
org.apache.mahout.h2obindings.H2OBlockMatrix Maven / Gradle / Ivy
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.mahout.h2obindings;
import org.apache.mahout.math.Matrix;
import org.apache.mahout.math.Vector;
import org.apache.mahout.math.AbstractMatrix;
import org.apache.mahout.math.DenseMatrix;
import org.apache.mahout.math.SparseMatrix;
import org.apache.mahout.math.flavor.MatrixFlavor;
import water.fvec.Chunk;
/**
* A Matrix implementation to represent a vertical Block of DRM.
*
* Creation of the matrix is an O(1) operation with negligible
* overhead, and will remain so as long as the matrix is only
* read from (no modifications).
*
* On the first modification, create a copy on write Matrix and
* all further operations happen on this cow matrix.
*
* The benefit is, mapBlock() closures which never modify the
* input matrix save on the copy overhead.
*/
public class H2OBlockMatrix extends AbstractMatrix {
/** Backing chunks which store the original matrix data */
private Chunk chks[];
/** Copy on write matrix created on demand when original matrix is modified */
private Matrix cow;
/** Class constructor. */
public H2OBlockMatrix(Chunk chks[]) {
super(chks[0].len(), chks.length);
this.chks = chks;
}
/**
* Internal method to create the copy on write matrix.
*
* Once created, all further operations are performed on the CoW matrix
*/
private void cow() {
if (cow != null) {
return;
}
if (chks[0].isSparse()) {
cow = new SparseMatrix(chks[0].len(), chks.length);
} else {
cow = new DenseMatrix(chks[0].len(), chks.length);
}
for (int c = 0; c < chks.length; c++) {
for (int r = 0; r < chks[0].len(); r++) {
cow.setQuick(r, c, chks[c].atd(r));
}
}
}
@Override
public void setQuick(int row, int col, double val) {
cow();
cow.setQuick(row, col, val);
}
@Override
public Matrix like(int nrow, int ncol) {
if (chks[0].isSparse()) {
return new SparseMatrix(nrow, ncol);
} else {
return new DenseMatrix(nrow, ncol);
}
}
@Override
public Matrix like() {
if (chks[0].isSparse()) {
return new SparseMatrix(rowSize(), columnSize());
} else {
return new DenseMatrix(rowSize(), columnSize());
}
}
@Override
public double getQuick(int row, int col) {
if (cow != null) {
return cow.getQuick(row, col);
} else {
return chks[col].atd(row);
}
}
@Override
public Matrix assignRow(int row, Vector v) {
cow();
cow.assignRow(row, v);
return cow;
}
@Override
public Matrix assignColumn(int col, Vector v) {
cow();
cow.assignColumn(col, v);
return cow;
}
@Override
public MatrixFlavor getFlavor() {
if (cow != null) {
return cow.getFlavor();
} else if (chks[0].isSparse()) {
return MatrixFlavor.SPARSELIKE;
} else {
return MatrixFlavor.DENSELIKE;
}
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy