def build_features(self):
first_features = [
Feature("my_first_feature",
self.compute_feature_1,
offset=offset) for offset in self.offsets
]
second_features = [
Feature("my_second_feature",
self.compute_feature_2,
offset=offset) for offset in self.offsets
]
return first_features + second_features
def test_feature_should_work_with_cache(self):
# Given
def fn(tokens, token_index):
value = tokens[token_index].value
return "%s_%s" % (value, len(value))
mocked_fn = MagicMock(side_effect=fn)
cache = [{TOKEN_NAME: token} for token in
tokenize("hello beautiful world", LANGUAGE_EN)]
feature = Feature("test_feature", mocked_fn, offset=0)
feature.compute(2, cache)
feature1 = Feature("test_feature", mocked_fn, offset=1)
feature2 = Feature("test_feature", mocked_fn, offset=2)
# When
res1 = feature1.compute(1, cache)
res1_bis = feature1.compute(0, cache)
res2 = feature2.compute(0, cache)
# Then
self.assertEqual(res1, "world_5")
self.assertEqual(res1_bis, "beautiful_9")
self.assertEqual(res2, "world_5")
self.assertEqual(mocked_fn.call_count, 2)
def test_feature_should_work_with_offset(self):
# Given
def fn(tokens, token_index):
value = tokens[token_index].value
return "%s_%s" % (value, len(value))
cache = [{TOKEN_NAME: token} for token in
tokenize("hello beautiful world", LANGUAGE_EN)]
feature = Feature("test_feature", fn, offset=1)
# When
res = feature.compute(1, cache)
# Then
self.assertEqual(res, "world_5")
def build_features(self):
return [
Feature(base_name=self.feature_name,
func=self.compute_feature,
offset=offset,
drop_out=self.drop_out) for offset in self.offsets
]
def test_feature_should_work_with_offset(self):
# Given
def fn(tokens, token_index):
value = tokens[token_index].value
return "%s_%s" % (value, len(value))
cache = [{
TOKEN_NAME: token
} for token in tokenize("hello beautiful world", LANGUAGE_EN)]
feature = Feature("test_feature", fn, offset=1)
# When
res = feature.compute(1, cache)
# Then
self.assertEqual(res, "world_5")
def build_features(self,
builtin_entity_parser=None,
custom_entity_parser=None):
return [
Feature(base_name=self.feature_name,
func=self.compute_feature,
offset=offset,
drop_out=self.drop_out) for offset in self.offsets
]
def build_features(self):
features = []
for name, collection in iteritems(self.collections):
# We need to call this wrapper in order to properly capture
# `collection`
collection_match = self._build_collection_match_fn(collection)
for offset in self.offsets:
feature = Feature("entity_match_%s" % name, collection_match,
offset, self.drop_out)
features.append(feature)
return features
def build_features(self):
features = []
for entity_name in self.entities:
# We need to call this wrapper in order to properly capture
# `entity_name`
entity_match = self._build_entity_match_fn(entity_name)
for offset in self.offsets:
feature = Feature("entity_match_%s" % entity_name,
entity_match, offset, self.drop_out)
features.append(feature)
return features
def build_features(self):
features = []
for builtin_entity in self.builtin_entities:
# We need to call this wrapper in order to properly capture
# `builtin_entity`
builtin_entity_match = self._build_entity_match_fn(builtin_entity)
for offset in self.offsets:
feature_name = "builtin_entity_match_%s" % builtin_entity
feature = Feature(feature_name, builtin_entity_match, offset,
self.drop_out)
features.append(feature)
return features
def test_feature_should_work_with_cache(self):
# Given
def fn(tokens, token_index):
value = tokens[token_index].value
return "%s_%s" % (value, len(value))
mocked_fn = MagicMock(side_effect=fn)
cache = [{
TOKEN_NAME: token
} for token in tokenize("hello beautiful world", LANGUAGE_EN)]
feature = Feature("test_feature", mocked_fn, offset=0)
feature.compute(2, cache)
feature1 = Feature("test_feature", mocked_fn, offset=1)
feature2 = Feature("test_feature", mocked_fn, offset=2)
# When
res1 = feature1.compute(1, cache)
res1_bis = feature1.compute(0, cache)
res2 = feature2.compute(0, cache)
# Then
self.assertEqual(res1, "world_5")
self.assertEqual(res1_bis, "beautiful_9")
self.assertEqual(res2, "world_5")
self.assertEqual(mocked_fn.call_count, 2)
