public PlayerTask Simulate(UCTNode currentNode, long accumulativeMillisecond)
{
if (root.IsCurrentPlayerTurnEnd())
{
return(root.currentPOGame.CurrentPlayer.Options()[0]);
}
if (root.IsGameOver())
{
return(root.currentPOGame.CurrentPlayer.Options()[0]);
}
double reward = 0;
int onePlayoutTimes = 0;
bool isStop = false;
long startMilliseconds = DateTimeOffset.Now.ToUnixTimeMilliseconds();
if (_debug_)
{
Log.Instance().Append(currentNode.FullPrint());
}
while (!isStop &&
onePlayoutTimes < maxOnePlayOutTimes &&
DateTimeOffset.Now.ToUnixTimeMilliseconds() - startMilliseconds < this.maxOptionCalculateTime &&
DateTimeOffset.Now.ToUnixTimeMilliseconds() - accumulativeMillisecond < this.maxTurnCalculateTime)
{
UCTNode currentPlayerLeafNode = OnePlayOut(currentNode, availableOptionTypes);
if (_debug_)
{
Log.Instance().Append(currentPlayerLeafNode.FullPathPrint());
}
if (_debug_)
{
Log.Instance().Append(currentPlayerLeafNode.FullPrint());
}
reward = EvaluateUCTNode(currentPlayerLeafNode, maxTurnCalculateTime);
if (_debug_)
{
Log.Instance().Append("reward : " + reward.ToString());
}
UpdateReward(currentPlayerLeafNode, reward);
onePlayoutTimes++;
}
reward = EvaluateUCTNode(root, maxTurnCalculateTime);
PlayerTask bestOption = GetBestOption(root, reward);
if (onePlayoutTimes > maxRealOnePlayOutTimes)
{
maxRealOnePlayOutTimes = onePlayoutTimes;
}
Debug.Assert(bestOption != null);
return(bestOption);
}
