public void TestAgainstPolicy1()
{
Rollout[] results = RolloutPerformer.Winner(PolicyProvider.Policy, SubmittedPolicyProvider.Policy1);
double winRate = results.Average(x => x.WinRate);
Assert.IsTrue(winRate > 0.5);
}
public void TestEqualMatch()
{
Rollout[] results =
RolloutPerformer.Winner(SubmittedPolicyProvider.Policy1, SubmittedPolicyProvider.Policy1)
.OrderBy(x => x.Matchup.Team0[0])
.ThenBy(x => x.Matchup.Team0[1])
.ThenBy(x => x.Matchup.Team1[0])
.ThenBy(x => x.Matchup.Team1[1])
.ToArray();
/*
* foreach (MatchupResult result in results) {
* Debug.WriteLine(
* "{0}\t{1}\t{2}\t{3}\t{4}",
* result.Matchup.Team0[0],
* result.Matchup.Team0[1],
* result.Matchup.Team1[0],
* result.Matchup.Team1[1],
* result.WinRate);
* }
*/
double winRate = results.Average(x => x.WinRate);
Assert.AreEqual(0.5, winRate, 0.05);
}
private PolicyCandidate EvaluatePolicy(Policy policy, Policy incumbent)
{
Rollout[] rollouts = RolloutPerformer.Rollout(policy, incumbent);
return(new PolicyCandidate {
Policy = policy,
Rollouts = rollouts,
WinRate = rollouts.Average(r => IntermediateWinner(r)),
});
}
public static void Main(string[] args)
{
#if DEBUGCUDA
Telogis.RouteCloud.GPUManagement.KernelManager.SynchroniseAfterEveryKernel = true;
#endif
string outputPath = args[0];
Console.WriteLine("Environment.ProcessorCount: " + Environment.ProcessorCount);
Console.WriteLine("64-bit process? " + Environment.Is64BitProcess);
Console.WriteLine("World.NumHeroesPerTeam = " + World.NumHeroesPerTeam);
Console.WriteLine("World.SpawnMinions = " + World.SpawnMinions);
Console.WriteLine("World.EnableBuySell = " + World.EnableBuySell);
Console.WriteLine("World.EnableSpells = " + World.EnableSpells);
using (Trainer trainer = new Trainer()) {
Policy policy = PolicyProvider.Policy;
Policy[] incumbents = SubmittedPolicyProvider.Submissions.ToArray();
double[] incumbentWinRates =
incumbents
.Select(oldPolicy => RolloutPerformer.Winner(policy, oldPolicy).Average(x => x.WinRate)).ToArray();
Console.WriteLine("Initial win rates: " + string.Join(" ", incumbentWinRates));
int improvements = 0;
for (int iteration = 0; ; ++iteration)
{
Policy incumbent = policy;
Console.WriteLine(string.Format("Iteration #{0}", iteration));
TrainingResult trainingResult = trainer.StrengthenPolicy(policy, incumbent);
policy = trainingResult.Policy;
List <Rollout[]> incumbentResults = incumbents.Select(oldPolicy => RolloutPerformer.Winner(policy, oldPolicy)).ToList();
incumbentWinRates = incumbentResults.Select(submissionResult => submissionResult.Average(x => x.WinRate)).ToArray();
Console.WriteLine("Against incumbents: " + string.Join(" ", incumbentWinRates));
policy.HeroMatchups = MatchupOptimizer.Optimize(incumbentResults.SelectMany(x => x));
if (trainingResult.IsImprovement)
{
++improvements;
Console.WriteLine(string.Format("Improvement #{0}", improvements));
}
if (outputPath != null)
{
File.WriteAllText(outputPath, PolicySerializer.Serialize(policy));
}
Console.WriteLine();
}
}
}
