private static void UseDpToFindOptimalPolicy()
{
const double probabilityOfHeads = 0.4;
const int dollarsToWin = 100;
const int evaluationSweepsPerPolicyUpdate = 1;
var world = new GamblersWorld(probabilityOfHeads, dollarsToWin);
var rewarder = new GamblersWorldRewarder(world);
Console.WriteLine("Policy optimiser:");
var(policy, values) = DpPolicyOptimiser
.FindOptimalPolicy(world, rewarder, evaluationSweepsPerPolicyUpdate);
Console.WriteLine("Optimal policy values:");
PrintAllValues(world, values);
Console.WriteLine("Optimal policy stakes:");
PrintPolicyActions(world, policy);
}
private static void EvaluatePolicyValues()
{
const double probabilityOfHeads = 0.4;
const int dollarsToWin = 100;
var world = new GamblersWorld(probabilityOfHeads, dollarsToWin);
var rewarder = new GamblersWorldRewarder(world);
Console.WriteLine("Optimal policy:");
var(policy, values) = DpPolicyOptimiser.FindOptimalPolicy(world, rewarder);
var evalValues = new ValueTable <GamblersWorldState, GamblersWorldAction>(world);
evalValues.Evaluate(policy, rewarder);
PrintPolicyActions(world, policy);
PrintAllValues(world, values);
// Console.WriteLine("Always stake max policy:");
// var policy = new AlwaysStakeMaxPolicy(world);
// PrintPolicyActionsAndValues(world, policy, rewarder);
}
