keel.Algorithms.Preprocess.Basic

Class Metodo

java.lang.Object

keel.Algorithms.Preprocess.Basic.Metodo

Direct Known Subclasses:

ADASYN, ADOMS, AHCClustering, AllKNN, AllKNN, Borderline_SMOTE, BSE, BSE, CCIS, CCIS, CHC, CHC, CNN, CNN, CNN, CNN_TomekLinks, CoCoIS, CoCoIS, CPM, CPruner, CPruner, Depur, DROP1, DROP1, DROP2, DROP2, DROP3, DROP3, DROP3LVQ3, DROP3PSO, DROP3SFLSDE, EHS_CHC, ENN, ENN, ENNRS, ENNRS, ENNTh, ENNTh, ENRBF, ENRBF, Explore, Explore, FCNN, FCNN, GA_MSE_CC_FSM, GA_MSE_CC_FSM, GCNN, GCNN, GG, GG, GGA, GGA, HMNEI, HMNEI, IB2, IB2, IB3, IB3, ICF, ICF, ICFLVQ3, ICFPSO, ICFSFLSDE, IGA, IGA, IKNN, IKNN, KNN, LVQ, MCNN, MCNN, MCS, MCS, MENN, MENN, MNV, MNV, ModelCS, ModelCS, MSS, MSS, Multiedit, Multiedit, NCL, NCNEdit, NCNEdit, NRMCS, NRMCS, OSS, PBIL, PBIL, POP, POP, PSC, PSC, PSRCG, PSRCG, RandomOverSampling, RandomUnderSampling, Reconsistent, Reconsistent, RENN, RENN, RMHC, RMHC, RNG, RNG, RNN, RNN, Safe_Level_SMOTE, SBC, SGA, SGA, Shrink, Shrink, SMOTE, SMOTE, SMOTE_ENN, SMOTE_RSB, SMOTE_TomekLinks, SNN, SNN, SPIDER, SPIDER2, SSMA, SSMA, SSMALVQ3, SSMAPSO, SSMASFLSDE, SVBPS, SVBPS, TCNN, TCNN, TomekLinks, TRKNN, TRKNN, VSM, VSM, ZhangTS, ZhangTS

public class Metodo
extends java.lang.Object

File: Metodo.java An auxiliary class to initialize Instance Selection algorithms

Since:

JDK1.5

Version:

0.1

Author:

Written by Salvador García (University of Granada) 20/07/2004, Modified by Isaac Triguero (University of Granada) 22/06/2010, Modified by Victoria Lopez (University of Granada) 21/09/2010

Field Summary

Fields

Modifier and Type	Field and Description
protected int[]	clasesTest Test output data.
protected int[]	clasesTrain Training output data.
protected double[][]	datosTest Test input data.
protected double[][]	datosTrain Training input data.
protected boolean	distanceEu True if euclidean distance is used, false if HVDM is used.
protected Attribute[]	entradas Inputs attributes
protected java.lang.String[]	ficheroSalida Output files names
protected java.lang.String	ficheroTest Test file name
protected java.lang.String	ficheroTraining Train file name
protected java.lang.String	ficheroValidation Reference file name
protected int	nEntradas Number of input attributes
protected static double[][][]	nominalDistance Nominal distances.
protected int[][]	nominalTrain Training Nominal input data.
protected boolean[][]	nulosTrain Training null data.
protected double[][]	realTest Test input data.
protected double[][]	realTrain Training input data.
protected java.lang.String	relation Relation string.
protected Attribute	salida Output attribute
protected static double[]	stdDev Standard deviation.
protected InstanceSet	test Test dataset
protected InstanceSet	training Training dataset

Constructor Summary

Constructors

Constructor and Description
Metodo() Default builder
Metodo(java.lang.String ficheroScript) Builder.
Metodo(java.lang.String ficheroScript, InstanceSet train) Parameter constructor.

Method Summary

Modifier and Type	Method and Description
void	leerConfiguracion(java.lang.String ficheroScript) Reads the parameters of the algorithm.
protected void	normalizar() This function builds the data matrix for reference data and normalizes inputs values

Methods inherited from class java.lang.Object

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

Field Detail

ficheroTraining

protected java.lang.String ficheroTraining

Train file name

ficheroValidation

protected java.lang.String ficheroValidation

Reference file name

ficheroTest

protected java.lang.String ficheroTest

Test file name

ficheroSalida

protected java.lang.String[] ficheroSalida

Output files names

training

protected InstanceSet training

Training dataset

test

protected InstanceSet test

Test dataset

entradas

protected Attribute[] entradas

Inputs attributes

salida

protected Attribute salida

Output attribute

nEntradas

protected int nEntradas

Number of input attributes

relation

protected java.lang.String relation

Relation string.

datosTrain

protected double[][] datosTrain

Training input data.

clasesTrain

protected int[] clasesTrain

Training output data.

datosTest

protected double[][] datosTest

Test input data.

realTest

protected double[][] realTest

Test input data.

clasesTest

protected int[] clasesTest

Test output data.

nulosTrain

protected boolean[][] nulosTrain

Training null data.

nominalTrain

protected int[][] nominalTrain

Training Nominal input data.

realTrain

protected double[][] realTrain

Training input data.

distanceEu

protected boolean distanceEu

True if euclidean distance is used, false if HVDM is used.

nominalDistance

protected static double[][][] nominalDistance

Nominal distances.

stdDev

protected static double[] stdDev

Standard deviation.

Constructor Detail

Metodo

public Metodo()

Default builder

Metodo

public Metodo(java.lang.String ficheroScript)

Builder. Creates the basic structures of the algorithm

Parameters:

ficheroScript - Configuration script

Metodo

public Metodo(java.lang.String ficheroScript,
              InstanceSet train)

Parameter constructor. Construct the main class of the algorithm by reading the configuration file and using as training and test, the dataset given as a parameter.

Parameters:

ficheroScript - configuration filename.

train - given dataset.

Method Detail

normalizar

protected void normalizar()
                   throws CheckException

This function builds the data matrix for reference data and normalizes inputs values

Throws:

CheckException - if the dataset is not appropiate.

leerConfiguracion

public void leerConfiguracion(java.lang.String ficheroScript)

Reads the parameters of the algorithm. Must be implemented in the subclass.

Parameters:

ficheroScript - Configuration script