source: branches/HeuristicLab.Problems.GrammaticalOptimization-gkr/Main/Program.cs @ 12290

using System;
using System.Diagnostics;
using System.Globalization;
using HeuristicLab.Algorithms.Bandits.BanditPolicies;
using HeuristicLab.Algorithms.Bandits.GrammarPolicies;
using HeuristicLab.Algorithms.GrammaticalOptimization;
using HeuristicLab.Algorithms.MonteCarloTreeSearch;
using HeuristicLab.Algorithms.MonteCarloTreeSearch.Simulation;
using HeuristicLab.Problems.GrammaticalOptimization;

// NOTES: gkronber
// TODO: feature extraction for full symbolic expressions and experiment for all benchmark problems
// TODO: why does GaussianThompsonSampling work so well with MCTS for the artificial ant problem?
// TODO: research thompson sampling for max bandit?
// TODO: verify TA implementation using example from the original paper      
// TODO: implement thompson sampling for gaussian mixture models
// TODO: gleichzeitige modellierung von transformierter zielvariable (y, 1/y, log(y), exp(y), sqrt(y), ...)
// TODO: vergleich bei complete-randomly möglichst kurze sÀtze generieren vs. einfach zufÀllig alternativen wÀhlen
// TODO: reward discounting (fÌr verÀnderliche reward distributions Ìber zeit). speziellen unit-test dafÌr erstellen
// TODO: constant optimization 
using HeuristicLab.Problems.GrammaticalOptimization.SymbReg;


namespace Main {
  class Program {
    static void Main(string[] args) {
      CultureInfo.DefaultThreadCurrentCulture = CultureInfo.InvariantCulture;

      RunDemo();
    }


    private static void RunDemo() {

      for (int i = 0; i < 100; i++) {
        int maxIterations = 2000000;
        int iterations = 0;

        var globalStatistics = new SentenceSetStatistics();
        var random = new Random();

        var problem = new SymbolicRegressionPoly10Problem();
        //var problem = new SantaFeAntProblem();              
        //var problem = new RoyalPairProblem(25);
        //var problem = new FindPhrasesProblem(random, 10, 5, 3, 5, 5, 1.0, 0.9, true);
        //var problem = new PrimePolynomialProblem();
        //var problem = new SymbolicRegressionProblem(random,
        //  //@"C:\reps\HeuristicLab\branches\HeuristicLab.Problems.GrammaticalOptimization\HeuristicLab.Problems.GrammaticalOptimization.SymbReg\nht-train.csv",
        //  @"C:\reps\fhooe-new\research\Datasets\Benchmark\kommenda-1.csv",
        //  1.0,
        //  true);
        // //var problem = new PrimePolynomialProblem();
        // var alg = new SequentialSearch(problem, 25, random, 0,
        //   new HeuristicLab.Algorithms.Bandits.GrammarPolicies.GenericGrammarPolicy(problem, new UCB1TunedPolicy()));
        var alg = new SequentialSearch(problem, 25, random, 0,
          new GenericPolicy(problem, new HeuristicLab.Algorithms.Bandits.BanditPolicies.EpsGreedyPolicy(0.1)));
        //var alg = new MonteCarloTreeSearch(problem, 23, random, new UCB1Policy(), new RandomSimulation(problem, random, 30));


        alg.FoundNewBestSolution += (sentence, quality) => {
          //Console.WriteLine("{0}", globalStatistics);
        };

        alg.SolutionEvaluated += (sentence, quality) => {
          iterations++;
          globalStatistics.AddSentence(sentence, quality);

          // comment this if you don't want to see solver statistics
          if (iterations % 100 == 0) {
            if (iterations % 1000 == 0) Console.Clear();
            Console.SetCursorPosition(0, 0);
            alg.PrintStats();
          }

          // uncomment this if you want to collect statistics of the generated sentences
          //if (iterations % 100 == 0) {
          //  Console.WriteLine("{0}", globalStatistics);
          //}
        };

        var sw = new Stopwatch();
        sw.Start();
        alg.Run(maxIterations);
        sw.Stop();

        Console.WriteLine(globalStatistics);

        Console.WriteLine("{0:F2} sec {1,10:F1} sols/sec {2,10:F1} ns/sol",
          sw.Elapsed.TotalSeconds,
          maxIterations / (double)sw.Elapsed.TotalSeconds,
          (double)sw.ElapsedMilliseconds * 1000 / maxIterations);
      }
    }
  }
}