// adatok: http://repository.seasr.org/Datasets/UCI/arff/
// Tutorial: http://weka.wikispaces.com/Use+Weka+in+your+Java+code
// további info: Weka API 

import java.io.BufferedReader;
import java.io.FileReader;


import weka.classifiers.Classifier;
import weka.classifiers.Evaluation;
import weka.classifiers.trees.J48;
import weka.clusterers.ClusterEvaluation;
import weka.clusterers.Clusterer;
import weka.clusterers.SimpleKMeans;
import weka.core.Instances;
import weka.filters.Filter;


public class HelloWekaWorld {

	// Ez a metodus "kihamozza" a klaszterezes eredmenyekent kapott stringbol 
	// a helytelenul klaszterezett peldanyok szamat, es visszaadja ezt szamkent
	public static float getIncorrectlyClusteredInstances(String clustering_result) {
		String pattern = "Incorrectly clustered instances :";
		String s = clustering_result.substring(
			clustering_result.indexOf(pattern)+pattern.length());
		s = s.trim();
		return new Float(s.split("\t")[0]);
	}

	public static void main(String[] args) throws Exception {
		
		Instances data = new Instances(new BufferedReader(
		                 new FileReader(" (...) \\breast-cancer.arff")));  
		data.setClassIndex(data.numAttributes()-1);
		
		Instances traindata = data.trainCV(10, 0);
		Instances testdata  = data.testCV(10, 0);
	
		Classifier c = new J48();
		((J48) c).setReducedErrorPruning(true);
		((J48) c).setMinNumObj(5);
		c.buildClassifier(traindata);
		
		for (int i=0;i<testdata.numInstances();i++) {
			double predicted_class = c.classifyInstance(testdata.instance(i));
			System.out.println(i+"  "+predicted_class);
		}
		
		
		
		 Evaluation eval = new Evaluation(traindata);
		 eval.evaluateModel(c, testdata);
		 System.out.println(eval.toSummaryString("\nResults\n======\n", false));
		
		 for (int i=0;i<data.numClasses();i++) {
			 System.out.println("Area Under ROC Curve (class="+i+")\t"+
                                          eval.areaUnderROC(i));
		 }
		 
		 System.out.println("Accuracy:\t"+ (eval.pctCorrect()/100) );
		 
		 

		 // Klaszterezés
		 
		 weka.filters.unsupervised.attribute.Remove filter = 
		     new weka.filters.unsupervised.attribute.Remove();
		 filter.setAttributeIndices("" + (data.classIndex() + 1));
		 filter.setInputFormat(data);
		 Instances dataClusterer = Filter.useFilter(data, filter);
		 
		 Clusterer clus = new SimpleKMeans();
		 //( (SimpleKMeans) clus).setDistanceFunction(new weka.core.ChebyshevDistance());
		 ( (SimpleKMeans) clus).setNumClusters(2);
		 
		 /* Szintén kipróbálni:
		    Clusterer clus = new HierarchicalClusterer();
		 	((HierarchicalClusterer)clus).setOptions(new String[] 
			  { "-N", "3", "-L", "SINGLE", "-A", "weka.core.ChebyshevDistance" });
		 */
		 // Link options: SINGLE COMPLETE AVERAGE
		 // Distance functions: ChebyshevDistance, EuclideanDistance, ManhattanDistance
		 
		 clus.buildClusterer(dataClusterer);

		 ClusterEvaluation eval1 = new ClusterEvaluation();
		 eval1.setClusterer(clus);
		 eval1.evaluateClusterer(data);
		 System.out.println(eval1.clusterResultsToString());
		 
		 

		 // Igy tudjuk szamkent, a clustering_error elnevezesu float tipusu 
		 // valtozoban megkapni a helytelenul klaszterezett peldanyok szamat 
		 String cluster_results = eval1.clusterResultsToString();
		 float clustering_error = getIncorrectlyClusteredInstances(cluster_results);
		 System.out.println(clustering_error);

		 
		 // Asszociációs szabályok
		 
		 Apriori a = new Apriori();
		 a.setOptions(new String[] { "-M", "0.3", "-C", "0.7", "-N", "10000"});
		 a.buildAssociations(data);
		 System.out.println(a.toString());
		 
		 
		 FastVector[] rules = a.getAllTheRules();
		 int[] items = ((AprioriItemSet) rules[0].elementAt(0)).items();
		 for (int i=0;i<items.length;i++) System.out.println( items[i] );	 
	}
}

// További feladatok: 
//    - más osztályozók kipróbálása, 
//    - 10-fold crossvalidation-t végéigcsinálni, 
//    - hiperparaméterkeresés