public void evaluateBestIndividual(IntegerVectorIndividual ind, EvolutionState state)
throws Exception {
int[] genome = ((IntegerVectorIndividual) ind).genome;
// ajustar com parametros "globais"
// ClusWrapper.initialization(Dataset.getPath()+Dataset.getFileName() +
// "-train.arff",Dataset.getPath()+Dataset.getFileName() + "-train.arff", Main.targets,false);
int currentFold = Dataset.getCurrentFold();
String trainSet =
Dataset.getPath() + Dataset.getFileName() + "_fold" + (currentFold) + ".train";
String trainValidSet =
Dataset.getPath() + Dataset.getFileName() + "_fold" + (currentFold) + ".trainvalid";
String valSet = Dataset.getPath() + Dataset.getFileName() + "_fold" + (currentFold) + ".valid";
String testSet = Dataset.getPath() + Dataset.getFileName() + "_fold" + (currentFold) + ".test";
// String trainSet = Dataset.getPath()+Dataset.getFileName() +
// "_train_fold_"+(currentFold+1)+".arff";
// String testSet = Dataset.getPath()+Dataset.getFileName() +
// "_test_fold_"+(currentFold+1)+".arff";
// ClusWrapper.initialization(Dataset.getPath()+Dataset.getFileName() + "-train.arff",
// Dataset.getPath()+Dataset.getFileName() + "-test.arff", Main.targets,Main.randomForest);
myMeasures measures = new myMeasures();
if (Main.mlTask > 0) {
double auroc[] = new double[3];
double auprc[] = new double[3];
/*if (Main.mlTask == 2) { // multi-threads)
ClusWrapperNonStatic object = new ClusWrapperNonStatic();
object.initialization(trainSet,valSet, Main.targets,Main.randomForest,true);
measures = object.evaluateIndividualClassification(genome,true);
}
else { // ==1, single thread
// true = classification or false = regression
*/
ClusWrapper.clus = null;
ClusWrapper.initialization(trainSet, valSet, Main.targets, Main.randomForest, true);
measures = ClusWrapper.evaluateIndividualClassification(genome, true);
// }
auroc[1] = measures.getAUROC()[1];
auprc[1] = measures.getAUPRC()[1];
/*if (Main.mlTask == 2) { // multi-threads)
ClusWrapperNonStatic object = new ClusWrapperNonStatic();
object.initialization(trainValidSet,testSet, Main.targets,Main.randomForest,true);
measures = object.evaluateIndividualClassification(genome,true);
}
else { // ==1, single thread
*/
// true = classification or false = regression
ClusWrapper.clus =
null; // we really need to do this! otherwise, cluswrapper does not re-read the
// corresponding partition.
ClusWrapper.initialization(trainValidSet, testSet, Main.targets, Main.randomForest, true);
measures = ClusWrapper.evaluateIndividualClassification(genome, true);
// }
auroc[0] = measures.getAUROC()[0];
auprc[0] = measures.getAUPRC()[0];
auroc[2] = measures.getAUROC()[1];
auprc[2] = measures.getAUPRC()[1];
;
/* System.err.println("trainValidSet: "+ trainValidSet);
System.err.println("testSet: "+ testSet);
System.err.println("auroc[0]: "+ auroc[0]);
System.err.println("auroc[2]: "+ auroc[2]);
*/
for (int i = 0; i < 3; i++) {
Main.measuresSingle[i][0][Dataset.getCurrentFold()][(Integer) state.job[0]] = auroc[i];
Main.measuresSingle[i][1][Dataset.getCurrentFold()][(Integer) state.job[0]] = auprc[i];
}
// Printing results
String test = new String();
for (int i = 0; i < 2; i++)
test =
test
+ Main.measuresSingle[2][i][Dataset.getCurrentFold()][(Integer) state.job[0]]
+ ",";
Main.pwTest.println(test);
String full = new String();
for (int i = 0; i < 2; i++) {
for (int j = 0; j < 3; j++) {
full =
full
+ Main.measuresSingle[j][i][Dataset.getCurrentFold()][(Integer) state.job[0]]
+ ",";
}
}
Main.pwAll.println(full);
} else {
double mae[] = new double[3];
double mse[] = new double[3];
double rmse[] = new double[3];
// true = classification or false = regression
ClusWrapper.initialization(trainSet, valSet, Main.targets, Main.randomForest, false);
measures = ClusWrapper.evaluateIndividual(genome, true);
mae[1] = measures.getMAE()[1];
mse[1] = measures.getMSE()[1];
rmse[1] = measures.getRMSE()[1]; // wrmse[1] = measures.getWRMSE()[0];
ClusWrapper.initialization(trainValidSet, testSet, Main.targets, Main.randomForest, false);
measures = ClusWrapper.evaluateIndividual(genome, true);
mae[0] = measures.getMAE()[0];
mse[0] = measures.getMSE()[0];
rmse[0] = measures.getRMSE()[0]; // wrmse[0] = measures.getWRMSE()[0];
mae[2] = measures.getMAE()[1];
mse[2] = measures.getMSE()[1];
rmse[2] = measures.getRMSE()[1]; // wrmse[2] = measures.getWRMSE()[1];
for (int i = 0; i < 3; i++) {
Main.measuresSingle[i][0][Dataset.getCurrentFold()][(Integer) state.job[0]] = mae[i];
Main.measuresSingle[i][1][Dataset.getCurrentFold()][(Integer) state.job[0]] = mse[i];
Main.measuresSingle[i][2][Dataset.getCurrentFold()][(Integer) state.job[0]] = rmse[i];
}
// Printing results
String test = new String();
for (int i = 0; i < 3; i++)
test =
test
+ Main.measuresSingle[2][i][Dataset.getCurrentFold()][(Integer) state.job[0]]
+ ",";
Main.pwTest.println(test);
String full = new String();
for (int i = 0; i < 3; i++) {
for (int j = 0; j < 3; j++) {
full =
full
+ Main.measuresSingle[j][i][Dataset.getCurrentFold()][(Integer) state.job[0]]
+ ",";
}
}
Main.pwAll.println(full);
}
// Main.pEvolution.println(mae[0] +","+ mse[0] +","+ rmse[0] +","+ mae[1] +","+ mse[1] +","+
// rmse[1]);
}
