/*
 * Main.java
 *
 * Copyright (C) August Mayer, 2001-2004. All rights reserved.
 * Please consult the Boone LICENSE file for additional rights granted to you.
 *
 * Created on 25. November 2002, 14:47
 */

package samples.programs;

import boone.NetFactory;
import boone.NeuralNet;
import boone.PatternSet;
import boone.Trainer;
import boone.training.RpropTrainer;
import boone.util.Conversion;

import java.io.File;
import java.io.IOException;

/**
 * Test program: the XOR-Problem
 *
 * @author August Mayer
 * @version $Id: XORTest.java 2292 2016-01-15 16:52:18Z helmut $
 */
public class XORTest {

	/**
	 * XOR programs network, using back-propagation
	 *
	 * @param args the command line arguments
	 */
	public static void main(String[] args) {

		System.out.println("*** Creating feed forward network...");

		NeuralNet net = NetFactory.createFeedForward(new int[]{2, 2, 1}, false, new boone.map.Function.Sigmoid(), new RpropTrainer(), null, null);

		double[][] inPatterns = new double[][]{{0, 0}, {0, 1}, {1, 0}, {1, 1}};
		double[][] outPatterns = new double[][]{{0}, {1}, {1}, {0}};
		PatternSet patterns = new PatternSet();
		for (int i = 0; i < inPatterns.length; i++) {
			patterns.getInputs().add(Conversion.asList(inPatterns[i]));
			patterns.getTargets().add(Conversion.asList(outPatterns[i]));
		}
		int steps = 100;
		int epochs = 10;
		Trainer trainer = net.getTrainer();
		trainer.setTrainingData(patterns);
		trainer.setTestData(patterns);
		trainer.setEpochs(epochs);
		trainer.setStepMode(true);											// training in steps
		System.out.println("*** Training " + (steps * epochs) + " epochs...");
//		trainer.setShuffle(true);
//		((BackpropTrainer)trainer).setMomentum(0.8);
		System.out.println("Error: ");
		for (int i = 0; i < steps; i++) {
			trainer.train();
			System.out.println((i * epochs) + ". - " + net.getTrainer().test());
		}

		System.out.println("\n*** Testing the network...");

		System.out.println();
		for (int i = 0; i < patterns.size(); i++)
			System.out.println("Error " + i + " = " + net.getTrainer().test(patterns.getInputs().get(i), patterns.getTargets().get(i)));

		System.out.println("Saving network");

		try {
			net.save(new File("test/samples/data/XORTest.xnet"));
		} catch (IOException e) {
			e.printStackTrace();
		}
		System.out.println("Printing the net:\n" + net);
		System.out.println("Done.");
	}

}