def test_make_train_test_split_seed_works(self):
a1, a2 = corpus.make_train_test_split("a")
b1, b2 = corpus.make_train_test_split("b")
c1, c2 = corpus.make_train_test_split("a")
self.assertEqual(a1, c1)
self.assertEqual(a2, c2)
self.assertNotEqual(a1, b1)
self.assertNotEqual(a2, b2)
def test_make_train_test_split_seed_works(self):
a1, a2 = corpus.make_train_test_split("a")
b1, b2 = corpus.make_train_test_split("b")
c1, c2 = corpus.make_train_test_split("a")
self.assertEqual(a1, c1)
self.assertEqual(a2, c2)
self.assertNotEqual(a1, b1)
self.assertNotEqual(a2, b2)
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
self.realTest = corpus.iter_test_corpus()
self.corpus = buildVocabulary(getPhrases(self.train + self.test + self.realTest))
# self.training = buildVocabulary(getPhrases(self.train))
# self.testing = buildVocabulary(getPhrases(self.test))
self.size = 50
def test_make_train_test_split_no_shared_sentences(self):
"""
Test that train and test don't share sent ids.
"""
train, test = corpus.make_train_test_split("semis")
train_ids = set(x.sentenceid for x in train)
test_ids = set(x.sentenceid for x in test)
self.assertEqual(train_ids & test_ids, set())
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
self.samples = 50000
self.xTrain = buildVocabulary(getPhrases(self.train[:self.samples]))
self.xTest = buildVocabulary(getPhrases(self.test[:self.samples]))
self.size = 150
self.labelsTrain = getLabels(self.train)
self.labelsTest = getLabels(self.test)
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
self.realTest = corpus.iter_test_corpus()
self.corpus = buildVocabulary(
getPhrases(self.train + self.test + self.realTest))
# self.training = buildVocabulary(getPhrases(self.train))
# self.testing = buildVocabulary(getPhrases(self.test))
self.size = 50
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
self.samples = 50000
self.xTrain = buildVocabulary(getPhrases(self.train[:self.samples]))
self.xTest = buildVocabulary(getPhrases(self.test[:self.samples]))
self.size = 150
self.labelsTrain = getLabels(self.train)
self.labelsTest = getLabels(self.test)
def test_make_train_test_split_no_shared_sentences(self):
"""
Test that train and test don't share sent ids.
"""
train, test = corpus.make_train_test_split("semis")
train_ids = set(x.sentenceid for x in train)
test_ids = set(x.sentenceid for x in test)
self.assertEqual(train_ids & test_ids, set())
def cross_validation(factory, seed, K=10, callback=None):
seed = str(seed)
scores = []
for k in range(K):
train, test = make_train_test_split(seed + str(k))
predictor = factory()
predictor.fit(train)
score = predictor.score(test)
if callback:
callback(score)
scores.append(score)
return sum(scores) / len(scores)
def test_simple_predict(self):
train, test = corpus.make_train_test_split("inhaler")
predictor = PhraseSentimentPredictor()
predictor.fit(train)
predictions = predictor.predict(test)
# Same amount of predictions than input values
self.assertEqual(len(predictions), len(test))
# Predicted labels where seen during training
train_labels = set(x.sentiment for x in train)
predicted_labels = set(predictions)
self.assertEqual(predicted_labels - train_labels, set())
def test_simple_error_matrix(self):
train, test = corpus.make_train_test_split("reflektor", proportion=0.4)
predictor = PhraseSentimentPredictor()
predictor.fit(train)
error = predictor.error_matrix(test)
for real, predicted in error.keys():
self.assertNotEqual(real, predicted)
score = predictor.score(test)
assert score > 0, "Test is valid only if score is more than 0"
N = float(len(test))
wrong = sum(len(xs) for xs in error.values())
self.assertEqual((N - wrong) / N, score)
def test_simple_predict(self):
train, test = corpus.make_train_test_split("inhaler")
predictor = PhraseSentimentPredictor()
predictor.fit(train)
predictions = predictor.predict(test)
# Same amount of predictions than input values
self.assertEqual(len(predictions), len(test))
# Predicted labels where seen during training
train_labels = set(x.sentiment for x in train)
predicted_labels = set(predictions)
self.assertEqual(predicted_labels - train_labels, set())
def test_simple_error_matrix(self):
train, test = corpus.make_train_test_split("reflektor", proportion=0.4)
predictor = PhraseSentimentPredictor()
predictor.fit(train)
error = predictor.error_matrix(test)
for real, predicted in error.keys():
self.assertNotEqual(real, predicted)
score = predictor.score(test)
assert score > 0, "Test is valid only if score is more than 0"
N = float(len(test))
wrong = sum(len(xs) for xs in error.values())
self.assertEqual((N - wrong) / N, score)
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
self.config = json.load(open("../data/model2.json"))
# self.samples = len(self.train)
self.samples = len(self.train)
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
def test_make_train_test_split_simple(self):
train, test = corpus.make_train_test_split("blitz")
self.assertIn("word play", [x.phrase for x in train + test])
self.assertEqual(len(set(x.sentenceid for x in test)), 1)
self.assertEqual(len(set(x.sentenceid for x in test + train)), 4)
def setUp(self):
self.train, self.test = corpus.make_train_test_split("mySeed")
def test_fit_returns_self(self):
train, _ = corpus.make_train_test_split("defiant order")
predictor = PhraseSentimentPredictor()
s = predictor.fit(train)
self.assertEqual(predictor, s)
def test_fit_returns_self(self):
train, _ = corpus.make_train_test_split("defiant order")
predictor = PhraseSentimentPredictor()
s = predictor.fit(train)
self.assertEqual(predictor, s)
def getTrainingAndTestSplitOnSize(size):
train, test = corpus.make_train_test_split("mySeed")
trainingDataWithLabels, trainingDataShortWithLabels = getWordListsGreaterThan(size, train)
testDataWithLabels, testDataShortWithLabels = getWordListsGreaterThan(size, test)
return trainingDataWithLabels, trainingDataShortWithLabels, testDataWithLabels, testDataShortWithLabels
def test_make_train_test_split_simple(self):
train, test = corpus.make_train_test_split("blitz")
self.assertIn("word play", [x.phrase for x in train + test])
self.assertEqual(len(set(x.sentenceid for x in test)), 1)
self.assertEqual(len(set(x.sentenceid for x in test + train)), 4)
def getVocabularyOfSizeGreaterThan(size):
train, test = corpus.make_train_test_split("mySeed")
trainingDataWithLabels, dummy = getWordListsGreaterThan(size, train)
testDataWithLabels, dummy = getWordListsGreaterThan(size, test)
return trainingDataWithLabels, testDataWithLabels
