def print_trees(self, text):
parses = self.parse_text(text)
for p in parses:
s = Sentence(p[1])
print(p[0])
s.print_tree()
print('')
def read_text(self, text, aux_text=None, reset_context=True):
if self.parser is None:
self.debug_msg('creating parser...')
self.parser = Parser()
self.disamb = Disambiguation(self.hg, self.parser)
nlp_parses = self.parser.parse_text(text)
if reset_context:
self.aux_text = text
if aux_text:
self.aux_text = '%s\n%s' % (text, aux_text)
parses = [(p[0], self.read_sentence(Sentence(p[1])))
for p in nlp_parses]
for p in parses:
self.debug_msg('== extra ==')
for edg in p[1].edges:
self.debug_msg(ed.edge2str(edg))
return parses
def test(infile, model_type='rf'):
parses = read_parses(infile, test_set=True)
acc_total = 0
acc_wrong = 0
acc_predictions = Counter()
acc_true_values = Counter()
for parse in parses:
# sentence_str = parse[0].strip()
json_str = parse[1].strip()
# outcome_str = parse[2].strip()
sentence = Sentence(json_str=json_str)
transfs = [int(token) for token in parse[3].split(',')]
total = len(transfs)
hgforest = Hypergen(model_type=model_type)
hgforest.test(sentence, transfs)
wrong = hgforest.wrong
print('%s / %s' % (wrong, total))
acc_total += total
acc_wrong += wrong
acc_predictions = sum(
(acc_predictions, Counter(hgforest.test_predictions)), Counter())
acc_true_values = sum(
(acc_true_values, Counter(hgforest.test_true_values)), Counter())
acc_predictions = dict(acc_predictions)
acc_true_values = dict(acc_true_values)
print('PREDICTIONS:')
for transf in acc_predictions:
print('%s: %s' % (hgtransf.to_string(transf), acc_predictions[transf]))
print('TRUE_VALUES:')
for transf in acc_true_values:
print('%s: %s' % (hgtransf.to_string(transf), acc_true_values[transf]))
error_rate = (float(acc_wrong) / float(acc_total)) * 100.
print('error rate: %.3f%%' % error_rate)
for i in range(len(token_seq)):
token_seq[i].position_in_sentence = i
sents.append((sentence_text, token_seq))
return sents
def print_trees(self, text):
parses = self.parse_text(text)
for p in parses:
s = Sentence(p[1])
print(p[0])
s.print_tree()
print('')
if __name__ == '__main__':
test_text = u"""Some subspecies of mosquito might be 1st to be genetically wiped out."""
# test_text = u"""Des millions de Français n’ont pas accès à une connexion."""
print('Starting parser...')
parser = Parser()
print('Parsing...')
result = parser.parse_text(test_text)
for r in result:
sentence = Sentence(r[1])
print(r[0])
sentence.print_tree()
print('')
sents = []
for span in parsed_data.sents:
sent = [self.__spacy2token(parsed_data[i]) for i in range(span.start, span.end)]
sents.append(sent)
return sents
if __name__ == '__main__':
test_text = u"""
Alan Mathison Turing was a pioneering English computer scientist, mathematician, logician, cryptanalyst and
theoretical biologist. He was highly influential in the development of theoretical computer science, providing a
formalisation of the concepts of algorithm and computation with the Turing machine, which can be considered a model
of a general purpose computer. Turing is widely considered to be the father of theoretical computer science
and artificial intelligence.
The psychologist George Kelley (1955) noted that humans do not enter a world that is inherently structured; we must
give the world a structure that we ourselves create.
Some subspecies of mosquito might be 1st to be genetically wiped out.
"""
print('Starting parser...')
parser = Parser()
print('Parsing...')
result = parser.parse_text(test_text)
for r in result:
sentence = Sentence(r)
print(sentence)
sentence.print_tree()
print('')
def process_sentence(self, sentence):
self.tree.root_id = self.process_token(sentence.root())
self.tree.remove_redundant_nesting()
return self.tree
def transform(sentence):
alpha = AlphaStage()
return alpha.process_sentence(sentence)
if __name__ == "__main__":
test_text = """
My name is James Bond.
"""
print("Starting parser...")
parser = Parser()
print("Parsing...")
result = parser.parse_text(test_text)
print(result)
for r in result:
s = Sentence(r)
print(s)
s.print_tree()
t = transform(s)
print(t)
return elem_id, transf
def process_sentence(self, sentence):
self.tree.root_id = self.process_token(sentence.root())[0]
return ParserOutput(sentence, self.tree)
def transform(sentence):
alpha = AlphaForest()
return alpha.process_sentence(sentence)
if __name__ == '__main__':
# learn('cases.csv', 'alpha_forest.model')
test_text = """
Satellites from NASA and other agencies have been tracking sea ice changes since 1979.
"""
# test_text = 'Telmo is going to the gym.'
print('Starting parser...')
parser = Parser()
print('Parsing...')
result = parser.parse_text(test_text)
for r in result:
s = Sentence(r[1])
t = transform(s)
print(t.tree.to_hyperedge_str(with_namespaces=False))