private void EndToEnd(MwtExplorer <SimpleContext> mwtt, IExplorer <SimpleContext> explorer, TestRecorder <SimpleContext> recorder)
{
uint numActions = 10;
Random rand = new Random();
List <float> rewards = new List <float>();
for (int i = 0; i < 1000; i++)
{
Feature[] f = new Feature[rand.Next(800, 1201)];
for (int j = 0; j < f.Length; j++)
{
f[j].Id = (uint)(j + 1);
f[j].Value = (float)rand.NextDouble();
}
SimpleContext c = new SimpleContext(f);
mwtt.ChooseAction(explorer, i.ToString(), c);
rewards.Add((float)rand.NextDouble());
}
var testInteractions = recorder.GetAllInteractions();
Interaction[] partialInteractions = new Interaction[testInteractions.Count];
for (int i = 0; i < testInteractions.Count; i++)
{
partialInteractions[i] = new Interaction()
{
ApplicationContext = new OldSimpleContext(testInteractions[i].Context.GetFeatures(), null),
ChosenAction = testInteractions[i].Action,
Probability = testInteractions[i].Probability,
Id = testInteractions[i].UniqueKey
};
}
MwtRewardReporter mrr = new MwtRewardReporter(partialInteractions);
for (int i = 0; i < partialInteractions.Length; i++)
{
Assert.AreEqual(true, mrr.ReportReward(partialInteractions[i].GetId(), rewards[i]));
}
Interaction[] completeInteractions = mrr.GetAllInteractions();
MwtOptimizer mop = new MwtOptimizer(completeInteractions, numActions);
string modelFile = "model";
mop.OptimizePolicyVWCSOAA(modelFile);
Assert.IsTrue(System.IO.File.Exists(modelFile));
float evaluatedValue = mop.EvaluatePolicyVWCSOAA(modelFile);
Assert.IsFalse(float.IsNaN(evaluatedValue));
System.IO.File.Delete(modelFile);
}
