source: branches/HeuristicLab.Problems.GrammaticalOptimization/Evaluation/MainWindow.xaml.cs @ 12741

using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Diagnostics;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Data;
using System.Windows.Documents;
using System.Windows.Input;
using System.Windows.Media;
using System.Windows.Media.Imaging;
using System.Windows.Navigation;
using System.Windows.Shapes;
using System.Windows.Threading;
using Evaluation.ViewModel;
using HeuristicLab.Algorithms.Bandits;
using HeuristicLab.Algorithms.Bandits.BanditPolicies;
using HeuristicLab.Algorithms.GeneticProgramming;
using HeuristicLab.Algorithms.GrammaticalOptimization;
using HeuristicLab.Algorithms.MonteCarloTreeSearch;
using HeuristicLab.Algorithms.MonteCarloTreeSearch.Simulation;
using HeuristicLab.Problems.GrammaticalOptimization;
using Microsoft.Research.DynamicDataDisplay;
using Microsoft.Research.DynamicDataDisplay.DataSources;

namespace Evaluation
{
    /// <summary>
    /// Interaction logic for MainWindow.xaml
    /// </summary>
    public partial class MainWindow : Window
    {
        private BackgroundWorker worker = new BackgroundWorker();
        private EvaluationViewModel vm;
        private List<EvaluationStat> stats = new List<EvaluationStat>();

        private Stack<EvaluationStat> newStats = new Stack<EvaluationStat>();
        private DispatcherTimer updateCollectionTimer;

        public MainWindow()
        {
            InitializeComponent();
            this.DataContext = vm = new EvaluationViewModel();
            this.worker.WorkerSupportsCancellation = true;
            this.worker.DoWork += worker_DoWork;
            this.worker.ProgressChanged += worker_ProgressChanged;
            this.worker.RunWorkerCompleted += worker_RunWorkerCompleted;

            ////updateCollectionTimer = new DispatcherTimer();
            ////updateCollectionTimer.Interval = TimeSpan.FromMilliseconds(100);
            ////updateCollectionTimer.Tick += updateCollectionTimer_Tick;
            ////updateCollectionTimer.Start();

            vm.HorizontalAxisString = "Evaluations";
            vm.VerticalAxisString = "BestKnownQuality";

        }

        ////void updateCollectionTimer_Tick(object sender, EventArgs e)
        ////{
        ////    while (newStats.Count > 0)
        ////    {
        ////        EvaluationStat stat = newStats.Pop();
        ////        stats.Add(stat);
        ////    }
        ////}

        void worker_RunWorkerCompleted(object sender, RunWorkerCompletedEventArgs e)
        {
            if (stats.Count > 0)
            {
                TimeSpan timeNeeded = stats[stats.Count - 1].Time - stats[0].Time;
                vm.EvaluationsPerSec = Math.Round(vm.Evaluations / timeNeeded.TotalSeconds, 2);
            }
            var ds = new EnumerableDataSource<EvaluationStat>(stats);
            ds.SetXMapping(x => x.Iteration);
            ds.SetYMapping(y => y.CurrentBestQuality);

            LineGraph graph = new LineGraph(ds);

            graph.StrokeThickness = 2;
            graph.AddToPlotter(ChartPlotter);

            Debug.WriteLine("DONE");
            ButtonRun.IsEnabled = true;
        }
        void worker_ProgressChanged(object sender, ProgressChangedEventArgs e)
        {
            Debug.WriteLine(e.UserState);
        }

        void worker_DoWork(object sender, DoWorkEventArgs e)
        {
            Type algorithmType = vm.SelectedAlgorithm;

            ISymbolicExpressionTreeProblem problem = vm.SelectedProblem;

            Random random = new Random(DateTime.Now.Millisecond);

            Type policy = vm.SelectedPolicy;
            IBanditPolicy policyInstance = null;

            if (policy == typeof(UCTPolicy))
            {
                policyInstance = new UCTPolicy();
            }
            else if (policy == typeof (ThresholdAscentPolicy))
            {
                policyInstance = new ThresholdAscentPolicy();
            }
            else
            {
                policyInstance = (IBanditPolicy)Activator.CreateInstance(policy);
            }

            vm.MaxLen = 1000;
            vm.MaxEvaluations = 250000;

            vm.BestKnownQuality = problem.BestKnownQuality(vm.MaxLen);

            vm.Evaluations = 0;
            vm.CurrentBestQuality = 0;

            stats.Clear();

            SolverBase solver = null;

            if (algorithmType == typeof(MonteCarloTreeSearch))
            {
                solver = new MonteCarloTreeSearch(problem, vm.MaxLen, random, policyInstance, new RandomSimulation(problem, random, vm.MaxLen));
            }
            else if (algorithmType == typeof(SequentialSearch))
            {
                solver = new SequentialSearch(problem, vm.MaxLen, random, 0,
                    new HeuristicLab.Algorithms.Bandits.GrammarPolicies.GenericGrammarPolicy(problem, policyInstance));
            }
            else if (algorithmType == typeof(RandomSearch))
            {
                solver = new RandomSearch(problem, random, vm.MaxLen);
            }
            else if (algorithmType == typeof(StandardGP))
            {
                solver = new StandardGP(problem, random);
            }
            else if (algorithmType == typeof(OffspringSelectionGP))
            {
                solver = new OffspringSelectionGP(problem, random);
            }

            solver.FoundNewBestSolution += (sentence, quality) => vm.BestSolutionFoundAt = vm.Evaluations;
            solver.SolutionEvaluated += (sentence, quality) =>
            {
                vm.Evaluations++;
                if (vm.CurrentBestQuality < quality)
                {
                    vm.CurrentBestQuality = quality;
                }
                stats.Add(new EvaluationStat(DateTime.Now, vm.Evaluations, quality, vm.CurrentBestQuality));
            };

            solver.Run(vm.MaxEvaluations);
        }

        private void ButtonRun_OnClick(object sender, RoutedEventArgs e)
        {
            ChartPlotter.Children.RemoveAll<LineGraph>();
            ButtonRun.IsEnabled = false;
            worker.RunWorkerAsync();
        }

        private void ButtonPause_OnClick(object sender, RoutedEventArgs e)
        {
        }

        private void ButtonStop_OnClick(object sender, RoutedEventArgs e)
        {
            worker.CancelAsync();
        }

        private void ComboBoxAlgorithms_OnSelectionChanged(object sender, SelectionChangedEventArgs e)
        {
            if (vm.SelectedAlgorithm == typeof(MonteCarloTreeSearch)
                || vm.SelectedAlgorithm == typeof(SequentialSearch))
            {
                ComboBoxPolicies.IsEnabled = true;
            }
            else
            {
                ComboBoxPolicies.IsEnabled = false;
            }
        }
    }
}