using System;
using System.Diagnostics;
using System.Globalization;
using HeuristicLab.Algorithms.Bandits.BanditPolicies;
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 sätze generieren vs. einfach zufällig alternativen wählen
// TODO: reward discounting (für veränderliche reward distributions über zeit). speziellen unit-test dafür erstellen
// TODO: constant optimization 


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

            RunDemo();
        }


        private static void RunDemo()
        {


            int maxIterations = 100000;
            int iterations = 0;

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

            //var problem = new SymbolicRegressionPoly10Problem();
            //var problem = new SantaFeAntProblem();              
            var problem = new RoyalPairProblem();
            //var problem = new EvenParityProblem();
            //var alg = new SequentialSearch(problem, 23, random, 0,
            // new HeuristicLab.Algorithms.Bandits.GrammarPolicies.GenericGrammarPolicy(problem, new UCB1TunedPolicy()));
            var alg = new MonteCarloTreeSearch(problem, 23, random, new UCB1Policy(), new RandomSimulation(problem, random, 23));


            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 % 10000 == 0) Console.Clear();
                    Console.SetCursorPosition(0, 0);
                    alg.PrintStats();
                }

                // uncomment this if you want to collect statistics of the generated sentences
                // if (iterations % 1000 == 0) {
                //   Console.WriteLine("{0}", globalStatistics);
                // }
            };

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

            Console.Clear();
            alg.PrintStats();
            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);
        }
    }
}