public void TauFirst()
{
uint numActions = 10;
uint tau = 0;
string uniqueKey = "ManagedTestId";
TestRecorder <TestContext> recorder = new TestRecorder <TestContext>();
TestPolicy policy = new TestPolicy();
MwtExplorer <TestContext> mwtt = new MwtExplorer <TestContext>("mwt", recorder);
TestContext testContext = new TestContext()
{
Id = 100
};
var explorer = new TauFirstExplorer <TestContext>(policy, tau, numActions);
uint expectedAction = policy.ChooseAction(testContext);
uint chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext);
Assert.AreEqual(expectedAction, chosenAction);
var interactions = recorder.GetAllInteractions();
Assert.AreEqual(0, interactions.Count);
}
public void UsageBadVariableActionContext()
{
int numExceptionsCaught = 0;
int numExceptionsExpected = 5;
var tryCatchArgumentException = (Action <Action>)((action) => {
try
{
action();
}
catch (ArgumentException ex)
{
if (ex.ParamName.ToLower() == "ctx")
{
numExceptionsCaught++;
}
}
});
tryCatchArgumentException(() => {
var mwt = new MwtExplorer <TestContext>("test", new TestRecorder <TestContext>());
var policy = new TestPolicy <TestContext>();
var explorer = new EpsilonGreedyExplorer <TestContext>(policy, 0.2f);
mwt.ChooseAction(explorer, "key", new TestContext());
});
tryCatchArgumentException(() =>
{
var mwt = new MwtExplorer <TestContext>("test", new TestRecorder <TestContext>());
var policy = new TestPolicy <TestContext>();
var explorer = new TauFirstExplorer <TestContext>(policy, 10);
mwt.ChooseAction(explorer, "key", new TestContext());
});
tryCatchArgumentException(() =>
{
var mwt = new MwtExplorer <TestContext>("test", new TestRecorder <TestContext>());
var policies = new TestPolicy <TestContext> [2];
for (int i = 0; i < 2; i++)
{
policies[i] = new TestPolicy <TestContext>(i * 2);
}
var explorer = new BootstrapExplorer <TestContext>(policies);
mwt.ChooseAction(explorer, "key", new TestContext());
});
tryCatchArgumentException(() =>
{
var mwt = new MwtExplorer <TestContext>("test", new TestRecorder <TestContext>());
var scorer = new TestScorer <TestContext>(10);
var explorer = new SoftmaxExplorer <TestContext>(scorer, 0.5f);
mwt.ChooseAction(explorer, "key", new TestContext());
});
tryCatchArgumentException(() =>
{
var mwt = new MwtExplorer <TestContext>("test", new TestRecorder <TestContext>());
var scorer = new TestScorer <TestContext>(10);
var explorer = new GenericExplorer <TestContext>(scorer);
mwt.ChooseAction(explorer, "key", new TestContext());
});
Assert.AreEqual(numExceptionsExpected, numExceptionsCaught);
}
public void EpsilonGreedy()
{
uint numActions = 10;
float epsilon = 0f;
string uniqueKey = "ManagedTestId";
TestRecorder <TestContext> recorder = new TestRecorder <TestContext>();
TestPolicy policy = new TestPolicy();
MwtExplorer <TestContext> mwtt = new MwtExplorer <TestContext>("mwt", recorder);
TestContext testContext = new TestContext();
testContext.Id = 100;
var explorer = new EpsilonGreedyExplorer <TestContext>(policy, epsilon, numActions);
uint expectedAction = policy.ChooseAction(testContext);
uint chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext);
Assert.AreEqual(expectedAction, chosenAction);
chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext);
Assert.AreEqual(expectedAction, chosenAction);
var interactions = recorder.GetAllInteractions();
Assert.AreEqual(2, interactions.Count);
Assert.AreEqual(testContext.Id, interactions[0].Context.Id);
}
public void EpsilonGreedyFixedActionUsingVariableActionInterface()
{
uint numActions = 10;
float epsilon = 0f;
var policy = new TestPolicy <TestVarContext>();
var testContext = new TestVarContext(numActions);
var explorer = new EpsilonGreedyExplorer <TestVarContext>(policy, epsilon);
EpsilonGreedyWithContext(numActions, testContext, policy, explorer);
}
public void EpsilonGreedy()
{
uint numActions = 10;
float epsilon = 0f;
var policy = new TestPolicy <TestContext>();
var testContext = new TestContext();
var explorer = new EpsilonGreedyExplorer <TestContext>(policy, epsilon, numActions);
EpsilonGreedyWithContext(numActions, testContext, policy, explorer);
}
public void TauFirstFixedActionUsingVariableActionInterface()
{
uint numActions = 10;
uint tau = 0;
var testContext = new TestVarContext(numActions)
{
Id = 100
};
var policy = new TestPolicy <TestVarContext>();
var explorer = new TauFirstExplorer <TestVarContext>(policy, tau);
TauFirstWithContext(numActions, testContext, policy, explorer);
}
public void TauFirst()
{
uint numActions = 10;
uint tau = 0;
TestContext testContext = new TestContext()
{
Id = 100
};
var policy = new TestPolicy <TestContext>();
var explorer = new TauFirstExplorer <TestContext>(policy, tau, numActions);
TauFirstWithContext(numActions, testContext, policy, explorer);
}
public void Bootstrap()
{
uint numActions = 10;
uint numbags = 2;
string uniqueKey = "ManagedTestId";
TestRecorder <TestContext> recorder = new TestRecorder <TestContext>();
TestPolicy[] policies = new TestPolicy[numbags];
for (int i = 0; i < numbags; i++)
{
policies[i] = new TestPolicy(i * 2);
}
TestContext testContext1 = new TestContext()
{
Id = 99
};
TestContext testContext2 = new TestContext()
{
Id = 100
};
MwtExplorer <TestContext> mwtt = new MwtExplorer <TestContext>("mwt", recorder);
var explorer = new BootstrapExplorer <TestContext>(policies, numActions);
uint expectedAction = policies[0].ChooseAction(testContext1);
uint chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext1);
Assert.AreEqual(expectedAction, chosenAction);
chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext2);
Assert.AreEqual(expectedAction, chosenAction);
var interactions = recorder.GetAllInteractions();
Assert.AreEqual(2, interactions.Count);
Assert.AreEqual(testContext1.Id, interactions[0].Context.Id);
Assert.AreEqual(testContext2.Id, interactions[1].Context.Id);
}
public void BootstrapFixedActionUsingVariableActionInterface()
{
uint numActions = 10;
uint numbags = 2;
var testContext1 = new TestVarContext(numActions)
{
Id = 99
};
var testContext2 = new TestVarContext(numActions)
{
Id = 100
};
var policies = new TestPolicy <TestVarContext> [numbags];
for (int i = 0; i < numbags; i++)
{
policies[i] = new TestPolicy <TestVarContext>(i * 2);
}
var explorer = new BootstrapExplorer <TestVarContext>(policies);
BootstrapWithContext(numActions, testContext1, testContext2, policies, explorer);
}
public void Bootstrap()
{
uint numActions = 10;
uint numbags = 2;
TestContext testContext1 = new TestContext()
{
Id = 99
};
TestContext testContext2 = new TestContext()
{
Id = 100
};
var policies = new TestPolicy <TestContext> [numbags];
for (int i = 0; i < numbags; i++)
{
policies[i] = new TestPolicy <TestContext>(i * 2);
}
var explorer = new BootstrapExplorer <TestContext>(policies, numActions);
BootstrapWithContext(numActions, testContext1, testContext2, policies, explorer);
}
private static void EpsilonGreedyWithContext <TContext>(uint numActions, TContext testContext, TestPolicy <TContext> policy, IExplorer <TContext> explorer)
where TContext : TestContext
{
string uniqueKey = "ManagedTestId";
TestRecorder <TContext> recorder = new TestRecorder <TContext>();
MwtExplorer <TContext> mwtt = new MwtExplorer <TContext>("mwt", recorder);
testContext.Id = 100;
uint expectedAction = policy.ChooseAction(testContext);
uint chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext);
Assert.AreEqual(expectedAction, chosenAction);
chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext);
Assert.AreEqual(expectedAction, chosenAction);
var interactions = recorder.GetAllInteractions();
Assert.AreEqual(2, interactions.Count);
Assert.AreEqual(testContext.Id, interactions[0].Context.Id);
// Verify that policy action is chosen all the time
explorer.EnableExplore(false);
for (int i = 0; i < 1000; i++)
{
chosenAction = mwtt.ChooseAction(explorer, uniqueKey, testContext);
Assert.AreEqual(expectedAction, chosenAction);
}
}
