def test_prepare_cv(self):
cf_cv = ClassifierCv(self.labels, self.texts)
cf_cv.prepare_pipeline([('vect', CountVectorizer()),
('tfidf', TfidfTransformer()),
('crf', MultinomialNB())])
cf_cv.prepare_cv(3)
self.assertEqual(cf_cv.kf.n_splits, 3)
self.assertIsNotNone(cf_cv.unique_labels)
def test_random_search(self):
cf_cv = ClassifierCv(self.labels, self.texts)
cf_cv.prepare_pipeline([('vect', CountVectorizer()),
('tfidf', TfidfTransformer()),
('crf', MultinomialNB())])
param_dist = {'crf__alpha': [0.01, 0.02, 0.05, 0.07, 0.09]}
cf_cv.perform_random_search(param_dist)
params = cf_cv.get_top_random_search_parameters(1)
params_print = cf_cv.print_top_random_search(1)
self.assertEqual(type(params), dict)
self.assertEqual(len(params), 1)
self.assertEqual(params_print, None)
def test_prepare_pipeline(self):
cf_cv = ClassifierCv(self.labels, self.texts)
pipeline = [
('tfidf', TfidfTransformer()),
('clf',
SGDClassifier(loss='hinge',
penalty='l2',
alpha=1e-3,
random_state=42,
max_iter=5,
tol=None)),
]
cf_cv.prepare_pipeline(pipeline)
self.assertEqual(cf_cv.text_clf._final_estimator.loss, 'hinge')
self.assertEqual(cf_cv.text_clf._final_estimator.max_iter, 5)
self.assertEqual(cf_cv.text_clf._final_estimator.penalty, 'l2')
self.assertEqual(cf_cv.text_clf._final_estimator.random_state, 42)