def test_init(self):
corpus = Corpus.from_file('deerwester')
c2n = CorpusToNetwork(corpus)
self.assertIsInstance(c2n.corpus, Corpus)
self.assertEqual(len(c2n.corpus), 9)
with self.assertRaises(Exception):
c2n = CorpusToNetwork(corpus.domain)
def test_call_word(self):
corpus = Corpus.from_file('deerwester')
c2n = CorpusToNetwork(corpus)
result = c2n(document_nodes=False,
window_size=1,
threshold=1,
freq_threshold=1)
items = c2n.get_current_items(False)
self.assertIsInstance(result, Network)
self.assertIsInstance(items, Table)
self.assertEqual(len(items), result.number_of_nodes())
self.assertGreater(result.number_of_nodes(), len(corpus))
def test_empty(self):
corpus = Corpus.from_file('deerwester')[:0]
c2n = CorpusToNetwork(corpus)
result = c2n(document_nodes=True,
window_size=1,
threshold=1,
freq_threshold=1)
self.assertEqual(result.number_of_nodes(), 0)
self.assertEqual(result.number_of_edges(), 0)
def test_cache(self):
corpus = Corpus.from_file('deerwester')
c2n = CorpusToNetwork(corpus)
result1 = c2n(document_nodes=True,
window_size=1,
threshold=1,
freq_threshold=1)
result2 = c2n(document_nodes=True,
window_size=1,
threshold=1,
freq_threshold=1)
self.assertIs(result1, result2)
def set_data(self, data):
if Network is None:
self.Error.no_network_addon()
return
self.cancel()
self._task_state = "running"
self.button.setText("Stop")
if not data:
self._corpus_to_network = None
self.clear_outputs()
return
self.corpus = data
self._corpus_to_network = CorpusToNetwork(corpus=data)
self.commit()
def run(corpus_to_network: CorpusToNetwork, document_nodes: bool,
threshold: int, window_size: int, freq_threshold: int,
state: TaskState) -> Tuple[Network, Table]:
def advance(progress):
if state.is_interruption_requested():
raise InterruptedError
state.set_progress_value(progress)
network = corpus_to_network(document_nodes=document_nodes,
window_size=window_size,
threshold=threshold,
freq_threshold=freq_threshold,
progress_callback=advance)
items = corpus_to_network.get_current_items(document_nodes)
return (network, items)
def set_data(self, data):
if Network is None:
self.Error.no_network_addon()
return
self.cancel()
self._task_state = "running"
self.button.setText("Stop")
if not data:
self._corpus_to_network = None
self.info.set_input_summary(self.info.NoInput)
self.clear_outputs()
return
self.corpus = data
summary = str(len(self.corpus))
details = "Corpus with {} documents.".format(len(self.corpus))
self.info.set_input_summary(summary, details)
self._corpus_to_network = CorpusToNetwork(corpus=data)
self.commit()
def test_params(self):
corpus = Corpus.from_file('deerwester')
c2n = CorpusToNetwork(corpus)
result1 = c2n(document_nodes=False,
window_size=1,
threshold=1,
freq_threshold=1)
result2 = c2n(document_nodes=False,
window_size=1,
threshold=100,
freq_threshold=1)
self.assertGreater(result1.number_of_edges(),
result2.number_of_edges())
result2 = c2n(document_nodes=False,
window_size=10,
threshold=1,
freq_threshold=1)
self.assertLess(result1.number_of_edges(), result2.number_of_edges())
result2 = c2n(document_nodes=False,
window_size=1,
threshold=1,
freq_threshold=100)
self.assertGreater(result1.number_of_nodes(),
result2.number_of_nodes())