def test_simple(self):
X = ["This was a good summer", "The food was awful"]
m = InquirerLexTransform()
Z = m.transform(X)
self.assertEqual(len(Z), 2)
self.assertTrue(isinstance(Z[0], str) and isinstance(Z[1], str))
self.assertIn("positiv", Z[0].lower())
self.assertIn("negativ", Z[1].lower())
self.assertNotIn("good", Z[0].lower())
self.assertNotIn("awful", Z[1].lower())
def build_lex_extraction(binary, min_df, ngram):
return make_pipeline(InquirerLexTransform(),
CountVectorizer(binary=binary,
tokenizer=lambda x: x.split(),
min_df=min_df,
ngram_range=(1, ngram)),
Densifier())
def test_empty(self):
m = InquirerLexTransform()
Z = m.transform([])
self.assertEqual(len(Z), 0)
def test_fit_returns_self(self):
m = InquirerLexTransform()
s = m.fit([])
self.assertEqual(s, m)
