Package ch.idsia.tools

Examples of ch.idsia.tools.EvaluationOptions

31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
        // System.out.println(inputs.length+" "+inputs[0].length);
        // System.out.println(recurrent.length+" "+recurrent[0].length);
        // System.out.println(output.length+" "+output[0].length);
        SRN srn = new SRN (inputs, recurrent, output, recurrent.length, output[0].length);
        Agent agent = new LargeSRNAgent(srn);
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        final int startingSeed = 0;
        options.setLevelRandSeed(seed);
        options.setMaxAttempts(1);
        options.setVisualization(false);
        options.setMaxFPS(true);
        options.setLevelDifficulty(level);
        options.setPauseWorld(false);
        agent.reset();
        options.setAgent(agent);
        Evaluator evaluator = new Evaluator (options);
        EvaluationInfo result = evaluator.evaluate().get(0);
       // System.out.print(".");
        double score = result.computeDistancePassed();
         System.out.println("score: " +score);
View Full Code Here
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
    final static int generations = 100;
    final static int populationSize = 100;

    public static void main(String[] args) {
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        options.setMaxAttempts(1);
        options.setPauseWorld(false);
        Evolvable initial = new SmallMLPAgent();
        RegisterableAgent.registerAgent ((Agent) initial);
        options.setMaxFPS(true);
        options.setVisualization(false);
        MultiDifficultyProgressTask task = new MultiDifficultyProgressTask(options);

        ES es = new ES (task, initial, populationSize);
        System.out.println("Evolving " + initial + " with task " + task);
        final String fileName = "evolved" + (int) (Math.random () * Integer.MAX_VALUE) + ".xml";
View Full Code Here
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
    final static int repetitions = 10;

    public static void main(String[] args) {
        Agent controller = RegisterableAgent.load (args[0]);
        RegisterableAgent.registerAgent (controller);
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        options.setAgent(controller);
        options.setPauseWorld (false);
        Task task = new ProgressTask(options);
        options.setMaxFPS(true);
        options.setVisualization(false);     
        options.setMaxAttempts(1);
        options.setMatlabFileName("");
        task.setOptions(options);
        for (int i = 0; i < repetitions; i++) {
            System.out.println ("Score: " + task.evaluate (controller)[0]);
        }
    }
View Full Code Here
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
        // System.out.println(inputs.length+" "+inputs[0].length);
        // System.out.println(recurrent.length+" "+recurrent[0].length);
        // System.out.println(output.length+" "+output[0].length);
        SRN srn = new SRN (inputs, recurrent, output, recurrent.length, output[0].length);
        Agent agent = new LargeSRNAgent(srn);
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        options.setLevelRandSeed(seed);
        options.setMaxAttempts(1);
        options.setVisualization(true);
        options.setMaxFPS(false);
        options.setLevelDifficulty(level);
        options.setPauseWorld(false);
        agent.reset();
        options.setAgent(agent);
        Evaluator evaluator = new Evaluator (options);
        EvaluationInfo result = evaluator.evaluate().get(0);
       // System.out.print(".");
        double score = result.computeDistancePassed();
         System.out.println("score: " +score);
View Full Code Here
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
    final static int generations = 100;
    final static int populationSize = 100;
   

    public static void main(String[] args) {
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        options.setMaxAttempts(1);
        options.setPauseWorld(false);
        Evolvable initial = new SimpleMLPAgent();
        RegisterableAgent.registerAgent ((Agent) initial);
        options.setMaxFPS(true);
        options.setLevelDifficulty(0);
        options.setVisualization(false);
        ProgressTask task = new ProgressTask(options);
        options.setLevelRandSeed((int) (Math.random () * Integer.MAX_VALUE));
        ES es = new ES (task, initial, populationSize);
        System.out.println("Evolving " + initial + " with task " + task);
        final String fileName = "evolved" + (int) (Math.random () * Integer.MAX_VALUE) + ".xml";
        for (int gen = 0; gen < generations; gen++) {
            es.nextGeneration();
View Full Code Here
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
    final static int generations = 100;
    final static int populationSize = 100;


    public static void main(String[] args) {
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        options.setMaxAttempts(1);
        Evolvable initial = new LargeSRNAgent();
        if (args.length > 0) {
            initial = (Evolvable) RegisterableAgent.load (args[0]);
        }
        RegisterableAgent.registerAgent ((Agent) initial);
        options.setMaxFPS(true);
        options.setPauseWorld(false);
        options.setVisualization(false);
        ProgressTask task = new ProgressTask(options);
        int seed = (int) (Math.random () * Integer.MAX_VALUE);
        ES es = new ES (task, initial, populationSize);
        System.out.println("Evolving " + initial + " with task " + task);
        int difficulty = 0;
        final String fileName = "evolved" + (int) (Math.random () * Integer.MAX_VALUE) + ".xml";
        options.setLevelRandSeed(seed);
        for (int gen = 0; gen < generations; gen++) {
            es.nextGeneration();
            double bestResult = es.getBestFitnesses()[0];
            System.out.println("Generation " + gen + " diff " + difficulty + "  best " + bestResult);
            Easy.save (es.getBests()[0], fileName);
            if (bestResult > 4000) {
                difficulty++;
                options.setLevelDifficulty(difficulty);
                options.setLevelRandSeed(seed);
            }
        }
        Stats.main(new String[]{fileName, "0"});
        System.out.println("\n\n\n\n\n\n\n\n\n");
    }
View Full Code Here
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
    }

    public static void doStats (Agent agent, int startingSeed) {
        TimingAgent controller = new TimingAgent (agent);
        RegisterableAgent.registerAgent (controller);
        EvaluationOptions options = new CmdLineOptions(new String[0]);

        options.setMaxAttempts(1);
        options.setVisualization(false);
        options.setMaxFPS(true);
        options.setLevelLength(100000);
        System.out.println("Testing controller " + controller + " with starting seed " + startingSeed);

        double competitionScore = 0;

//        competitionScore += testConfig (controller, options, startingSeed, 0, true);
View Full Code Here
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
     
      GlobalOptions.currentController = controller.getName();
      GlobalOptions.writeFrames = false; //set to true to write frames to disk
      GlobalOptions.TimerOn = false;
      GlobalOptions.dontResetWindowPosition = true;
      EvaluationOptions options = new CmdLineOptions(new String[0]);
      options.setAgent(controller);
      Task task = new ProgressTask(options);
      options.setMaxFPS(false);
      options.setVisualization(true);
      options.setMaxAttempts(1);
      options.setMatlabFileName("");
      options.setLevelLength(length);
      options.setLevelRandSeed(seed);
      options.setLevelDifficulty(difficulty);
      options.setTimeLimit(0);
     
      task.setOptions(options);

      System.out.println("Score: " + ArrayUtils.toString(task.evaluate(controller)));
      System.out.println("Seed: " + options.getLevelRandSeed());
      System.out.println("Difficulty: " + options.getLevelDifficulty());
      seed++;
    }
  }
View Full Code Here
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
    final static int generations = 100;
    final static int populationSize = 100;


    public static void main(String[] args) {
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        options.setMaxAttempts(1);
        options.setPauseWorld(true);
        Evolvable initial = new SimpleMLPAgent();
        if (args.length > 0) {
            initial = (Evolvable) RegisterableAgent.load (args[0]);           
        }
        RegisterableAgent.registerAgent ((Agent) initial);
        for (int difficulty = 0; difficulty < 11; difficulty++)
        {
            System.out.println("New EvolveIncrementally phase with difficulty = " + difficulty + " started.");
            options.setLevelDifficulty(difficulty);
            options.setMaxFPS(true);
            options.setVisualization(false);
            //Task task = new ProgressTask(options);
            MultiSeedProgressTask task = new MultiSeedProgressTask(options);
            task.setNumberOfSeeds(3);
            task.setStartingSeed(0);
            ES es = new ES (task, initial, populationSize);
            System.out.println("Evolving " + initial + " with task " + task);
            for (int gen = 0; gen < generations; gen++) {
                es.nextGeneration();
                double bestResult = es.getBestFitnesses()[0];
                System.out.println("Generation " + gen + " best " + bestResult);
                options.setVisualization(gen % 5 == 0 || bestResult > 4000);
                options.setMaxFPS(true);
                Agent a = (Agent) es.getBests()[0];
                a.setName(((Agent)initial).getName() + gen);
                RegisterableAgent.registerAgent(a);
                double result = task.evaluate(a)[0];
                options.setVisualization(false);
                options.setMaxFPS(true);
                Easy.save (es.getBests()[0], "evolved.xml");
                if (result > 4000) {
                    initial = es.getBests()[0];
                    break; // Go to next difficulty.
                }
View Full Code Here
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
        GeneticAgent controller = new GeneticAgent(seed); // This line uses the agent you imported above.
        /*if (args.length > 0) {
            controller = RegisterableAgent.load (args[0]);
            RegisterableAgent.registerAgent (controller);
        }*/
        EvaluationOptions options = new CmdLineOptions(new String[0]);
        //options.setAgent(controller);
        Task task = new ProgressTask(options);
        options.setMaxFPS(false);
        options.setVisualization(false);
        options.setMaxAttempts(1);
        options.setMatlabFileName("");
        options.setLevelRandSeed(seed);
        options.setLevelDifficulty(10);
        GlobalOptions.FPS = GlobalOptions.InfiniteFPS;
        task.setOptions(options);
       
        GA vivarium = new GA(task, new GeneticAgent(), 60);
       
        for (int i = 0; i < 30; i++)
        {
          System.out.println("Best of Gen " + i + ": " + vivarium.getBestFitnesses()[0]);
          vivarium.nextGeneration();
          System.out.flush();
        }

        System.out.println("Score: " + ArrayUtils.toString(task.evaluate(controller)));
        System.out.println("Seed: " + options.getLevelRandSeed());
        System.out.println("Difficulty: " + options.getLevelDifficulty());
        System.out.println("Best Genome: " + vivarium.getBests()[0]);
    }
View Full Code Here
TOP

Related Classes of ch.idsia.tools.EvaluationOptions

  • ch.idsia.scenarios.CompetitionScore
  • ch.idsia.scenarios.Evolve
  • ch.idsia.scenarios.EvolveIncrementally
  • ch.idsia.scenarios.EvolveMultiSeed
  • ch.idsia.scenarios.GeneticPlay
  • ch.idsia.scenarios.MainRun
  • ch.idsia.scenarios.Play
  • ch.idsia.scenarios.PlayForever
  • ch.idsia.scenarios.Stats
  • ch.idsia.scenarios.test.EvaluateJLink
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.