Package org.esa.snap.core.nn

Class NNffbpAlphaTabFast

java.lang.Object

org.esa.snap.core.nn.NNffbpAlphaTabFast

public class NNffbpAlphaTabFast
extends Object

This class is for using a Neural Net (NN) of type ffbp in a Java program. The program for training such a NN "ffbp1.0" was written in C by

Constructor Summary

Constructors

Constructor	Description
NNffbpAlphaTabFast(InputStream neuralNetStream)	Creates a neural net by reading the definition from the input stream.
NNffbpAlphaTabFast(String neuralNet)	Creates a neural net by reading the definition from the string.

Method Summary

Modifier and Type	Method	Description
double[]	calc(double[] nninp)	Method calc The NN is used.
NNCalc	calcJacobi(double[] nnInp)	Method calcJacobi The NN is used.
double[]	getInmax()
double[]	getInmin()
double[]	getOutmax()
double[]	getOutmin()
void	setInmax(double[] inmax)
void	setInmin(double[] inmin)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

NNffbpAlphaTabFast

public NNffbpAlphaTabFast(String neuralNet)
                   throws IOException

Creates a neural net by reading the definition from the string.

Parameters:

neuralNet - the neural net definition as a string

Throws:

IOException - if the neural net could not be read

NNffbpAlphaTabFast

public NNffbpAlphaTabFast(InputStream neuralNetStream)
                   throws IOException

Creates a neural net by reading the definition from the input stream.

Parameters:

neuralNetStream - the neural net definition as a input stream

Throws:

IOException - if the neural net could not be read

Method Detail

getInmin

public double[] getInmin()

setInmin

public void setInmin(double[] inmin)

getInmax

public double[] getInmax()

setInmax

public void setInmax(double[] inmax)

getOutmin

public double[] getOutmin()

getOutmax

public double[] getOutmax()

calcJacobi

public NNCalc calcJacobi(double[] nnInp)

Method calcJacobi The NN is used. For a given input vector the corresponding output vector together with the corresponding Jacobi matrix is returned as an instance of class NNCalc.

Parameters:

nnInp - The vector contains the nn_ininput parameters (must be in right order).

Returns:

The output and corresponding Jacobi matrix of the NN.

calc

public double[] calc(double[] nninp)

Method calc The NN is used. For a given input vector the corresponding output vector is returned.

Parameters:

nninp - The vector contains the nn_ininput parameters (must be in right order).

Returns:

The nn_out-long output vector.