public bool[] RunFrame()
{
var action1 = agent1.Act(gs, 1);
var action2 = agent2.Act(gs, 2);
bool[] frameResult = PacManGameState.Step(gs, action1, action2, speed);
agent1.Obs((frameResult[0] ? 1F : 0F), frameResult[2]);
agent1.Obs((frameResult[1] ? 1F : 0F), frameResult[2]);
return(frameResult);
}
public MovementIntent Act(PacManGameState gs, int playerNumber)
{
int bestActionScore = 0;
int j;
MovementIntent bestAction = 0;
foreach (var action in movementIntentValues)
{
int actionScore = 0;
for (var i = 0; i < RolloutIterations; i++)
{
gsCopy.CopyGS(gs);
var randActionIndex = UnityEngine.Random.Range(0, movementIntentValues.Length);
MovementIntent randAction = (MovementIntent)movementIntentValues.GetValue(randActionIndex);
var result = PacManGameState.Step(gsCopy, action, randAction, Speed);
for (j = 0; !result[2] && j < ExplorationFrames; j++) // While not terminal state
{
randActionIndex = UnityEngine.Random.Range(0, movementIntentValues.Length);
MovementIntent randAction1 = (MovementIntent)movementIntentValues.GetValue(randActionIndex);
randActionIndex = UnityEngine.Random.Range(0, movementIntentValues.Length);
MovementIntent randAction2 = (MovementIntent)movementIntentValues.GetValue(randActionIndex);
result = PacManGameState.Step(gsCopy, randAction1, randAction2, Speed);
}
actionScore += result[playerNumber] ? 1 : 0;
}
if (actionScore > bestActionScore)
{
bestActionScore = actionScore;
bestAction = action;
}
}
return(bestAction);
}