def evaluate(self, test=None):
'''
'''
t0 = t = time()
self.logger.info('started evaluation')
#
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
#
if not self.test:
self.logger.error('cannot evaluate without the test data')
return
#
y_true = self.test.labels
y_pred = self.tagger.predict(self.test.features)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's generated predictions')
#
accuracy = 100 * flat_accuracy_score(y_true, y_pred)
self.logger.info('Accuracy : ' + '{:.2f}'.format(accuracy))
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's evaluated test data')
return accuracy
def tag(self, test=None, save=None):
'''
if save is not given write to stdout
'''
'''
'''
t0 = t = time()
self.logger.info('started tagging')
#
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
#
if not self.test:
self.logger.error('cannot tag without the test data')
return
#
y_pred = self.tagger.predict(self.test.features)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's generated predictions')
# print accuracy, if given data contains labels
accuracy = 0.0
y_true = self.test.labels
if [tag for s_true in y_true for tag in s_true if tag != '_']:
accuracy = 100 * flat_accuracy_score(y_true, y_pred)
self.logger.info('Accuracy : ' + '{:.2f}'.format(accuracy))
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's tagged test data')
# set ccg categories of the sentences
self.test.update_tags(y_pred)
#
if not save:
for sent in self.test.sentences:
print(sent)
else:
with open(save, 'w', encoding='utf-8') as f:
for sent in self.test.sentences:
f.write(str(sent) + '\n')
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's saved as ' + save)
return accuracy
def __init__(self,
train=None,
test=None,
log_level=LOG_LEVEL,
sent_cls=CCG_CoNLL_UD):
'''
Constructor
'''
self.logger = None
self.init_logging(log_level)
self.sent_cls = sent_cls
self.train = None
self.test = None
# if given read data
if train or test:
t0 = t = time()
self.logger.info('started feature extraction')
if train:
self.train = Data(train, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted training features')
self.logger.info('processed ' + str(self.train.num_sents) +
' sentences')
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
# conditional random fields
self.tagger = sklearn_crfsuite.CRF(algorithm='lbfgs',
c1=0.1,
c2=0.1,
max_iterations=100,
all_possible_transitions=True)
def tag(self, test=None, save=None):
'''
if save is not given write to stdout
'''
'''
'''
t0 = t = time()
self.logger.info('started evaluation')
#
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
#
if not self.test:
self.logger.error('cannot evaluate without the test data')
return
#
y_pred = self.tagger.predict(self.test.features)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's generated predictions')
# set ccg categories of the sentences
self.test.update_tags(y_pred)
#
if not save:
for sent in self.test.sentences:
print(sent)
else:
with open(save, 'w', encoding='utf-8') as f:
for sent in self.test.sentences:
f.write(str(sent) + '\n')
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's saved as ' + save)
def fit(self, train=None):
'''
'''
t0 = t = time()
self.logger.info('started training')
#
if train:
self.train = Data(train, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted training features')
self.logger.info('processed ' + str(self.train.num_sents) +
' sentences')
#
if not self.train:
self.logger.error('cannot train without the training data')
return
#
self.tagger.fit(self.train.features, self.train.labels)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's trained model')
class Supertagger(object):
'''
CRF instead Logistic
'''
def __init__(self,
train=None,
test=None,
log_level=LOG_LEVEL,
sent_cls=CCG_CoNLL_UD):
'''
Constructor
'''
self.logger = None
self.init_logging(log_level)
self.sent_cls = sent_cls
self.train = None
self.test = None
# if given read data
if train or test:
t0 = t = time()
self.logger.info('started feature extraction')
if train:
self.train = Data(train, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted training features')
self.logger.info('processed ' + str(self.train.num_sents) +
' sentences')
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
# conditional random fields
self.tagger = sklearn_crfsuite.CRF(algorithm='lbfgs',
c1=0.1,
c2=0.1,
max_iterations=100,
all_possible_transitions=True)
def fit(self, train=None):
'''
'''
t0 = t = time()
self.logger.info('started training')
#
if train:
self.train = Data(train, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted training features')
self.logger.info('processed ' + str(self.train.num_sents) +
' sentences')
#
if not self.train:
self.logger.error('cannot train without the training data')
return
#
self.tagger.fit(self.train.features, self.train.labels)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's trained model')
def evaluate(self, test=None):
'''
'''
t0 = t = time()
self.logger.info('started evaluation')
#
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
#
if not self.test:
self.logger.error('cannot evaluate without the test data')
return
#
y_true = self.test.labels
y_pred = self.tagger.predict(self.test.features)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's generated predictions')
#
accuracy = 100 * flat_accuracy_score(y_true, y_pred)
self.logger.info('Accuracy : ' + '{:.2f}'.format(accuracy))
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's evaluated test data')
return accuracy
def tag(self, test=None, save=None):
'''
if save is not given write to stdout
'''
'''
'''
t0 = t = time()
self.logger.info('started tagging')
#
if test:
self.test = Data(test, sent_cls=self.sent_cls)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) +
's extracted test features')
self.logger.info('processed ' + str(self.test.num_sents) +
' sentences')
#
if not self.test:
self.logger.error('cannot tag without the test data')
return
#
y_pred = self.tagger.predict(self.test.features)
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's generated predictions')
# print accuracy, if given data contains labels
accuracy = 0.0
y_true = self.test.labels
if [tag for s_true in y_true for tag in s_true if tag != '_']:
accuracy = 100 * flat_accuracy_score(y_true, y_pred)
self.logger.info('Accuracy : ' + '{:.2f}'.format(accuracy))
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's tagged test data')
# set ccg categories of the sentences
self.test.update_tags(y_pred)
#
if not save:
for sent in self.test.sentences:
print(sent)
else:
with open(save, 'w', encoding='utf-8') as f:
for sent in self.test.sentences:
f.write(str(sent) + '\n')
t0, t = t, time()
self.logger.info('{:.2f}'.format(t - t0) + 's saved as ' + save)
return accuracy
def tag_sent(self, sent):
'''
'''
def save(self, file_name):
'''
'''
f = open(file_name, 'wb')
pickle.dump(self.tagger, f)
f.close()
self.logger.info('saved model to ' + file_name)
def load(self, file_name):
'''
'''
f = open(file_name, 'rb')
self.tagger = pickle.load(f)
f.close()
self.logger.info('loaded model from ' + file_name)
def init_logging(self, log_level):
'''
logging config and init
'''
if not self.logger:
formatter = logging.Formatter(
'%(asctime)s-|%(name)s:%(funcName)12s|-%(levelname)s-> %(message)s'
)
self.handler = logging.StreamHandler()
self.handler.setLevel(log_level)
self.handler.setFormatter(formatter)
self.logger = logging.getLogger(self.__class__.__name__)
self.logger.addHandler(self.handler)
self.logger.setLevel(log_level)
def __del__(self):
# remove handler or else duplicate logs
self.logger.removeHandler(self.handler)
# ?
del self.logger