org.encog.util.simple

Class EncogUtility

java.lang.Object

org.encog.util.simple.EncogUtility

public final class EncogUtility
extends java.lang.Object

General utility class for Encog. Provides for some common Encog procedures.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	EncogUtility.FalsePositiveReport

Method Summary

Modifier and Type	Method and Description
static double	calculateClassificationError(MLClassification method, MLDataSet data) Calculate the classification error.
static EncogUtility.FalsePositiveReport	calculatePositiveNegative(MLRegression method, MatrixMLDataSet data)
static double	calculateRegressionError(MLRegression method, MLDataSet data)
static void	convertCSV2Binary(java.io.File csvFile, CSVFormat format, java.io.File binFile, int[] input, int[] ideal, boolean headers)
static void	convertCSV2Binary(java.io.File csvFile, java.io.File binFile, int inputCount, int outputCount, boolean headers) Convert a CSV file to a binary training file.
static void	convertCSV2Binary(java.lang.String csvFile, java.lang.String binFile, int inputCount, int outputCount, boolean headers) Convert a CSV file to a binary training file.
static void	evaluate(MLRegression network, MLDataSet training) Evaluate the network and display (to the console) the output for every value in the training set.
static void	explainErrorMSE(MLRegression method, MatrixMLDataSet training)
static void	explainErrorRMS(MLRegression method, MatrixMLDataSet training)
static java.lang.String	formatNeuralData(MLData data) Format neural data as a list of numbers.
static MLDataSet	loadCSV2Memory(java.lang.String filename, int input, int ideal, boolean headers, CSVFormat format, boolean significance) Load CSV to memory.
static MLDataSet	loadEGB2Memory(java.io.File filename)
static void	saveCSV(java.io.File targetFile, CSVFormat format, MLDataSet set)
static void	saveEGB(java.io.File f, MLDataSet data) Save a training set to an EGB file.
static BasicNetwork	simpleFeedForward(int input, int hidden1, int hidden2, int output, boolean tanh) Create a simple feedforward neural network.
static MLDataSet[]	splitTrainValidate(MLDataSet trainingSet, GenerateRandom rnd, double trainingPercent)
static void	trainConsole(BasicNetwork network, MLDataSet trainingSet, int minutes) Train the neural network, using SCG training, and output status to the console.
static void	trainConsole(MLTrain train, BasicNetwork network, MLDataSet trainingSet, int minutes) Train the network, using the specified training algorithm, and send the output to the console.
static void	trainToError(MLMethod method, MLDataSet dataSet, double error) Train the method, to a specific error, send the output to the console.
static void	trainToError(MLTrain train, double error) Train to a specific error, using the specified training method, send the output to the console.

Methods inherited from class java.lang.Object

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

Method Detail

convertCSV2Binary

public static void convertCSV2Binary(java.io.File csvFile,
                                     java.io.File binFile,
                                     int inputCount,
                                     int outputCount,
                                     boolean headers)

Convert a CSV file to a binary training file.

Parameters:

csvFile - The CSV file.

binFile - The binary file.

inputCount - The number of input values.

outputCount - The number of output values.

headers - True, if there are headers on the3 CSV.

loadCSV2Memory

public static MLDataSet loadCSV2Memory(java.lang.String filename,
                                       int input,
                                       int ideal,
                                       boolean headers,
                                       CSVFormat format,
                                       boolean significance)

Load CSV to memory.

Parameters:

filename - The CSV file to load.

input - The input count.

ideal - The ideal count.

headers - True, if headers are present.

format - The loaded dataset.

significance - True, if there is a significance column.

Returns:

The loaded dataset.

evaluate

public static void evaluate(MLRegression network,
                            MLDataSet training)

Evaluate the network and display (to the console) the output for every value in the training set. Displays ideal and actual.

Parameters:

network - The network to evaluate.

training - The training set to evaluate.

formatNeuralData

public static java.lang.String formatNeuralData(MLData data)

Format neural data as a list of numbers.

Parameters:

data - The neural data to format.

Returns:

The formatted neural data.

simpleFeedForward

public static BasicNetwork simpleFeedForward(int input,
                                             int hidden1,
                                             int hidden2,
                                             int output,
                                             boolean tanh)

Create a simple feedforward neural network.

Parameters:

input - The number of input neurons.

hidden1 - The number of hidden layer 1 neurons.

hidden2 - The number of hidden layer 2 neurons.

output - The number of output neurons.

tanh - True to use hyperbolic tangent activation function, false to use the sigmoid activation function.

Returns:

The neural network.

trainConsole

public static void trainConsole(BasicNetwork network,
                                MLDataSet trainingSet,
                                int minutes)

Train the neural network, using SCG training, and output status to the console.

Parameters:

network - The network to train.

trainingSet - The training set.

minutes - The number of minutes to train for.

trainConsole

public static void trainConsole(MLTrain train,
                                BasicNetwork network,
                                MLDataSet trainingSet,
                                int minutes)

Train the network, using the specified training algorithm, and send the output to the console.

Parameters:

train - The training method to use.

network - The network to train.

trainingSet - The training set.

minutes - The number of minutes to train for.

trainToError

public static void trainToError(MLMethod method,
                                MLDataSet dataSet,
                                double error)

Train the method, to a specific error, send the output to the console.

Parameters:

method - The method to train.

dataSet - The training set to use.

error - The error level to train to.

trainToError

public static void trainToError(MLTrain train,
                                double error)

Train to a specific error, using the specified training method, send the output to the console.

Parameters:

train - The training method.

error - The desired error level.

loadEGB2Memory

public static MLDataSet loadEGB2Memory(java.io.File filename)

convertCSV2Binary

public static void convertCSV2Binary(java.lang.String csvFile,
                                     java.lang.String binFile,
                                     int inputCount,
                                     int outputCount,
                                     boolean headers)

Convert a CSV file to a binary training file.

Parameters:

csvFile - The binary file.

binFile - The binary file.

inputCount - The number of input values.

outputCount - The number of output values.

headers - True, if there are headers on the CSV.

convertCSV2Binary

public static void convertCSV2Binary(java.io.File csvFile,
                                     CSVFormat format,
                                     java.io.File binFile,
                                     int[] input,
                                     int[] ideal,
                                     boolean headers)

calculateRegressionError

public static double calculateRegressionError(MLRegression method,
                                              MLDataSet data)

saveCSV

public static void saveCSV(java.io.File targetFile,
                           CSVFormat format,
                           MLDataSet set)

calculateClassificationError

public static double calculateClassificationError(MLClassification method,
                                                  MLDataSet data)

Calculate the classification error.

Parameters:

method - The method to check.

data - The data to check.

Returns:

The error.

saveEGB

public static void saveEGB(java.io.File f,
                           MLDataSet data)

Save a training set to an EGB file.

Parameters:

f - The file.

data - The data.

explainErrorMSE

public static void explainErrorMSE(MLRegression method,
                                   MatrixMLDataSet training)

explainErrorRMS

public static void explainErrorRMS(MLRegression method,
                                   MatrixMLDataSet training)

calculatePositiveNegative

public static EncogUtility.FalsePositiveReport calculatePositiveNegative(MLRegression method,
                                                                         MatrixMLDataSet data)

splitTrainValidate

public static MLDataSet[] splitTrainValidate(MLDataSet trainingSet,
                                             GenerateRandom rnd,
                                             double trainingPercent)