Examples of org.apache.hama.commons.io.MatrixWritable

org.apache.hama.commons.io.MatrixWritable
Majorly designed for dense matrices, can be extended for sparse ones as well.

    // write the number of neurons for each layer
    for (int i = 0; i < this.numberOfLayers; ++i) {
      output.writeInt(this.layerSizeArray[i]);
    }
    for (int i = 0; i < numberOfLayers - 1; ++i) {
      MatrixWritable matrixWritable = new MatrixWritable(this.weightMatrice[i]);
      matrixWritable.write(output);
    }


    // serialize the feature transformer
    Class<? extends FeatureTransformer> featureTransformerCls = this.featureTransformer
        .getClass();

View Full Code Here

      BSPPeer<Text, VectorWritable, Text, VectorWritable, MapWritable> peer,
      DoubleMatrix featureMatrix, IntWritable msgFeatureMatrix, boolean broadcast) 
          throws IOException, SyncException, InterruptedException {
    // send to master peer
    MapWritable msg = new MapWritable();
    MatrixWritable mtx = new MatrixWritable(featureMatrix);
    msg.put(msgFeatureMatrix, mtx);
    String master = peer.getPeerName(peer.getNumPeers()/2);
    peer.send(master, msg);
    peer.sync();


    // normalize
    DoubleMatrix res = null;
    if (peer.getPeerName().equals(master)) {
      res = new DenseDoubleMatrix(featureMatrix.getRowCount(),
                                  featureMatrix.getColumnCount(), 0);
      int incomingMsgCount = 0;
      while ( (msg = peer.getCurrentMessage()) != null) {
        MatrixWritable tmp = (MatrixWritable) msg.get(msgFeatureMatrix);
        res.add(tmp.getMatrix());
        incomingMsgCount++;
      }
      res.divide(incomingMsgCount);
    }


    if (broadcast) {
      if (peer.getPeerName().equals(master)) {
        // broadcast to all
        msg = new MapWritable();
        msg.put(msgFeatureMatrix, new MatrixWritable(res));
        // send to all
        for (String peerName : peer.getAllPeerNames()) {
          peer.send(peerName, msg);
        }
      }

View Full Code Here

    // write the number of neurons for each layer
    for (int i = 0; i < this.numberOfLayers; ++i) {
      output.writeInt(this.layerSizeArray[i]);
    }
    for (int i = 0; i < numberOfLayers - 1; ++i) {
      MatrixWritable matrixWritable = new MatrixWritable(this.weightMatrice[i]);
      matrixWritable.write(output);
    }


    // serialize the feature transformer
    Class<? extends FeatureTransformer> featureTransformerCls = this.featureTransformer
        .getClass();

View Full Code Here

TOP

Related Classes of org.apache.hama.commons.io.MatrixWritable

org.apache.hama.commons.math.DenseDoubleMatrix

org.apache.hama.commons.math.DoubleMatrix

org.apache.hama.ml.perception.SmallMultiLayerPerceptron

org.apache.hama.ml.recommendation.cf.OnlineTrainBSP

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.