private double GetMaxQ(VirusBoard state)
{
double max = -10;
Move[] actions = state.GetPossibleMoves(playerNumber);
foreach (Move a in actions)
{
double value = 0;
if (!Q[state.CustomHash()].ContainsKey(a.CustomHash()))
{
value = 0;
}
else
{
value = Q[state.CustomHash()][a.CustomHash()];
}
if (value > max)
{
max = value;
}
}
if (Q[state.CustomHash()].ContainsKey(0))
{
if (Q[state.CustomHash()][0] > max)
{
max = Q[state.CustomHash()][0];
}
}
return(max);
}
public override Move Move(Virus percept)
{
VirusBoard currentState = percept.GetBoardCopy();
Move[] actions = currentState.GetPossibleMoves(playerNumber);
Move action = actions[0];
double max = double.NegativeInfinity;
foreach (Move a in actions)
{
VirusBoard newState = currentState.GetUpdated(a);
double q = 0;
if (Q.ContainsKey(currentState.CustomHash()))
{
if (Q[currentState.CustomHash()].ContainsKey(a.CustomHash()))
{
q = Q[currentState.CustomHash()][a.CustomHash()];
}
}
q += MinValue(newState, 0);
if (q > max)
{
max = q;
action = a;
}
if (max == 1)
{
break;
}
}
return(action);
}
private Move GetMaxExplorationFunctionA(VirusBoard state)
{
double max = double.NegativeInfinity;
Move action = default(Move);
Move[] actions = state.GetPossibleMoves(playerNumber);
bool berandom = random.NextDouble() < RandomRate;
foreach (Move a in actions)
{
double value = 0;
if (Q[state.CustomHash()].ContainsKey(a.CustomHash()))
{
if (Q[state.CustomHash()][a.CustomHash()] >= 1)
{
value = 1;
max = value;
action = a;
break;
}
else if (Q[state.CustomHash()][a.CustomHash()] <= -1)
{
value = -1;
}
else
{
if (berandom)
{
value = random.NextDouble();
}
else
{
value = Q[state.CustomHash()][a.CustomHash()] + ((explore && !(RandomRate > 0)) ? ExplorationRate(N[state.CustomHash()][a.CustomHash()]) : 0);
}
}
}
else
{
value = 1;
}
if (value > max)
{
max = value;
action = a;
}
}
return(action);
}
//Calc maxValue
private double MaxValue(VirusBoard state, int iteration)
{
iteration++;
if (state.winner == playerNumber)
{
return(double.PositiveInfinity);
}
if (state.winner != playerNumber && state.winner != 0)
{
return(double.NegativeInfinity);
}
if (iteration < searchLength)
{
Move[] actions = state.GetPossibleMoves(playerNumber);
double max = double.NegativeInfinity;
foreach (Move a in actions)
{
VirusBoard newState = state.GetUpdated(a);
double q = Utility(state, newState);
if (Q.ContainsKey(state.CustomHash()))
{
if (Q[state.CustomHash()].ContainsKey(a.CustomHash()))
{
q = Q[state.CustomHash()][a.CustomHash()];
}
}
q += MinValue(newState, iteration);
if (q > max)
{
max = q;
}
if (max == double.PositiveInfinity)
{
return(max);
}
}
return(max);
}
else
{
return(0);
}
}
public override Move Move(Virus percept)
{
//Checking if we're at an terminal state
byte winner = percept.Winner;
VirusBoard newState = percept.GetBoardCopy();
if (!Q.ContainsKey(newState.CustomHash()))
{
Q.Add(newState.CustomHash(), new Dictionary <UInt32, double>());
}
if (learn && !prevState.Equals(default(VirusBoard)))
{
if (!N.ContainsKey(prevState.CustomHash()))
{
N.Add(prevState.CustomHash(), new Dictionary <UInt32, int>());
}
if (!N[prevState.CustomHash()].ContainsKey(prevAction.CustomHash()))
{
N[prevState.CustomHash()].Add(prevAction.CustomHash(), 0);
}
if (!Q.ContainsKey(prevState.CustomHash()))
{
Q.Add(prevState.CustomHash(), new Dictionary <UInt32, double>());
}
if (!Q[prevState.CustomHash()].ContainsKey(prevAction.CustomHash()))
{
Q[prevState.CustomHash()].Add(prevAction.CustomHash(), initvalue);
}
if (winner == playerNumber)
{
if (!Q[newState.CustomHash()].ContainsKey(0))
{
Q[newState.CustomHash()].Add(0, 1);
}
}
else if (winner != playerNumber && winner != 0)
{
if (!Q[newState.CustomHash()].ContainsKey(0))
{
Q[newState.CustomHash()].Add(0, -1);
}
}
N[prevState.CustomHash()][prevAction.CustomHash()]++;
Q[prevState.CustomHash()][prevAction.CustomHash()] =
Q[prevState.CustomHash()][prevAction.CustomHash()]
+ LearningRate(N[prevState.CustomHash()][prevAction.CustomHash()])
* (prevReward + discount * GetMaxQ(newState) - Q[prevState.CustomHash()][prevAction.CustomHash()]);
}
prevState = newState;
prevAction = GetMaxExplorationFunctionA(newState);
prevReward = 0;
return(prevAction);
}
public override Move Move(Virus percept)
{
VirusBoard newState = percept.GetBoardCopy();
if (!Q.ContainsKey(newState.CustomHash()))
{
Q.Add(newState.CustomHash(), new Dictionary <UInt32, double>());
}
if (learn && !prevState.Equals(default(VirusBoard)))
{
if (!N.ContainsKey(prevState.CustomHash()))
{
N.Add(prevState.CustomHash(), new Dictionary <UInt32, int>());
}
if (!N[prevState.CustomHash()].ContainsKey(prevAction.CustomHash()))
{
N[prevState.CustomHash()].Add(prevAction.CustomHash(), 0);
}
N[prevState.CustomHash()][prevAction.CustomHash()]++;
Q[prevState.CustomHash()][prevAction.CustomHash()] =
Q[prevState.CustomHash()][prevAction.CustomHash()]
+ LearningRate(N[prevState.CustomHash()][prevAction.CustomHash()])
* (prevReward + discount * GetMaxQ(newState) - Q[prevState.CustomHash()][prevAction.CustomHash()]);
}
prevState = newState;
prevAction = GetMaxExplorationFunctionA(newState);
prevReward = 0;
if (learn && !Q[prevState.CustomHash()].ContainsKey(prevAction.CustomHash()))
{
Q[prevState.CustomHash()].Add(prevAction.CustomHash(), initvalue);
}
return(prevAction);
}
public override void EndGame(Virus percept)
{
if (learn)
{
double reward = 0;
byte winner = percept.Winner;
if (winner == playerNumber)
{
reward = 1;
}
else if (winner != playerNumber && winner != 0)
{
reward = -1;
}
else
{
reward = 0;
}
if (!N.ContainsKey(prevState.CustomHash()))
{
N.Add(prevState.CustomHash(), new Dictionary <UInt32, int>());
}
if (!N[prevState.CustomHash()].ContainsKey(prevAction.CustomHash()))
{
N[prevState.CustomHash()].Add(prevAction.CustomHash(), 0);
}
N[prevState.CustomHash()][prevAction.CustomHash()]++;
Q[prevState.CustomHash()][prevAction.CustomHash()] =
Q[prevState.CustomHash()][prevAction.CustomHash()]
+ LearningRate(N[prevState.CustomHash()][prevAction.CustomHash()])
* (reward - Q[prevState.CustomHash()][prevAction.CustomHash()]);
}
prevState = default(VirusBoard);
prevAction = default(Move);
prevReward = 0;
}