PruningCriteriaTest.java

package func.test;

import shared.DataSet;
import shared.DataSetDescription;
import shared.Instance;
import func.dtree.DecisionTreeSplit;
import func.dtree.DecisionTreeSplitStatistics;
import func.dtree.ChiSquarePruningCriteria;
import func.dtree.StandardDecisionTreeSplit;

/**
 * Test the class
 * @author Andrew Guillory gtg008g@mail.gatech.edu
 * @version 1.0
 */
public class PruningCriteriaTest {
    
    /**
     * Test main
     * @param args ignored
     */
    public static void main(String[] args) {
        Instance[] instances =  {
            new Instance(new double[] {0, 0, 0, 1}, 1),
            new Instance(new double[] {1, 0, 0, 0}, 1),
            new Instance(new double[] {1, 0, 0, 0}, 1),
            new Instance(new double[] {1, 0, 0, 0}, 1),
            new Instance(new double[] {1, 0, 0, 1}, 0),
            new Instance(new double[] {1, 0, 0, 1}, 0),
            new Instance(new double[] {1, 0, 0, 1}, 0),
            new Instance(new double[] {1, 0, 0, 1}, 0)
        };
        DataSet set = new DataSet(instances);
        set.setDescription(new DataSetDescription(set));
        ChiSquarePruningCriteria cspc = new ChiSquarePruningCriteria(1);
        for (int i = 0; i < 4; i++) {
            DecisionTreeSplit split = 
                new StandardDecisionTreeSplit(i, 2);
            DecisionTreeSplitStatistics stats = 
                new DecisionTreeSplitStatistics(split, set);
            System.out.println("\nAttribute " + i);
            System.out.println("Should prune? " + cspc.shouldPrune(stats));
        }
    }
}