def test_tag_distribution(corpus):
ut = unigramtagger.UnigramTagger()
ut.train(corpus)
expected_distribution = {'VERB': 2 / 3, 'NOUN': 1 / 3}
assert ut.N == 3
assert ut.tag_distribution == expected_distribution
def test_train(corpus):
ut = unigramtagger.UnigramTagger()
expected_model = {'train': {'VERB': 2, 'NOUN': 1}}
ut.train(corpus)
assert ut.model == expected_model
def test_train(self):
corpus = self.corpus()
ut = unigramtagger.UnigramTagger()
expected_model = {'train': {'VERB': 2, 'NOUN': 1}}
ut.train(corpus)
self.assertEqual(ut.model, expected_model)
def test_load(corpus):
testfile = tempfile.NamedTemporaryFile()
ut = unigramtagger.UnigramTagger()
# For saving a file to load later
_ut_saver = unigramtagger.UnigramTagger()
_ut_saver.train(corpus)
_ut_saver.save(testfile.name)
expected_distribution = {'VERB': 2 / 3, 'NOUN': 1 / 3}
expected_model = {'train': Counter(VERB=2, NOUN=1)}
ut.load(testfile.name)
assert (expected_model == ut.model
and expected_distribution == ut.tag_distribution)
def test_tag_distribution(self):
corpus = self.corpus()
ut = unigramtagger.UnigramTagger()
ut.train(corpus)
expected_distribution = {'VERB': 2 / 3, 'NOUN': 1 / 3}
self.assertEqual(ut.N, 3)
self.assertEqual(ut.tag_distribution, expected_distribution)
def test_tag(corpus):
ut = unigramtagger.UnigramTagger()
ut.train(corpus)
expected_list = [('train', 'VERB')]
words_to_tag = ['train']
tagged_words = ut.tag(words_to_tag)
assert expected_list == tagged_words
def test_tag(self):
corpus = self.corpus()
ut = unigramtagger.UnigramTagger()
ut.train(corpus)
expected_list = [('train', 'VERB')]
words_to_tag = ['train']
tagged_words = ut.tag(words_to_tag)
self.assertEqual(expected_list, tagged_words)
def test_save(corpus):
testfile = tempfile.NamedTemporaryFile()
ut = unigramtagger.UnigramTagger()
ut.train(corpus)
expected_data = ({
'train': Counter(VERB=2, NOUN=1)
}, {
'VERB': 2 / 3,
'NOUN': 1 / 3
})
ut.save(testfile.name)
actual_data = pickle.load(testfile)
assert expected_data == actual_data