public void TestEstimateAnomalyLikelihoods()
{
// Generate an estimate using fake distribution of anomaly scores.
List <Sample> data = GenerateSampleData(0.2, 0.2, 0.2, 0.2).Take(1000).ToList();
AnomalyLikelihoodMetrics metrics = an.EstimateAnomalyLikelihoods(data, 10, 0);
Assert.AreEqual(1000, metrics.GetLikelihoods().Length);
Assert.AreEqual(1000, metrics.GetAvgRecordList().AveragedRecords.Count);
Assert.IsTrue(an.IsValidEstimatorParams(metrics.GetParams()));
// Get the total
double total = 0;
foreach (Sample sample in metrics.GetAvgRecordList().AveragedRecords)
{
total = total + sample.score;
}
// Check that the estimated mean is correct
Statistic statistic = (Statistic)metrics.GetParams().Distribution();
Assert.IsTrue(
AssertWithinEpsilon(
statistic.mean, (total / (double)metrics.GetAvgRecordList().AveragedRecords.Count)
)
);
int count = ArrayUtils.Where(metrics.GetLikelihoods(), d => d < 0.02).Length;
Assert.IsTrue(count <= 50);
Assert.IsTrue(count >= 1);
}
public void TestSerializeAnomalyLikelihoodForUpdates()
{
Parameters @params = Parameters.Empty();
@params.SetParameterByKey(Parameters.KEY.ANOMALY_KEY_MODE, Anomaly.Mode.LIKELIHOOD);
AnomalyLikelihood an = (AnomalyLikelihood)Anomaly.Create(@params);
// Serialize the Anomaly Computer without errors
SerialConfig config = new SerialConfig("testSerializeAnomalyLikelihood", SerialConfig.SERIAL_TEST_DIR);
IPersistenceAPI api = Persistence.Get(config);
byte[] bytes = api.Write(an);
// Deserialize the Anomaly Computer and make sure its usable (same tests as AnomalyTest.java)
AnomalyLikelihood serializedAn = api.Read <AnomalyLikelihood>(bytes);
Assert.IsNotNull(serializedAn);
//----------------------------------------
// Step 1. Generate an initial estimate using fake distribution of anomaly scores.
List <Sample> data1 = AnomalyLikelihoodTest.GenerateSampleData(0.2, 0.2, 0.2, 0.2).Take(1000).ToList();
AnomalyLikelihoodMetrics metrics1 = serializedAn.EstimateAnomalyLikelihoods(data1, 5, 0);
//----------------------------------------
// Step 2. Generate some new data with a higher average anomaly
// score. Using the estimator from step 1, to compute likelihoods. Now we
// should see a lot more anomalies.
List <Sample> data2 = AnomalyLikelihoodTest.GenerateSampleData(0.6, 0.2, 0.2, 0.2).Take(300).ToList();
AnomalyLikelihoodMetrics metrics2 = serializedAn.UpdateAnomalyLikelihoods(data2, metrics1.GetParams());
// Serialize the Metrics too just to be sure everything can be serialized
SerialConfig metricsConfig = new SerialConfig("testSerializeMetrics", SerialConfig.SERIAL_TEST_DIR);
api = Persistence.Get(metricsConfig);
api.Write(metrics2);
// Deserialize the Metrics
AnomalyLikelihoodMetrics serializedMetrics = api.Read <AnomalyLikelihoodMetrics>();
Assert.IsNotNull(serializedMetrics);
Assert.AreEqual(serializedMetrics.GetLikelihoods().Length, data2.Count);
Assert.AreEqual(serializedMetrics.GetAvgRecordList().Count, data2.Count);
Assert.IsTrue(serializedAn.IsValidEstimatorParams(serializedMetrics.GetParams()));
// The new running total should be different
Assert.IsFalse(metrics1.GetAvgRecordList().Total == serializedMetrics.GetAvgRecordList().Total);
// We should have many more samples where likelihood is < 0.01, but not all
int conditionCount = ArrayUtils.Where(serializedMetrics.GetLikelihoods(), d => d < 0.1).Length;
Assert.IsTrue(conditionCount >= 25);
Assert.IsTrue(conditionCount <= 250);
}
