All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.ejml.data.DenseD2Matrix64F Maven / Gradle / Ivy

Go to download

A fast and easy to use dense and sparse matrix linear algebra library written in Java.

There is a newer version: 0.43.1
Show newest version
/*
 * Copyright (c) 2009-2020, Peter Abeles. All Rights Reserved.
 *
 * This file is part of Efficient Java Matrix Library (EJML).
 *
 * Licensed 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.ejml.data;

import java.io.Serializable;
import java.util.Arrays;


/**
 * A row matrix where the array is stored as a 2D array.
 *
 * @author Peter Abeles
 */
public class DenseD2Matrix64F implements Serializable, ReshapeMatrix, DMatrix {

    /**
     * Where the raw data for the matrix is stored.  The format is type dependent.
     */
    public double[][] data;

    /**
     * Number of rows in the matrix.
     */
    public int numRows;
    /**
     * Number of columns in the matrix.
     */
    public int numCols;

    public DenseD2Matrix64F( int numRows , int numCols ) {
        data = new double[ numRows ][numCols];
        this.numRows = numRows;
        this.numCols = numCols;
    }

//    public double[][] getData() {
//        return data;
//    }

    @Override
    public void reshape(int numRows, int numCols ) {
        if( numRows <= data.length ) {
            this.numRows = numRows;
        } else {
            throw new IllegalArgumentException("Requested number of rows is too great.");
        }

        if( numCols <= data[0].length ) {
            this.numCols = numCols;
        } else {
            throw new IllegalArgumentException("Requested number of columns is too great.");
        }
    }

    @Override
    public void zero() {
        Arrays.fill(data, 0, getNumElements(), 0.0);
    }

    @Override
    public double get(int row, int col) {
        return data[row][col];
    }

    @Override
    public void set(int row, int col, double val) {
        data[row][col] = val;
    }

    @Override
    public double unsafe_get( int row, int col) {
        return get(row,col);
    }

    @Override
    public void unsafe_set( int row, int col, double val) {
        set(row,col,val);
    }

    @Override
    public int getNumElements() {
        return numRows*numCols;
    }

    @Override
    public int getNumRows() {
        return numRows;
    }

    @Override
    public int getNumCols() {
        return numCols;
    }

    @Override
    public void print() {
    }

    @Override
    public void print(String format) {

    }

    @Override
    public  T copy() {
        return null;
    }

    @Override
    public DenseD2Matrix64F createLike() {
        return new DenseD2Matrix64F(numRows,numCols);
    }

    @Override
    public DenseD2Matrix64F create(int numRows, int numCols) {
        return new DenseD2Matrix64F(numRows,numCols);
    }

    @Override
    public void set(Matrix original) {
        throw new RuntimeException("Not yet supported");
    }

    @Override
    public MatrixType getType() {
        return MatrixType.UNSPECIFIED;
    }
}




© 2015 - 2024 Weber Informatics LLC | Privacy Policy