static void Main(string[] args)
{
var env = new CliffWalkingEnvironment();
var sarsaAgent = new SarsaCliffWalker();
var values = sarsaAgent.ImproveEstimates(env, out var tdDiags, 10000);
System.Console.WriteLine("td 0 avg Values:");
StateActionValuesConsoleRenderer.RenderAverageValues(values);
System.Console.WriteLine("td 0 highest Values:");
StateActionValuesConsoleRenderer.RenderHighestValues(values);
System.Console.WriteLine("");
System.Console.WriteLine("td 0 Greedy path:");
ConsolePathRenderer.RenderPath(GreedyPath(env, values));
var qAgent = new QLearningCliffWalker();
var qValues = qAgent.ImproveEstimates(env, out var qDiags, 10000);
System.Console.WriteLine("");
System.Console.WriteLine("q learning avg Values:");
StateActionValuesConsoleRenderer.RenderAverageValues(qValues);
System.Console.WriteLine("q learning highest Values:");
StateActionValuesConsoleRenderer.RenderHighestValues(qValues);
System.Console.WriteLine("");
System.Console.WriteLine("q learning Greedy path:");
ConsolePathRenderer.RenderPath(GreedyPath(env, qValues));
}
public void AfterImprovingEstimates_StartingPosition_HasActionValues()
{
var env = new CliffWalkingEnvironment();
var agent = new SarsaCliffWalker();
var values = agent.ImproveEstimates(env, out var diags);
Assert.IsNotEmpty(values.ActionValues(new Position(0, 0)));
}