def test_arguments(self):
try:
CountVectorizer(max_doc_freq=-1)
except ValueError:
assert True
try:
CountVectorizer(max_features=-1)
except ValueError:
assert True
def test_transform(self):
vectorizer = CountVectorizer(max_doc_freq=2,
min_freq=1,
max_features=1)
vectorizer.fit(self.docs)
sequences, X = vectorizer.transform(self.docs)
npt.assert_array_equal(X.A, np.asarray([[0], [2], [0]]))
vectorizer.binary = True
_, X1 = vectorizer.fit_transform(self.docs)
_, X2 = vectorizer.transform(self.docs)
npt.assert_array_equal(X1.A, X2.A)
def test_with_special_tokens(self):
vectorizer = CountVectorizer(max_doc_freq=2,
min_freq=1,
max_features=1)
vectorizer.fit(self.docs)
new_vocab = Vocabulary(vectorizer.vocab.idx2tok,
use_special_tokens=True)
vectorizer.vocab = new_vocab
sequences, X = vectorizer.transform(self.docs)
npt.assert_array_equal(X.A, np.asarray([[0], [2], [0]]))
def test_bad_freq_arguments(self):
vectorizer = CountVectorizer(max_doc_freq=2, min_freq=3)
try:
vectorizer.fit(self.docs)
except ValueError:
assert True