def test_low_ratio_cosine_similarity_classifier(self):
train_data = list()
train_data.append("Hello world")
train_data.append("Uncle Bob")
test_data = list()
test_data.append("It's an outlier")
test_data.append("Hello world")
classifier = ClusterClassifierFactory(
train_repository=InMemoryRepository(data=train_data),
test_repository=InMemoryRepository(data=test_data),
notifier=InMemoryBroker())
classifier.add_pipeline(CosineSimilarityPipeline(ratio=.01))
classifier = classifier.compile()
outliers = classifier.detect_outliers()
self.assertEqual(len(outliers), 1, "it should retrieve 1 outliers")
self.assertEqual(len(classifier.notifier.queue.get()), 1,
"it should retrieve 1 outliers notified to Broker")
def test_cosine_similarity_classifier_with_persisted_train_clusters(self):
outlier_persister, train_persister = self.get_persisters()
test_data, train_data = self.get_train_and_test_data()
train_persister.save(object=None)
classifier = ClusterClassifierFactory(
train_repository=InMemoryRepository(data=train_data),
test_repository=InMemoryRepository(data=test_data),
notifier=InMemoryBroker())
classifier.add_outlier_persister(outlier_persister)
classifier.add_pipeline(CosineSimilarityPipeline(ratio=.70))
classifier = classifier.compile()
outliers = classifier.detect_outliers()
self.assertEqual(len(outliers), 3, "it should retrieve 3 outliers")
self.assertEqual(len(classifier.notifier.queue.get()), 3,
"it should retrieve 3 outliers notified to Broker")
self.assertEqual(len(classifier.outlier_persister.get()), 3,
"it should retrieve 3 outliers clusters")
classifier.outlier_persister.remove()