protected override void Backpropagate(MCTSNode node, Reward reward)
{
/*
* while v is not null do
* N(v) <- N(v) + 1
* Q(v) <- Q(v) + r(v, Player(Parent(v)))
* v <- parent of v
*/
int player = 0;
MCTSNode currentNode = node;
Pair <int, Action> par = new Pair <int, Action>(0, new Action("asdasdas"));
while (currentNode != null)
{
currentNode.N = currentNode.N + 1;
currentNode.Q = currentNode.Q + reward.GetRewardForNode(node);
if (currentNode.Parent != null)
{
par.Left = currentNode.Parent.PlayerID;
par.Right = currentNode.Action;
ActionHistory.Add(par);
}
currentNode = currentNode.Parent;
if (currentNode != null)
{
player = currentNode.PlayerID;
/* protected List<Pair<int, Action>> ActionHistory { get; set; }*/
foreach (MCTSNode child in currentNode.ChildNodes)
{
par.Left = player;
par.Right = child.Action;
if (ActionHistory.Contains(par))//Pair<player,child.Action>)
{
child.NRAVE = child.NRAVE + 1;
child.QRAVE = child.QRAVE + reward.GetRewardForNode(child);
}
}
}
/*if v is not null do
* p < -Player(v)
* foreach (c in Children(v))
* if ((p, A(c)) in actionHistory) do
* Nrave(c) < -Nrave(c) + 1
* Qrave(c) < -Qrave(c) + r(c, p)*/
}
}
protected virtual void Backpropagate(MCTSNode node, Reward reward)
{
MCTSNode currentNode = node;
while (currentNode != null)
{
currentNode.N = currentNode.N + 1;
currentNode.Q = currentNode.Q + reward.GetRewardForNode(node);
currentNode = currentNode.Parent;
}
}
protected override Reward Playout(WorldModel initialPlayoutState)
{
WorldModel state = initialPlayoutState.GenerateChildWorldModel();
while (!state.IsTerminal())
{
getRandomAction(state).ApplyActionEffects(state);
}
Reward r = new Reward();
r.Value = r.GetRewardForNode(new MCTSNode(state));
return(r);
}