source: branches/HeuristicLab.Problems.GrammaticalOptimization/HeuristicLab.Algorithms.Bandits/Policies/GenericThompsonSamplingPolicy.cs @ 13042

using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using HeuristicLab.Common;

namespace HeuristicLab.Algorithms.Bandits.BanditPolicies {
  public class GenericThompsonSamplingPolicy : IBanditPolicy {
    private readonly IModel model;

    public GenericThompsonSamplingPolicy(IModel model) {
      this.model = model;
    }

    public int SelectAction(Random random, IEnumerable<IBanditPolicyActionInfo> actionInfos) {
      var myActionInfos = actionInfos.OfType<ModelPolicyActionInfo>();
      int bestAction = -1;
      double bestQ = double.NegativeInfinity;
      var aIdx = -1;
      foreach (var aInfo in myActionInfos) {
        aIdx++;
        var q = aInfo.SampleExpectedReward(random);
        // very unlikely to be equal and we don't care
        if (q > bestQ) {
          bestQ = q;
          bestAction = aIdx;
        }
      }
      Debug.Assert(bestAction > -1);
      return bestAction;
    }

    public IBanditPolicyActionInfo CreateActionInfo() {
      return new ModelPolicyActionInfo((IModel)model.Clone());
    }

    public override string ToString() {
      return string.Format("GenericThompsonSamplingPolicy({0})", model);
    }
  }
}