
org.numenta.nupic.util.SparseObjectMatrix Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of htm.java Show documentation
Show all versions of htm.java Show documentation
The Java version of Numenta's HTM technology
The newest version!
/* ---------------------------------------------------------------------
* Numenta Platform for Intelligent Computing (NuPIC)
* Copyright (C) 2014, Numenta, Inc. Unless you have an agreement
* with Numenta, Inc., for a separate license for this software code, the
* following terms and conditions apply:
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero Public License version 3 as
* published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU Affero Public License for more details.
*
* You should have received a copy of the GNU Affero Public License
* along with this program. If not, see http://www.gnu.org/licenses.
*
* http://numenta.org/licenses/
* ---------------------------------------------------------------------
*/
package org.numenta.nupic.util;
import java.util.Arrays;
import org.numenta.nupic.model.Persistable;
import gnu.trove.map.TIntObjectMap;
import gnu.trove.map.hash.TIntObjectHashMap;
/**
* Allows storage of array data in sparse form, meaning that the indexes
* of the data stored are maintained while empty indexes are not. This allows
* savings in memory and computational efficiency because iterative algorithms
* need only query indexes containing valid data.
*
* @author David Ray
*
* @param
*/
public class SparseObjectMatrix extends AbstractSparseMatrix implements Persistable {
/** keep it simple */
private static final long serialVersionUID = 1L;
private TIntObjectMap sparseMap = new TIntObjectHashMap();
/**
* Constructs a new {@code SparseObjectMatrix}
* @param dimensions the dimensions of this array
*/
public SparseObjectMatrix(int[] dimensions) {
super(dimensions, false);
}
/**
* Constructs a new {@code SparseObjectMatrix}
* @param dimensions the dimensions of this array
* @param useColumnMajorOrdering where inner index increments most frequently
*/
public SparseObjectMatrix(int[] dimensions, boolean useColumnMajorOrdering) {
super(dimensions, useColumnMajorOrdering);
}
/**
* Sets the object to occupy the specified index.
*
* @param index the index the object will occupy
* @param object the object to be indexed.
*/
@Override
public SparseObjectMatrix set(int index, T object) {
sparseMap.put(index, (T)object);
return this;
}
/**
* Sets the specified object to be indexed at the index
* computed from the specified coordinates.
* @param object the object to be indexed.
* @param coordinates the row major coordinates [outer --> ,...,..., inner]
*/
@Override
public SparseObjectMatrix set(int[] coordinates, T object) {
set(computeIndex(coordinates), object);
return this;
}
/**
* Returns the T at the specified index.
*
* @param index the index of the T to return
* @return the T at the specified index.
*/
@Override
public T getObject(int index) {
return get(index);
}
/**
* Returns the T at the index computed from the specified coordinates
* @param coordinates the coordinates from which to retrieve the indexed object
* @return the indexed object
*/
@Override
public T get(int... coordinates) {
return get(computeIndex(coordinates));
}
/**
* Returns the T at the specified index.
*
* @param index the index of the T to return
* @return the T at the specified index.
*/
@Override
public T get(int index) {
return this.sparseMap.get(index);
}
/**
* Returns a sorted array of occupied indexes.
* @return a sorted array of occupied indexes.
*/
@Override
public int[] getSparseIndices() {
return reverse(sparseMap.keys());
}
/**
* {@inheritDoc}
*/
@Override
public String toString() {
return Arrays.toString(getDimensions());
}
/* (non-Javadoc)
* @see java.lang.Object#hashCode()
*/
@Override
public int hashCode() {
final int prime = 31;
int result = super.hashCode();
result = prime * result + ((sparseMap == null) ? 0 : sparseMap.hashCode());
return result;
}
/* (non-Javadoc)
* @see java.lang.Object#equals(java.lang.Object)
*/
@SuppressWarnings("rawtypes")
@Override
public boolean equals(Object obj) {
if(this == obj)
return true;
if(!super.equals(obj))
return false;
if(getClass() != obj.getClass())
return false;
SparseObjectMatrix other = (SparseObjectMatrix)obj;
if(sparseMap == null) {
if(other.sparseMap != null)
return false;
} else if(!sparseMap.equals(other.sparseMap))
return false;
return true;
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy