.numRows());
    }

    Configuration initialConf = getConf() == null ? new Configuration() : getConf();
    Configuration conf = MatrixMultiplicationJob.createMatrixMultiplyJobConf(initialConf, rowPath,
            other.rowPath, outPath, other.numCols);
    JobClient.runJob(new JobConf(conf));
    DistributedRowMatrix out = new DistributedRowMatrix(outPath, outputTmpPath, numCols, other.numCols());
    out.setConf(conf);
    return out;
}

From source file:at.illecker.hadoop.rootbeer.examples.matrixmultiplication.DistributedRowMatrix.java

License:Apache License

/**
 * This implements matrix multiplication A * B using MapReduce tasks on CPU or
 * GPU/*from w  w w  .  ja v a 2 s. c om*/
 * 
 * @param other a DistributedRowMatrix
 * @param outPath path to write result to
 * @param useGPU use GPU or CPU (default: false, use CPU)
 * @return a DistributedRowMatrix containing the product
 */
public DistributedRowMatrix multiplyMapReduce(DistributedRowMatrix other, Path outPath, boolean useGPU,
        boolean isMatrixATransposed, int tileWidth, boolean isDebugging) throws IOException {
    // Check if cols of MatrixA = rows of MatrixB
    // (l x m) * (m x n) = (l x n)
    if (numCols != other.numRows()) {
        throw new CardinalityException(numCols, other.numRows());
    }

    Configuration initialConf = (getConf() == null) ? new Configuration() : getConf();

    // Transpose Matrix within a new MapReduce Job
    DistributedRowMatrix transposed = this;
    if (!isMatrixATransposed) {
        transposed = transposed.transpose();
    }
    // Debug
    // System.out.println("DistributedRowMatrix transposed:");
    // transposed.printDistributedRowMatrix();

    // Build MatrixMultiplication job configuration
    Configuration conf = null;
    if (!useGPU) {
        conf = MatrixMultiplicationCpu.createMatrixMultiplicationCpuConf(initialConf, transposed.rowPath,
                other.rowPath, outPath, other.numCols, isDebugging);
    } else { // use GPU
        conf = MatrixMultiplicationGpu.createMatrixMultiplicationGpuConf(initialConf, transposed.rowPath,
                other.rowPath, outPath, other.numCols, tileWidth, isDebugging);
    }

    // Multiply Matrix with transposed one
    JobClient.runJob(new JobConf(conf));

    // Read resulting Matrix from HDFS
    DistributedRowMatrix out = new DistributedRowMatrix(outPath, outputTmpPath, this.numRows, other.numCols());
    out.setConf(conf);

    return out;
}

From source file:at.illecker.hadoop.rootbeer.examples.matrixmultiplication.DistributedRowMatrix.java

License:Apache License

/**
 * This implements matrix multiplication A * B in Java without using MapReduce
 * tasks//from  ww  w .jav  a 2  s  .com
 * 
 * @param other a DistributedRowMatrix
 * @param outPath path to write result to
 * 
 * @return a DistributedRowMatrix containing the product
 */
public DistributedRowMatrix multiplyJava(DistributedRowMatrix other, Path outPath) throws IOException {
    // Check if cols of MatrixA = rows of MatrixB
    // (l x m) * (m x n) = (l x n)
    if (numCols != other.numRows()) {
        throw new CardinalityException(numCols, other.numRows());
    }

    // Multiply Matrix with transposed one without new MapReduce Job
    final double[][] matrixA = this.toDoubleArray();
    final double[][] matrixB = other.toDoubleArray();
    final double[][] matrixC = new double[this.numRows][other.numCols];

    int m = this.numRows;
    int n = this.numCols;
    int p = other.numCols;
    for (int k = 0; k < n; k++) {
        for (int i = 0; i < m; i++) {
            for (int j = 0; j < p; j++) {
                matrixC[i][j] = matrixC[i][j] + matrixA[i][k] * matrixB[k][j];
            }
        }
    }

    // Save resulting Matrix to HDFS
    try {
        writeDistributedRowMatrix(this.conf, matrixC, this.numRows, other.numCols, outPath, false);
    } catch (Exception e) {
        e.printStackTrace();
    }

    // Read resulting Matrix from HDFS
    DistributedRowMatrix out = new DistributedRowMatrix(outPath, outputTmpPath, numCols, other.numCols());
    out.setConf(conf);

    return out;
}

From source file:at.illecker.hadoop.rootbeer.examples.matrixmultiplication.gpu.MatrixMultiplicationGpu.java

License:Apache License

@Override
public int run(String[] strings) throws Exception {
    addOption("numRowsA", "nra", "Number of rows of the first input matrix", true);
    addOption("numColsA", "nca", "Number of columns of the first input matrix", true);
    addOption("numRowsB", "nrb", "Number of rows of the second input matrix", true);
    addOption("numColsB", "ncb", "Number of columns of the second input matrix", true);
    addOption("tileWidth", "tw", "TileWidth denotes the size of a submatrix", true);
    addOption("debug", "db", "Enable debugging (true|false)", false);

    Map<String, List<String>> argMap = parseArguments(strings);
    if (argMap == null) {
        return -1;
    }//  ww w. jav a2  s .  c  o  m

    int numRowsA = Integer.parseInt(getOption("numRowsA"));
    int numColsA = Integer.parseInt(getOption("numColsA"));
    int numRowsB = Integer.parseInt(getOption("numRowsB"));
    int numColsB = Integer.parseInt(getOption("numColsB"));

    // TILE_WITH = 32
    // --> 2 * 32 = 1024 threads matches the blocksize
    int tileWidth = Integer.parseInt(getOption("tileWidth"));
    boolean isDebugging = Boolean.parseBoolean(getOption("debug"));

    LOG.info("numRowsA: " + numRowsA);
    LOG.info("numColsA: " + numColsA);
    LOG.info("numRowsB: " + numRowsB);
    LOG.info("numColsB: " + numColsB);
    LOG.info("tileWidth: " + tileWidth);
    LOG.info("isDebugging: " + isDebugging);
    LOG.info("outputPath: " + OUTPUT_DIR);

    if (numColsA != numRowsB) {
        throw new CardinalityException(numColsA, numRowsB);
    }

    Configuration conf = new Configuration(getConf());

    // Create random DistributedRowMatrix
    // use constant seeds to get reproducable results
    // Matrix A is stored transposed
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_TRANSPOSED_PATH, true);
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsB, numColsB, new Random(1337L),
            MATRIX_B_PATH, false);

    // Load DistributedRowMatrix a and b
    DistributedRowMatrix aTransposed = new DistributedRowMatrix(MATRIX_A_TRANSPOSED_PATH, OUTPUT_DIR, numRowsA,
            numColsA);
    aTransposed.setConf(conf);

    DistributedRowMatrix b = new DistributedRowMatrix(MATRIX_B_PATH, OUTPUT_DIR, numRowsB, numColsB);
    b.setConf(conf);

    // MatrixMultiply all within a new MapReduce job
    long startTime = System.currentTimeMillis();
    DistributedRowMatrix c = aTransposed.multiplyMapReduce(b, MATRIX_C_PATH, true, true, tileWidth,
            isDebugging);
    System.out.println("MatrixMultiplicationGpu using Hadoop finished in "
            + (System.currentTimeMillis() - startTime) / 1000.0 + " seconds");

    // Verification
    // Overwrite matrix A, NOT transposed for verification check
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_PATH, false);
    DistributedRowMatrix a = new DistributedRowMatrix(MATRIX_A_PATH, OUTPUT_DIR, numRowsA, numColsA);
    a.setConf(conf);

    DistributedRowMatrix d = a.multiplyJava(b, MATRIX_D_PATH);
    if (c.verify(d)) {
        System.out.println("Verify PASSED!");
    } else {
        System.out.println("Verify FAILED!");
    }

    if (isDebugging) {
        System.out.println("Matrix A:");
        a.printDistributedRowMatrix();
        System.out.println("Matrix A transposed:");
        aTransposed.printDistributedRowMatrix();
        System.out.println("Matrix B:");
        b.printDistributedRowMatrix();
        System.out.println("Matrix C:");
        c.printDistributedRowMatrix();
        System.out.println("Matrix D:");
        d.printDistributedRowMatrix();

        printOutput(conf);
    }
    return 0;
}

From source file:at.illecker.hama.rootbeer.examples.matrixmultiplication.compositeinput.cpu.MatrixMultiplicationBSPCpu.java

License:Apache License

public static void main(String[] args) throws Exception {

    // Defaults//from  ww  w  . jav  a 2  s  .c  om
    int numRowsA = 10;
    int numColsA = 10;
    int numRowsB = 10;
    int numColsB = 10;
    boolean isDebugging = false;

    Configuration conf = new HamaConfiguration();
    BSPJobClient jobClient = new BSPJobClient(conf);
    ClusterStatus cluster = jobClient.getClusterStatus(true);

    if (args.length > 0) {
        if (args.length == 6) {
            conf.setInt("bsp.peers.num", Integer.parseInt(args[0]));
            numRowsA = Integer.parseInt(args[1]);
            numColsA = Integer.parseInt(args[2]);
            numRowsB = Integer.parseInt(args[3]);
            numColsB = Integer.parseInt(args[4]);
            isDebugging = Boolean.parseBoolean(args[5]);

        } else {
            System.out.println("Wrong argument size!");
            System.out.println("    Argument1=numBspTask");
            System.out.println("    Argument2=numRowsA | Number of rows of the first input matrix");
            System.out.println("    Argument3=numColsA | Number of columns of the first input matrix");
            System.out.println("    Argument4=numRowsB | Number of rows of the second input matrix");
            System.out.println("    Argument5=numColsB | Number of columns of the second input matrix");
            System.out.println("    Argument6=debug | Enable debugging (true|false)");
            return;
        }
    } else {
        conf.setInt("bsp.peers.num", cluster.getMaxTasks());
    }

    conf.setInt("matrixmultiplication.bsp.cpu.numRowsA", numRowsA);
    conf.setInt("matrixmultiplication.bsp.cpu.numColsA", numColsA);
    conf.setInt("matrixmultiplication.bsp.cpu.numRowsB", numRowsB);
    conf.setInt("matrixmultiplication.bsp.cpu.numColsB", numRowsB);
    conf.setBoolean(DEBUG, isDebugging);

    LOG.info("NumBspTask: " + conf.getInt("bsp.peers.num", 0));
    LOG.info("numRowsA: " + numRowsA);
    LOG.info("numColsA: " + numColsA);
    LOG.info("numRowsB: " + numRowsB);
    LOG.info("numColsB: " + numColsB);
    LOG.info("isDebugging: " + isDebugging);
    LOG.info("outputPath: " + OUTPUT_DIR);

    if (numColsA != numRowsB) {
        throw new CardinalityException(numColsA, numRowsB);
    }

    // Create random DistributedRowMatrix
    // use constant seeds to get reproducable results
    // Matrix A is stored transposed
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_PATH, true);
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsB, numColsB, new Random(1337L),
            MATRIX_B_PATH, false);

    // Load DistributedRowMatrix a and b
    DistributedRowMatrix a = new DistributedRowMatrix(MATRIX_A_PATH, OUTPUT_DIR, numRowsA, numColsA);
    a.setConf(conf);

    DistributedRowMatrix b = new DistributedRowMatrix(MATRIX_B_PATH, OUTPUT_DIR, numRowsB, numColsB);
    b.setConf(conf);

    // MatrixMultiply all within a new BSP job
    long startTime = System.currentTimeMillis();
    DistributedRowMatrix c = a.multiplyBSP(b, MATRIX_C_PATH, false, false);
    System.out.println("MatrixMultiplicationCpu using Hama finished in "
            + (System.currentTimeMillis() - startTime) / 1000.0 + " seconds");

    // Verification
    // Overwrite matrix A, NOT transposed for verification check
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_PATH, false);
    a = new DistributedRowMatrix(MATRIX_A_PATH, OUTPUT_DIR, numRowsA, numColsA);
    a.setConf(conf);

    DistributedRowMatrix d = a.multiplyJava(b, MATRIX_D_PATH);
    if (c.verify(d)) {
        System.out.println("Verify PASSED!");
    } else {
        System.out.println("Verify FAILED!");
    }

    if (isDebugging) {
        System.out.println("Matrix A:");
        a.printDistributedRowMatrix();
        System.out.println("Matrix B:");
        b.printDistributedRowMatrix();
        System.out.println("Matrix C:");
        c.printDistributedRowMatrix();
        System.out.println("Matrix D:");
        d.printDistributedRowMatrix();

        printOutput(conf);
    }
}

From source file:at.illecker.hama.rootbeer.examples.matrixmultiplication.compositeinput.gpu.MatrixMultiplicationBSPGpu.java

License:Apache License

public static void main(String[] args) throws Exception {

    // Defaults/*from w ww  .jav  a2  s . com*/
    int numRowsA = 10;
    int numColsA = 10;
    int numRowsB = 10;
    int numColsB = 10;
    boolean isDebugging = false;

    Configuration conf = new HamaConfiguration();
    BSPJobClient jobClient = new BSPJobClient(conf);
    ClusterStatus cluster = jobClient.getClusterStatus(true);

    if (args.length > 0) {
        if (args.length == 6) {
            conf.setInt("bsp.peers.num", Integer.parseInt(args[0]));
            numRowsA = Integer.parseInt(args[1]);
            numColsA = Integer.parseInt(args[2]);
            numRowsB = Integer.parseInt(args[3]);
            numColsB = Integer.parseInt(args[4]);
            isDebugging = Boolean.parseBoolean(args[5]);

        } else {
            System.out.println("Wrong argument size!");
            System.out.println("    Argument1=numBspTask");
            System.out.println("    Argument2=numRowsA | Number of rows of the first input matrix");
            System.out.println("    Argument3=numColsA | Number of columns of the first input matrix");
            System.out.println("    Argument4=numRowsB | Number of rows of the second input matrix");
            System.out.println("    Argument5=numColsB | Number of columns of the second input matrix");
            System.out.println("    Argument6=debug | Enable debugging (true|false)");
            return;
        }
    } else {
        conf.setInt("bsp.peers.num", cluster.getMaxTasks());
    }

    conf.setInt("matrixmultiplication.bsp.gpu.numRowsA", numRowsA);
    conf.setInt("matrixmultiplication.bsp.gpu.numColsA", numColsA);
    conf.setInt("matrixmultiplication.bsp.gpu.numRowsB", numRowsB);
    conf.setInt("matrixmultiplication.bsp.gpu.numColsB", numRowsB);
    conf.setBoolean(DEBUG, isDebugging);

    LOG.info("NumBspTask: " + conf.getInt("bsp.peers.num", 0));
    LOG.info("numRowsA: " + numRowsA);
    LOG.info("numColsA: " + numColsA);
    LOG.info("numRowsB: " + numRowsB);
    LOG.info("numColsB: " + numColsB);
    LOG.info("isDebugging: " + isDebugging);
    LOG.info("outputPath: " + OUTPUT_DIR);

    if (numColsA != numRowsB) {
        throw new CardinalityException(numColsA, numRowsB);
    }

    // Create random DistributedRowMatrix
    // use constant seeds to get reproducable results
    // Matrix A is stored transposed
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_PATH, true);
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsB, numColsB, new Random(1337L),
            MATRIX_B_PATH, false);

    // Load DistributedRowMatrix a and b
    DistributedRowMatrix a = new DistributedRowMatrix(MATRIX_A_PATH, OUTPUT_DIR, numRowsA, numColsA);
    a.setConf(conf);

    DistributedRowMatrix b = new DistributedRowMatrix(MATRIX_B_PATH, OUTPUT_DIR, numRowsB, numColsB);
    b.setConf(conf);

    // MatrixMultiply all within a new BSP job
    long startTime = System.currentTimeMillis();
    DistributedRowMatrix c = a.multiplyBSP(b, MATRIX_C_PATH, true, false);
    System.out.println("MatrixMultiplicationCpu using Hama finished in "
            + (System.currentTimeMillis() - startTime) / 1000.0 + " seconds");

    // Verification
    // Overwrite matrix A, NOT transposed for verification check
    DistributedRowMatrix.createRandomDistributedRowMatrix(conf, numRowsA, numColsA, new Random(42L),
            MATRIX_A_PATH, false);
    a = new DistributedRowMatrix(MATRIX_A_PATH, OUTPUT_DIR, numRowsA, numColsA);
    a.setConf(conf);

    DistributedRowMatrix d = a.multiplyJava(b, MATRIX_D_PATH);
    if (c.verify(d)) {
        System.out.println("Verify PASSED!");
    } else {
        System.out.println("Verify FAILED!");
    }

    if (isDebugging) {
        System.out.println("Matrix A:");
        a.printDistributedRowMatrix();
        System.out.println("Matrix B:");
        b.printDistributedRowMatrix();
        System.out.println("Matrix C:");
        c.printDistributedRowMatrix();
        System.out.println("Matrix D:");
        d.printDistributedRowMatrix();

        printOutput(conf);
    }
}

From source file:at.illecker.hama.rootbeer.examples.matrixmultiplication.compositeinput.util.DistributedRowMatrix.java

License:Apache License

/**
 * This implements matrix multiplication A * B using MapReduce tasks on CPU or
 * GPU/*from www .j  a v a 2  s  . com*/
 * 
 * @param other a DistributedRowMatrix
 * @param outPath path to write result to
 * @param useGPU use GPU or CPU (default: false, use CPU)
 * @return a DistributedRowMatrix containing the product
 */
public DistributedRowMatrix multiplyBSP(DistributedRowMatrix other, Path outPath, boolean useGPU,
        boolean transposeMatrixA) throws IOException, ClassNotFoundException, InterruptedException {
    // Check if cols of MatrixA = rows of MatrixB
    // (l x m) * (m x n) = (l x n)
    if (numCols != other.numRows()) {
        throw new CardinalityException(numCols, other.numRows());
    }

    Configuration initialConf = (getConf() == null) ? new HamaConfiguration() : getConf();

    // Transpose Matrix within a new MapReduce Job
    DistributedRowMatrix transposed = this;
    if (transposeMatrixA) {
        transposed = transposed.transpose();
    }
    // Debug
    // System.out.println("DistributedRowMatrix transposed:");
    // transposed.printDistributedRowMatrix();

    // Build MatrixMultiplication job configuration
    BSPJob job = null;
    if (!useGPU) {
        job = MatrixMultiplicationBSPCpu.createMatrixMultiplicationBSPCpuConf(initialConf, transposed.rowPath,
                other.rowPath, outPath.getParent(), other.numCols);
    } else { // use GPU
        job = MatrixMultiplicationBSPGpu.createMatrixMultiplicationBSPGpuConf(initialConf, transposed.rowPath,
                other.rowPath, outPath.getParent(), other.numCols);
    }

    // Multiply Matrix with transposed one
    if (job.waitForCompletion(true)) {

        // Rename result file to output path
        Configuration conf = job.getConfiguration();
        FileSystem fs = outPath.getFileSystem(conf);
        FileStatus[] files = fs.listStatus(outPath.getParent());
        for (int i = 0; i < files.length; i++) {
            if ((files[i].getPath().getName().startsWith("part-")) && (files[i].getLen() > 97)) {
                fs.rename(files[i].getPath(), outPath);
                break;
            }
        }

        // Read resulting Matrix from HDFS
        DistributedRowMatrix out = new DistributedRowMatrix(outPath, outputTmpPath, this.numRows,
                other.numCols());
        out.setConf(conf);

        return out;
    }

    return null;
}

From source file:com.twitter.algebra.matrix.multiply.ABInnerHDFSBroadcastOfB.java

License:Apache License

/**
 * Perform A x B, where A and B are already wrapped in a DistributedRowMatrix
 * object. Refer to {@link ABInnerHDFSBroadcastOfB} for further details.
 * /*from www  .ja  v  a2  s  . com*/
 * @param conf the initial configuration
 * @param A matrix A
 * @param B matrix B
 * @param label the label for the output directory
 * @return AxB wrapped in a DistributedRowMatrix object
 * @throws IOException
 * @throws InterruptedException
 * @throws ClassNotFoundException
 */
public static DistributedRowMatrix run(Configuration conf, DistributedRowMatrix A, DistributedRowMatrix B,
        String label) throws IOException, InterruptedException, ClassNotFoundException {
    log.info("running " + ABInnerHDFSBroadcastOfB.class.getName());
    if (A.numCols() != B.numRows()) {
        throw new CardinalityException(A.numCols(), B.numRows());
    }
    Path outPath = new Path(A.getOutputTempPath(), label);
    FileSystem fs = FileSystem.get(outPath.toUri(), conf);
    ABInnerHDFSBroadcastOfB job = new ABInnerHDFSBroadcastOfB();

    if (!fs.exists(outPath)) {
        job.run(conf, A.getRowPath(), B.getRowPath(), outPath, B.numRows(), B.numCols());
    } else {
        log.warn("----------- Skip already exists: " + outPath);
    }
    DistributedRowMatrix distRes = new DistributedRowMatrix(outPath, A.getOutputTempPath(), A.numRows(),
            B.numCols());
    distRes.setConf(conf);
    return distRes;
}

From source file:com.twitter.algebra.matrix.multiply.ABOuterHDFSBroadcastOfA.java

License:Apache License

/**
 * Perform A x B, where A and B are already wrapped in a DistributedRowMatrix
 * object. Refer to {@link ABOuterHDFSBroadcastOfA} for further details.
 * /*from   w w  w  .j av a  2  s . co m*/
 * @param conf
 *          the initial configuration
 * @param A
 *          matrix A
 * @param B
 *          matrix B
 * @param label
 *          the label for the output directory
 * @return AxB wrapped in a DistributedRowMatrix object
 * @throws IOException
 * @throws InterruptedException
 * @throws ClassNotFoundException
 */
public static DistributedRowMatrix run(Configuration conf, DistributedRowMatrix A, DistributedRowMatrix B,
        String label) throws IOException, InterruptedException, ClassNotFoundException {
    log.info("running " + ABOuterHDFSBroadcastOfA.class.getName());
    if (A.numCols() != B.numRows()) {
        throw new CardinalityException(A.numCols(), B.numRows());
    }
    Path outPath = new Path(A.getOutputTempPath(), label);
    FileSystem fs = FileSystem.get(outPath.toUri(), conf);
    ABOuterHDFSBroadcastOfA job = new ABOuterHDFSBroadcastOfA();
    if (!fs.exists(outPath)) {
        job.run(conf, A.getRowPath(), B.getRowPath(), outPath, A.numRows(), A.numCols());
    } else {
        log.warn("----------- Skip already exists: " + outPath);
    }
    DistributedRowMatrix distRes = new DistributedRowMatrix(outPath, A.getOutputTempPath(), A.numRows(),
            B.numCols());
    distRes.setConf(conf);
    return distRes;
}