harp Documentation - Harp-DAAL-Mom

Overview

Moments are basic quantitative measures of data set characteristics such as location and dispersion. We compute the following low order characteristics: minimums/maximums, sums, means, sums of squares, sums of squared differences from the means, second order raw moments, variances, standard deviations, and variations.

Implementation of Harp-DAAL-Mom

Harp-DAAL-Mom is built upon the original DAAL distributed implementation of the algorithm using the Harp interface libraries for the collective-communication among the data nodes. The implementation of Mom in our Harp-DAAL consists of two levels. At the top Level, using Harp’s reduce operation to communication model data among mappers. At the bottom Level, using DAAL’s kernels to conduct local computations.

Brief background

There are primarily three kinds of nodes involved with the implementation.

name node
data node
- master node
- slave node

The Java language services provided by Intel as a wrapper to their C++ code.
The description of the intel daal’s JAVA API used can be found here

Code Walk-Through

The actual implemented code can be found here. The MapCollective function is defined here.

Step 1 (on data nodes)

The first step involves reading the files from the hdfs filesystem after splitting files between each mapper. Splitting is done by MultipleFileInputFormat class defined here. Data is converted into array which is eventually converted into the daal Numeric Table datastructure. In this example the files have been stored on the hdfs. The files have to be read in the daal table format as the local computations are performed by the daal libraries. Getting array from csv file -

  public static double[] loadPoints(String file,
    int pointsPerFile, int cenVecSize,
    Configuration conf) throws Exception {
    double[] points = new double[pointsPerFile * cenVecSize];
    Path pointFilePath = new Path(file);
    FileSystem fs =
      pointFilePath.getFileSystem(conf);
    FSDataInputStream in = fs.open(pointFilePath);
    int k =0;
    try{
      for(int i = 0; i < pointsPerFile;i++){
        String[] line = in.readLine().split(",");
        for(int j = 0; j < cenVecSize; j++){
          points[k] = Double.parseDouble(line[j]);
          k++;
        }
      }
    } finally{
      in.close();
    }
    return points;
  }

Converting array to Numeric Table -

List<double[]> pointArrays = MOMUtil.loadPoints(trainingDataFiles, pointsPerFile,
        vectorSize, conf, numThreads);
      ts2 = System.currentTimeMillis();
      load_time += (ts2 - ts1);

     // converting data to Numeric Table
      ts1 = System.currentTimeMillis();
      long nFeature = vectorSize;
      long nLabel = 1;
      long totalLengthFeature = 0;

      long[] array_startP_feature = new long[pointArrays.size()];
      double[][] array_data_feature = new double[pointArrays.size()][];

      for(int k=0;k<pointArrays.size();k++)
      {
       array_data_feature[k] = pointArrays.get(k);
       array_startP_feature[k] = totalLengthFeature;
       totalLengthFeature += pointArrays.get(k).length;
     }

     long featuretableSize = totalLengthFeature/nFeature;

   //initializing Numeric Table
     NumericTable featureArray_daal = new HomogenNumericTable(daal_Context, Double.class, nFeature, featuretableSize, NumericTable.AllocationFlag.DoAllocate);

Step 2 (on data nodes)

DistributedStep1Local is created for local computation. Computed partial results are sent to master node by reduce function.

DistributedStep1Local algorithm = new DistributedStep1Local(daal_Context, Float.class, Method.defaultDense);
    /* Set input objects for the algorithm */
    algorithm.input.set(InputId.data, featureArray_daal);

    /* Compute partial estimates on nodes */
    partialResult = algorithm.compute();
    ts2 = System.currentTimeMillis();
    compute_time += (ts2 - ts1);

    ts1 = System.currentTimeMillis();
    partialResultTable.addPartition(new Partition<>(this.getSelfID(), serializePartialResult(partialResult)));
    boolean reduceStatus = false;
    reduceStatus = this.reduce("mom", "sync-partialresult", partialResultTable, this.getMasterID());

Step 3 (only on master nodes)

Final low order characteristics are computed on master node.

 if(this.isMaster()){
      int[] pid = partialResultTable.getPartitionIDs().toIntArray();
    DistributedStep2Master algorithm = new DistributedStep2Master(daal_Context, Float.class, Method.defaultDense);
    ts1 = System.currentTimeMillis();
    for(int j = 0; j< pid.length; j++){
      try {
        algorithm.input.add(DistributedStep2MasterInputId.partialResults,
          deserializePartialResult(partialResultTable.getPartition(pid[j]).get())); 
      } catch (Exception e) 
      {  
        System.out.println("Fail to deserilize partialResultTable" + e.toString());
        e.printStackTrace();
      }
    }
    ts2 = System.currentTimeMillis();
    comm_time += (ts2 - ts1);

    ts1 = System.currentTimeMillis();
    algorithm.compute();
    result = algorithm.finalizeCompute();

    NumericTable minimum = result.get(ResultId.minimum);
    NumericTable maximum = result.get(ResultId.maximum);
    NumericTable sum = result.get(ResultId.sum);
    NumericTable sumSquares = result.get(ResultId.sumSquares);
    NumericTable sumSquaresCentered = result.get(ResultId.sumSquaresCentered);
    NumericTable mean = result.get(ResultId.mean);
    NumericTable secondOrderRawMoment = result.get(ResultId.secondOrderRawMoment);
    NumericTable variance = result.get(ResultId.variance);
    NumericTable standardDeviation = result.get(ResultId.standardDeviation);
    NumericTable variation = result.get(ResultId.variation);
    }

Running the codes

Make sure that the code is placed in the /harp/ml/daal directory. Run the harp-daal-mom.sh script here to run the code.

cd $HARP_ROOT/ml/daal
./test_scripts/harp-daal-mom.sh

The details of script is here