/***********************************************************************
This file is part of KEEL-software, the Data Mining tool for regression,
classification, clustering, pattern mining and so on.
Copyright (C) 2004-2010
F. Herrera (herrera@decsai.ugr.es)
L. Sánchez (luciano@uniovi.es)
J. Alcalá-Fdez (jalcala@decsai.ugr.es)
S. García (sglopez@ujaen.es)
A. Fernández (alberto.fernandez@ujaen.es)
J. Luengo (julianlm@decsai.ugr.es)
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
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 General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see http://www.gnu.org/licenses/
**********************************************************************/
//
// SSMA.javA HIBRIDO PSO
//
// Salvador Garc�a L�pez
//
// Created by Salvador Garc�a L�pez 3-10-2005.
// Copyright (c) 2004 __MyCompanyName__. All rights reserved.
//
package keel.Algorithms.Instance_Generation.DROP3PSO;
import keel.Algorithms.Preprocess.Basic.*;
import keel.Algorithms.Instance_Generation.Basic.PrototypeGenerator;
import keel.Algorithms.Instance_Generation.Basic.Prototype;
import keel.Algorithms.Instance_Generation.Basic.PrototypeGenerationAlgorithm;
import keel.Algorithms.Instance_Generation.Basic.PrototypeSet;
import keel.Algorithms.Instance_Generation.utilities.*;
//import keel.Algorithms.Instance_Generation.utilities.KNN.KNN;
import keel.Dataset.Attributes;
import keel.Dataset.InstanceAttributes;
import keel.Dataset.InstanceSet;
import org.core.*;
import java.util.StringTokenizer;
import java.util.Arrays;
import java.util.Vector;
public class DROP3PSO extends Metodo {
/*Own parameters of the algorithm*/
private int k;
private double semilla;
public String Script; // para releer par�metros..
private PrototypeSet trainingDataSet;
private PrototypeGenerator generador;
//Par�metros PSO
private int SwarmSize; // SwarmSize == P
private int ParticleSize; // ParticleSize == K (in the article)
private int MaxIter;
private double C1;
private double C2;
private double VMax;
private double Wstart;
private double Wend;
protected int numberOfClass;
protected int numberOfPrototypes; // Particle size is the percentage
protected int numberOfStrategies; // number of strategies in the pool
public DROP3PSO (String ficheroScript) {
super (ficheroScript);
}
/**
* Reads the prototype set from a data file.
* @param nameOfFile Name of data file to be read.
* @return PrototypeSet built with the data of the file.
*/
public static PrototypeSet readPrototypeSet(String nameOfFile)
{
Attributes.clearAll();//BUGBUGBUG
InstanceSet training = new InstanceSet();
try
{
//System.out.print("PROBANDO:\n"+nameOfFile);
training.readSet(nameOfFile, true);
training.setAttributesAsNonStatic();
InstanceAttributes att = training.getAttributeDefinitions();
Prototype.setAttributesTypes(att);
}
catch(Exception e)
{
System.err.println("readPrototypeSet has failed!");
e.printStackTrace();
}
return new PrototypeSet(training);
}
/**
* Implements the 1NN algorithm
* @param current Prototype which the algorithm will find its nearest-neighbor.
* @param dataSet Prototype set in which the algorithm will search.
* @return Nearest prototype to current in the prototype set dataset.
*/
public static Prototype _1nn(Prototype current, PrototypeSet dataSet)
{
Prototype nearestNeighbor = dataSet.get(0);
int indexNN = 0;
//double minDist = Distance.dSquared(current, nearestNeighbor);
//double minDist = Distance.euclideanDistance(current, nearestNeighbor);
double minDist =Double.POSITIVE_INFINITY;
double currDist;
int _size = dataSet.size();
// System.out.println("****************");
// current.print();
for (int i=0; i<_size; i++)
{
Prototype pi = dataSet.get(i);
//if(!current.equals(pi))
//{
// double currDist = Distance.dSquared(current, pi);
currDist = Distance.euclideanDistance(pi,current);
// System.out.println(currDist);
if(currDist >0){
if (currDist < minDist)
{
minDist = currDist;
// nearestNeighbor = pi;
indexNN =i;
}
}
//}
}
// System.out.println("Min dist =" + minDist + " Vecino Cercano = "+ indexNN);
return dataSet.get(indexNN);
}
public double classficationAccuracy1NN(PrototypeSet training, PrototypeSet test)
{
int wellClassificated = 0;
for(Prototype p : test)
{
Prototype nearestNeighbor = _1nn(p, training);
if(p.getOutput(0) == nearestNeighbor.getOutput(0))
++wellClassificated;
}
return 100.0* (wellClassificated / (double)test.size());
}
/**
* Generate a reduced prototype set by the PSOGenerator method.
* @return Reduced set by PSOGenerator's method.
*/
public PrototypeSet reduceSet(PrototypeSet initial)
{
System.out.print("\nThe algorithm is starting...\n Computing...\n");
//Algorithm
// First, we create the population, with SwarmSize.
// like a prototypeSet's vector.
PrototypeSet population [] = new PrototypeSet [SwarmSize];
PrototypeSet mejorPosicion [] = new PrototypeSet [SwarmSize];
PrototypeSet nominalPopulation = new PrototypeSet();
double fitness[] = new double[SwarmSize];
double fitness_bestPopulation[] = new double[SwarmSize];
PrototypeSet bestParticle = new PrototypeSet();
double inertia = ((Wstart-Wend)*(MaxIter))/ (MaxIter + Wend);
int mejorParticula =0; // The best particle in the population
double aleatorio;
//Each particle must have Particle Size %
//Initialization.
population[0]= new PrototypeSet(initial) ;
generador = new PrototypeGenerator(trainingDataSet);
nominalPopulation = new PrototypeSet();
nominalPopulation.formatear(population[0]);
fitness[0] = classficationAccuracy1NN(nominalPopulation,trainingDataSet);
this.numberOfClass = trainingDataSet.getPosibleValuesOfOutput().size();
System.out.println("Best initial fitness = "+ fitness[0]);
fitness_bestPopulation[0] = fitness[0];
for(int i=1; i< SwarmSize; i++){
population[i] = new PrototypeSet();
for(int j=0; j< population[0].size(); j++){
Prototype aux = new Prototype(trainingDataSet.getFromClass(population[0].get(j).getOutput(0)).getRandom());
population[i].add(aux);
}
nominalPopulation = new PrototypeSet();
nominalPopulation.formatear(population[i]);
fitness[i] = classficationAccuracy1NN(population[i],trainingDataSet); // PSOfitness
fitness_bestPopulation[i] = fitness[i]; // Initially the same fitness.
}
//We select the best initial particle
double bestFitness=fitness[0];
int bestFitnessIndex=0;
for(int i=1; i< SwarmSize;i++){
if(fitness[i]>bestFitness){
bestFitness = fitness[i];
bestFitnessIndex=i;
}
}
for(int j=0;j<SwarmSize;j++){
mejorPosicion[j] = population[j].clone(); // hard-copy.Save the best position of the particle.
//Initial mejorPosicion = initial population.
//Now, I establish the index of each prototype.
for(int i=0; i<population[j].size(); ++i)
population[j].get(i).setIndex(i);
}
double velocidad[][][] = new double[SwarmSize][][]; // tri-dimensional vector
int num_atribs = population[0].get(0).numberOfInputs();
for(int i=0; i<SwarmSize;i++){
velocidad[i]= new double[population[0].size()][]; // velocity matrix.
// Initially there is no velocity, no memory..
for(int j=0; j<population[0].size();j++){
velocidad[i][j] = new double[num_atribs];
for(int k = 0; k<num_atribs;k++){
velocidad[i][j][k] = RandomGenerator.Randdouble(-VMax, VMax)*1. ; // the initial velocity, a random number between -Vmax , Vmax
// System.out.println(velocidad[i][j][k]);
}
}
}
for(int iter=0; iter< MaxIter; iter++){ // Main loop
for(int i=0; i< SwarmSize; i++){
for(int k = 0; k< population[i].size();k++){
Prototype resta = mejorPosicion[i].get(k).sub(population[i].get(k));
Prototype restaBestParticle = mejorPosicion[bestFitnessIndex].get(k).sub(population[i].get(k));
for(int j=0; j< num_atribs ; j++){
velocidad[i][k][j]= inertia * velocidad[i][k][j] ; // Memory velocity.
aleatorio =RandomGenerator.Randdouble(0, 1) ;
velocidad[i][k][j]+= C1*aleatorio* resta.getInput(j) ; // Cognition part.
aleatorio =RandomGenerator.Randdouble(0, 1) ;
velocidad[i][k][j]+= C2*aleatorio * restaBestParticle.getInput(j) ; // Social part.
//System.out.print(aleatorio + "\t");
// Then we do xi = xi + vi.
if(velocidad[i][k][j]>VMax){
velocidad[i][k][j] = VMax; // The particles's velocities has a maximum velocity.
}else if(velocidad[i][k][j]< -VMax){
velocidad[i][k][j]=-VMax; // absolute value. �? or -VMax , Vmax. ?
}
//System.out.print("\nVelocidad ="+ velocidad[i][k][j] + "\n");
// System.out.print("\nvalor= "+ population[i].get(k).getInput(j)+ "\n");
double suma = population[i].get(k).getInput(j) + velocidad[i][k][j]*1.;
//if(suma>1) suma = 1;
//else if( suma<0) suma = 0; // Establish the normalize limits [0,1]
//System.out.print("\nSuma= "+ suma+ "\n");
population[i].get(k).setInput(j,suma); // We add the velocity to the attribute
population[i].get(k).applyThresholds();
}
}
}
//Now we have xi = xi + vi.for all particles.
// Particles has changed, We must calculate fitness and compare all.
for(int i=0; i< SwarmSize; i++){
/*
if(k<=population[i].size())
fitness[i] = absoluteclassficationAccuracy1NNKNN(population[i], trainingDataSet,k); // PSO fitness
else
fitness[i] = absoluteclassficationAccuracy1NNKNN(population[i],trainingDataSet,population[i].size());
*/
// Antes de calcular el fitness, tengo que "transformar los datos nominales.."
nominalPopulation = new PrototypeSet();
nominalPopulation.formatear(population[i]);
fitness[i] = classficationAccuracy1NN(nominalPopulation,trainingDataSet);
//fitness[i] = classficationAccuracy1NN(population[i],trainingDataSet);
}
for(int i=0; i< SwarmSize;i++){
// Where is the best?
if(fitness[i]>bestFitness){
bestFitness = fitness[i];
bestFitnessIndex=i;
}
//Save the best particles!
if(fitness[i]>fitness_bestPopulation[i]){
fitness_bestPopulation[i] = fitness[i];
mejorPosicion[i] = population[i].clone(); // Hard Copy.
}
}
//Calculate the new inertia.
inertia = ((Wstart-Wend)*(MaxIter-iter))/ (MaxIter + Wend);
}
System.err.println("Best Fitness "+ bestFitness);
nominalPopulation = new PrototypeSet();
nominalPopulation.formatear(mejorPosicion[bestFitnessIndex]);
System.err.println("\n% de acierto en training Nominal " + classficationAccuracy1NN(nominalPopulation,trainingDataSet));
return nominalPopulation;
}
/* MEzcla de algoritmos */
public void ejecutar () {
int i, j, l, m, n, o;
int nClases;
int claseObt;
boolean marcas[];
int nSel;
double conjS[][];
double conjR[][];
int conjN[][];
boolean conjM[][];
int clasesS[];
int vecinos[][];
Vector asociados[];
int aciertosSin;
int vecinosTemp[];
double distTemp[];
double dist, bestD;
boolean parar;
Referencia orden[];
int mayoria;
long tiempo = System.currentTimeMillis();
/*Getting the number of different classes*/
nClases = 0;
for (i=0; i<clasesTrain.length; i++)
if (clasesTrain[i] > nClases)
nClases = clasesTrain[i];
nClases++;
/*Inicialization of the instance flagged vector of the S set*/
marcas = new boolean[datosTrain.length];
for (i=0; i<datosTrain.length; i++) {
marcas[i] = true;
}
nSel = datosTrain.length;
/*Do ENN before sorting*/
for (i=0; i<datosTrain.length; i++) {
claseObt = KNN.evaluacionKNN2 (k, datosTrain, realTrain, nominalTrain, nulosTrain, clasesTrain, datosTrain[i], realTrain[i], nominalTrain[i], nulosTrain[i], nClases, distanceEu);
if (claseObt != clasesTrain[i]) { //is included in the solution set if it is agree with your mayority
marcas[i] = false;
nSel--;
}
}
/*Construction of an instance vector with distances to the nearest enemy*/
orden = new Referencia[datosTrain.length];
for (i=0; i<datosTrain.length; i++) {
bestD = Double.POSITIVE_INFINITY;
for (j=0; j<datosTrain.length; j++) {
if (clasesTrain[i] != clasesTrain[j]) {
dist = KNN.distancia (datosTrain[i], realTrain[i], nominalTrain[i], nulosTrain[i], datosTrain[j], realTrain[j], nominalTrain[j], nulosTrain[j], distanceEu);
if (dist < bestD)
bestD = dist;
}
}
orden[i] = new Referencia (i, bestD);
}
/*Sorting the previous vector*/
Arrays.sort(orden);
/*Inicialization of the data structures of neighbors and associates*/
distTemp = new double[k+1];
vecinosTemp = new int[k+1];
vecinos = new int[datosTrain.length][k+1];
asociados = new Vector[datosTrain.length];
for (i=0; i<datosTrain.length; i++)
asociados[i] = new Vector ();
/*Body of the algorithm DROP3 (same as DROP2).*/
for (i=0; i<datosTrain.length; i++) {
/*Get the k+1 nearest neighbors of each instance*/
if (marcas[i]) {
KNN.evaluacionKNN2 (k+1, datosTrain, realTrain, nominalTrain, nulosTrain, clasesTrain, datosTrain[i], realTrain[i], nominalTrain[i], nulosTrain[i], nClases, distanceEu, vecinos[i]);
for (j=0; j<vecinos[i].length; j++) {
if (vecinos[i][j] >= 0)
asociados[vecinos[i][j]].addElement (new Referencia (i,0));
}
}
}
/*Check if delete or not the instances considering the WITH and WITHOUT sets*/
for (o=0; o<datosTrain.length; o++){
i = orden[o].entero;
if (marcas[i]) { //only for instances haven�t noise filtered
aciertosSin = 0;
marcas[i] = false;
nSel--;
/*Construction of S set from the temporaly flags*/
conjS = new double[nSel][datosTrain[0].length];
conjR = new double[nSel][datosTrain[0].length];
conjN = new int[nSel][datosTrain[0].length];
conjM = new boolean[nSel][datosTrain[0].length];
clasesS = new int[nSel];
for (m=0, l=0; m<datosTrain.length; m++) {
if (marcas[m]) { //the instance will evaluate
for (j=0; j<datosTrain[0].length; j++) {
conjS[l][j] = datosTrain[m][j];
conjR[l][j] = realTrain[m][j];
conjN[l][j] = nominalTrain[m][j];
conjM[l][j] = nulosTrain[m][j];
}
clasesS[l] = clasesTrain[m];
l++;
}
}
marcas[i] = true;
nSel++;
/*Evaluation of associates without the instance in T*/
for (j=0; j<k+1; j++) {
if (vecinos[i][j] >= 0) {
claseObt = KNN.evaluacionKNN2 (k, conjS, conjR, conjN, conjM, clasesS, datosTrain[vecinos[i][j]], realTrain[vecinos[i][j]], nominalTrain[vecinos[i][j]], nulosTrain[vecinos[i][j]], nClases, distanceEu);
if (claseObt == clasesTrain[vecinos[i][j]]) //classify it correctly
aciertosSin++;
}
}
mayoria = (k+1) / 2;
if (aciertosSin > mayoria) {
/*Delete P from S*/
marcas[i] = false;
nSel--;
/*For each associate of P, search a new nearest neighbor*/
for (j=0; j<asociados[i].size(); j++) {
for (l=0; l<k+1; l++) {
vecinosTemp[l] = vecinos[((Referencia)(asociados[i].elementAt(j))).entero][l];
vecinos[((Referencia)(asociados[i].elementAt(j))).entero][l] = -1;
distTemp[l] = Double.POSITIVE_INFINITY;
}
for (l=0; l<datosTrain.length; l++) {
if (marcas[l]) { //it is in S
dist = KNN.distancia (datosTrain[((Referencia)(asociados[i].elementAt(j))).entero], realTrain[((Referencia)(asociados[i].elementAt(j))).entero], nominalTrain[((Referencia)(asociados[i].elementAt(j))).entero], nulosTrain[((Referencia)(asociados[i].elementAt(j))).entero], datosTrain[l], realTrain[l], nominalTrain[l], nulosTrain[l], distanceEu);
parar = false;
/*Calculate the nearest neighbors in this situation again*/
for (m=0; m<(k+1) && !parar; m++) {
if (dist < distTemp[m]) {
parar = true;
for (n=m+1; n<k+1; n++) {
distTemp[n] = distTemp[n-1];
vecinos[((Referencia)(asociados[i].elementAt(j))).entero][n] = vecinos[((Referencia)(asociados[i].elementAt(j))).entero][n-1];
}
distTemp[m] = dist;
vecinos[((Referencia)(asociados[i].elementAt(j))).entero][m] = l;
}
}
}
}
/*Add to the list of associates of the new neighbor this instance*/
for (l=0; l<k+1; l++) {
parar = false;
for (m=0; m<asociados[vecinosTemp[l]].size() && !parar; m++) {
if (((Referencia)(asociados[vecinosTemp[l]].elementAt(m))).entero == ((Referencia)(asociados[i].elementAt(j))).entero
&& vecinosTemp[l] != i) {
asociados[vecinosTemp[l]].removeElementAt(m);
parar = true;
}
}
}
for (l=0; l<k+1; l++) {
asociados[vecinos[((Referencia)(asociados[i].elementAt(j))).entero][l]].addElement(new Referencia (((Referencia)(asociados[i].elementAt(j))).entero,0));
}
}
}
}
}
/*Construction of the S set from the flags*/
conjS = new double[nSel][datosTrain[0].length];
conjR = new double[nSel][datosTrain[0].length];
conjN = new int[nSel][datosTrain[0].length];
conjM = new boolean[nSel][datosTrain[0].length];
clasesS = new int[nSel];
for (m=0, l=0; m<datosTrain.length; m++) {
if (marcas[m]) { //the instance will evaluate
for (j=0; j<datosTrain[0].length; j++) {
conjS[l][j] = datosTrain[m][j];
conjR[l][j] = realTrain[m][j];
conjN[l][j] = nominalTrain[m][j];
conjM[l][j] = nulosTrain[m][j];
}
clasesS[l] = clasesTrain[m];
l++;
}
}
System.out.println("DROP3 "+ relation + " " + (double)(System.currentTimeMillis()-tiempo)/1000.0 + "s");
OutputIS.escribeSalida(ficheroSalida[0], conjR, conjN, conjM, clasesS, entradas, salida, nEntradas, relation);
OutputIS.escribeSalida(ficheroSalida[1], test, entradas, salida, nEntradas, relation);
/** AHORA A�ADO MI PSO!! **/
Parameters.assertBasicArgs(ficheroSalida);
PrototypeGenerationAlgorithm.readParametersFile(this.Script);
PrototypeGenerationAlgorithm.printParameters();
PrototypeSet training = readPrototypeSet(ficheroSalida[0]);
// training.print(); // Conjunto devuelto POR SSMA
trainingDataSet = readPrototypeSet(this.ficheroTraining); // Conjunto inicial
PrototypeSet SADE = reduceSet(training); // LLAMO al SADE
SADE.save(ficheroSalida[0]); // Lo guardo
// COn conjS me vale.
int trainRealClass[][];
int trainPrediction[][];
trainRealClass = new int[datosTrain.length][1];
trainPrediction = new int[datosTrain.length][1];
//Working on training
for ( i=0; i<datosTrain.length; i++) {
trainRealClass[i][0] = clasesTrain[i];
trainPrediction[i][0] = KNN.evaluate(datosTrain[i],SADE.prototypeSetTodouble(), nClases, SADE.getClases(), 1);
}
KNN.writeOutput(ficheroSalida[0], trainRealClass, trainPrediction, entradas, salida, relation);
//Working on test
int realClass[][] = new int[datosTest.length][1];
int prediction[][] = new int[datosTest.length][1];
//Check time
for (i=0; i<realClass.length; i++) {
realClass[i][0] = clasesTest[i];
prediction[i][0]= KNN.evaluate(datosTest[i],SADE.prototypeSetTodouble(), nClases, SADE.getClases(), 1);
}
KNN.writeOutput(ficheroSalida[1], realClass, prediction, entradas, salida, relation);
}
public void leerConfiguracion (String ficheroScript) {
String fichero, linea, token;
StringTokenizer lineasFichero, tokens;
byte line[];
int i, j;
ficheroSalida = new String[2];
fichero = Fichero.leeFichero (ficheroScript);
lineasFichero = new StringTokenizer (fichero,"\n\r");
lineasFichero.nextToken();
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
token = tokens.nextToken();
/*Getting the name of training and test files*/
line = token.getBytes();
for (i=0; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroTraining = new String (line,i,j-i);
for (i=j+1; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroValidation = new String (line,i,j-i);
for (i=j+1; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroTest = new String (line,i,j-i);
//Parameters.assertBasicArgs(ficheroSalida);
/*Obtainin the path and the base name of the results files*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
token = tokens.nextToken();
/*Getting the name of output files*/
line = token.getBytes();
for (i=0; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroSalida[0] = new String (line,i,j-i);
for (i=j+1; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroSalida[1] = new String (line,i,j-i);
/*Getting the seed*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
semilla = Long.parseLong(tokens.nextToken().substring(1));
/*Getting the number of neighbors*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
k = Integer.parseInt(tokens.nextToken().substring(1));
/*Getting the type of distance function*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
distanceEu = tokens.nextToken().substring(1).equalsIgnoreCase("Euclidean")?true:false;
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.SwarmSize = Integer.parseInt(tokens.nextToken().substring(1));
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.MaxIter = Integer.parseInt(tokens.nextToken().substring(1));
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.C1 = Double.parseDouble(tokens.nextToken().substring(1));
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.C2 = Double.parseDouble(tokens.nextToken().substring(1));
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.VMax = Double.parseDouble(tokens.nextToken().substring(1));
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.Wstart = Double.parseDouble(tokens.nextToken().substring(1));
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
this.Wend = Double.parseDouble(tokens.nextToken().substring(1));
System.out.print("\nIsaac dice: Swar= "+SwarmSize+ " Maxiter= "+ MaxIter+" Wend= "+this.Wend+ "\n");
}
}