public Dictionary <Action, double> GetValue(State s, FriendQTable q)
{
var context = SolverContext.GetContext();
var model = context.CreateModel();
var actionDecisions = new List <Decision>();
foreach (var action in Enum.GetNames(typeof(Action)))
{
var decision = new Decision(Domain.RealNonnegative, action);
model.AddDecisions(decision);
actionDecisions.Add(decision);
}
var valueDecision = new Decision(Domain.RealNonnegative, "value");
model.AddDecisions(valueDecision);
model.AddConstraint("probSumConst", actionDecisions[0] + actionDecisions[1] + actionDecisions[2] + actionDecisions[3] + actionDecisions[4] == 1.0);
int constCount = 0;
foreach (Action playerOneAction in Enum.GetValues(typeof(Action)))
{
var qConstraintValues = new List <double>();
foreach (Action playerTwoAction in Enum.GetValues(typeof(Action)))
{
qConstraintValues.Add(q.GetQValue(s, playerOneAction, playerTwoAction));
}
model.AddConstraint("Const" + constCount, qConstraintValues[0] * actionDecisions[0] + qConstraintValues[1] * actionDecisions[1] + qConstraintValues[2] * actionDecisions[2] + qConstraintValues[3] * actionDecisions[3] + qConstraintValues[4] * actionDecisions[4] <= valueDecision);
++constCount;
}
model.AddGoal("MinimizeV", GoalKind.Minimize, valueDecision);
context.Solve(new SimplexDirective());
var pi_s = new Dictionary <Action, double>();
foreach (var actionDec in actionDecisions)
{
pi_s[(Action)Enum.Parse(typeof(Action), actionDec.Name)] = actionDec.GetDouble();
}
context.ClearModel();
return(pi_s);
}
private static void RunFriendQ()
{
var S = new StateSet();
var A = new JointActionSet();
var P = new Transition(S, A);
var Q_A = new FriendQTable();
var R = new Rewards();
var ERR = new Dictionary <int, double>();
var j = 0;
while (j <= 1000000)
{
//Initialize state according to Figure 4
var initialState = new State(2, 1, BallPossessor.B);
var done = false;
var currState = new State(3, 1, BallPossessor.B);
var actions = A.GetNextJointAction();
var playerAAction = actions.currplayer1Act;
var playerBAction = actions.player2act;
while (!done)
{
if (j % 20000 == 0)
{
Console.WriteLine(j);
}
var q_fig_4_initial = Q_A.getQval(initialState, Action.S, Action.X);
var nextState = P.GetNextState(currState, new JointAction(playerAAction, playerBAction));
var playerAReward = R.GetPlayerAReward(nextState);
var playerBReward = R.GetPlayerBReward(nextState);
Q_A.UpdateQval(currState, nextState, playerAAction, playerBAction, playerAReward);
if (playerAReward != 0.0 || playerBReward != 0.0)
{
done = true;
}
if (currState.Equals(initialState) && playerAAction == Action.S && playerBAction == Action.X)
{
var q_fig_4 = Q_A.getQval(initialState, Action.S, Action.X);
var diff = Math.Abs(q_fig_4 - q_fig_4_initial);
ERR.Add(j, diff);
}
currState = nextState;
actions = A.GetNextJointAction();
playerAAction = actions.currplayer1Act;
playerBAction = actions.player2act;
++j;
}
}
using (StreamWriter sw = File.CreateText("output.csv"))
{
foreach (var kvp in ERR)
{
sw.WriteLine(kvp.Key + "," + kvp.Value);
}
}
}