def test_predict_topic(self):
lda = CustomLda(self.text)
lda.train(2)
docs = [['good', 'one'], ['bad', 'one']]
topics = lda.predict_topic(docs)
self.assertEqual(type(topics), list)
self.assertEqual(len(topics), 2)
def test_lda_init(self):
lda = CustomLda(self.text)
self.assertEqual(type(lda.data), list)
self.assertEqual(type(lda.corpus), list)
self.assertGreater(len(lda.dictionary), 1)
self.assertGreater(len(lda.data), 1)
self.assertGreater(len(lda.corpus), 1)
def test_pickle_unpickle(self):
lda = CustomLda(self.text)
lda.train(2)
filename = os.path.join(self.test_dir, 'model.pickle')
lda.pickle(filename)
lda2 = CustomLda().unpickle(filename)
self.assertEqual(lda.corpus, lda2.corpus)
self.assertEqual(lda.data, lda2.data)
self.assertEqual(lda.dictionary, lda2.dictionary)
def test_save_lda(self):
lda = CustomLda(self.text)
lda.train(2)
filename = os.path.join(self.test_dir, 'model.lda')
lda.save_lda(filename)
self.assertTrue(os.path.isfile(filename))
def test_save_ldavis(self):
lda = CustomLda(self.text)
lda.train(2)
filename = os.path.join(self.test_dir, 'test_ldavis.html')
lda.save_ldavis(filename)
self.assertTrue(os.path.isfile(filename))
def test_get_perplexity(self):
lda = CustomLda(self.text)
lda.train(2)
perp = lda.get_preplexity()
self.assertEqual(type(perp), np.float64)
def test_get_topic_terms(self):
lda = CustomLda(self.text)
lda.train(2)
topics = lda.get_topic_terms(1)
self.assertEqual(type(topics), list)
self.assertGreater(len(topics), 1)
def test_get_coherence(self):
lda = CustomLda(self.text)
lda.train(2)
coherence = lda.get_coherence()
self.assertEqual(type(coherence), np.float64)
def test_train_multicore(self):
lda = CustomLda(self.text)
lda.train(2, workers=2)
self.assertEqual(lda.num_topics, 2)
def test_train(self):
lda = CustomLda(self.text)
lda.train(2)
self.assertEqual(lda.num_topics, 2)