jas.ai.ann
Class NodeLayer

java.lang.Object
  jas.ai.ann.NodeLayer

Direct Known Subclasses:

InputNodeLayer, OutputNodeLayer

public class NodeLayer
extends java.lang.Object

A node layer is a generic vector. It is a base class for specific implementations of the input and output layers for ANN architecture. The node layer is equipped with bias capability, so when the bias is enabled you can access the whole vector (with the bias in position 0) using the get/setValue methods and ignore the bias node using the get/setNode methods.
If, for instance, the bias is enabled the following instructions:
x = getValue(0);
y = getNode(0);
will return in variable x the first value of the vector (vector[0]) and in variable y the second value of the vector (vector[1]).

Title: JAS - Neural network package

Description: Java Agent-based Simulation library

Copyright: 2002-3 Michele Sonnessa

This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. This library 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 Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.

Author:

Michele Sonnessa

Constructor Summary

NodeLayer(int hiddenNodes)

NodeLayer(int nodes, boolean withBias)
Build a vector with a given number of nodes.

Method Summary

double[]	getArray() Return the internal vector.
double[]	getArrayCopy() Return a copy of the internal vector.
double	getBias() Return the value of bias.
int	getNodesNumber() Return the number of nodes, excluding the bias one.
double	getNodeValue(int nodeIndex) Return the value of the given node index.
Jama.Matrix	getRowVector() Return a copy of the internal vector, using the jas.maths.Jama.Matrix format.
Jama.Matrix	getTrasposedVector() Return a trasposed copy of the internal vector, using the jas.maths.Jama.Matrix format.
double	getValue(int arrayIndex) Return the value of the given index of the vector.
int	getVectorLength() Return the number of nodes, including the bias one.
boolean	hasBias() Check if bias has been enabled.
void	setBias(double value) Set the given value for the bias.
void	setNodeValue(int nodeIndex, double value) Set the given value for the given node index.
void	setValue(int arrayIndex, double value) Set the given value for the given index of the vector.
void	setVector(double[] values) Set node values the each cell of the vector.
java.lang.String	toString()

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

NodeLayer

public NodeLayer(int hiddenNodes)

Parameters:

hiddenNodes -

NodeLayer

public NodeLayer(int nodes,
                 boolean withBias)

Build a vector with a given number of nodes.

Parameters:

nodes - The number of nodes in the layers.

withBias - If true the position 0 of the vector will contain the bias and the total number of cells in the vector will be [nodes + 1].

Method Detail

getArray

public double[] getArray()

Return the internal vector. WARNING It is a reference to the vector: any change will affect the internal vector too.

Returns:

The internal vector itself.

getArrayCopy

public double[] getArrayCopy()

Return a copy of the internal vector.

Returns:

A copy of the internal vector.

getBias

public double getBias()

Return the value of bias. This method works only if bias has been enabled bt the constructor.

Returns:

current bias value.

Throws:

java.lang.UnsupportedOperationException - If bias has been not enabled during construction.

getNodesNumber

public int getNodesNumber()

Return the number of nodes, excluding the bias one. If the bias is activated the result will be vector.length - 1.

Returns:

The number of nodes without counting the bias one.

getNodeValue

public double getNodeValue(int nodeIndex)

Return the value of the given node index. If bias is enabled the value is the nodeIndex + 1 position of the vector.

Parameters:

nodeIndex - The 0 based index for the node. WARNING The index bounds are not checked!

Returns:

current node value for the given index.

getRowVector

public Jama.Matrix getRowVector()

Return a copy of the internal vector, using the jas.maths.Jama.Matrix format.

Returns:

A jas.maths.Jama.Matrix copy of the internal vector, with values on the first row of the matrix.

getTrasposedVector

public Jama.Matrix getTrasposedVector()

Return a trasposed copy of the internal vector, using the jas.maths.Jama.Matrix format.

Returns:

A jas.maths.Jama.Matrix copy of the internal vector, with values on the first column of the matrix.

getValue

public double getValue(int arrayIndex)

Return the value of the given index of the vector. If bias is enabled the 0 index corresponds to the bias position.

Parameters:

arrayIndex - The 0 based index for the vector. WARNING The index bounds are not checked!

Returns:

current node cell for the given index.

getVectorLength

public int getVectorLength()

Return the number of nodes, including the bias one. The result is always vector.length.

Returns:

The length of the vector.

hasBias

public boolean hasBias()

Check if bias has been enabled.

Returns:

True if bias is present, false otherwise.

setBias

public void setBias(double value)

Set the given value for the bias. This method works only if bias has been enabled bt the constructor.

Parameters:

value - The new bias value.

Throws:

java.lang.UnsupportedOperationException - If bias has been not enabled during construction.

setNodeValue

public void setNodeValue(int nodeIndex,
                         double value)

Set the given value for the given node index. If bias is enabled the value is the nodeIndex + 1 position of the vector.

Parameters:

nodeIndex - The 0 based index for the node. WARNING The index bounds are not checked!

value - The new double value for the given vector cell.

setValue

public void setValue(int arrayIndex,
                     double value)

Set the given value for the given index of the vector. If bias is enabled the 0 index corresponds to the bias position.

Parameters:

arrayIndex - The 0 based index for the vector. WARNING The index bounds are not checked!

value - The new double value for the given vector cell.

setVector

public void setVector(double[] values)

Set node values the each cell of the vector. If bias is enabled it is ignored.

Parameters:

values - A double[] array containig node values.

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object