def __init__(self,
model_base_dir,
pid,
features,
feature_exporter_field_map,
length_field='tokens'):
self.length_field = length_field
saved_vectorizers = self.get_vectorizers(model_base_dir, pid)
feature_names = [feature['name'] for feature in features]
feature_vectorizer_mapping = {
feature['name']: feature['vectorizer']['type']
for feature in features
}
self.preprocessors = dict()
indices, vocabs = self.create_vocabs(model_base_dir, pid,
feature_names)
self.feature_exporter_field_map = feature_exporter_field_map
self.FIELD_NAMES = list(set(self.feature_exporter_field_map.values()))
for feature in feature_vectorizer_mapping:
self.preprocessors[feature] = create_preprocessors(
preprocessor_type=feature_vectorizer_mapping[feature],
feature=feature,
vectorizer=saved_vectorizers[feature],
index=indices[feature],
vocab=vocabs[feature])
def __init__(self, model_base_dir, pid, features, feature_exporter_field_map, length_field='tokens'):
self.length_field = length_field
saved_vectorizers = self.get_vectorizers(model_base_dir, pid)
feature_names = [feature['name'] for feature in features]
feature_vectorizer_mapping = {feature['name']: feature['vectorizer']['type'] for feature in features}
self.preprocessors = dict()
indices, vocabs = self.create_vocabs(model_base_dir, pid, feature_names)
self.feature_exporter_field_map = feature_exporter_field_map
self.FIELD_NAMES = list(set(self.feature_exporter_field_map.values()))
for feature in feature_vectorizer_mapping:
self.preprocessors[feature] = create_preprocessors(preprocessor_type=feature_vectorizer_mapping[feature],
feature=feature,
vectorizer=saved_vectorizers[feature],
index=indices[feature],
vocab=vocabs[feature])
def __init__(self, model_base_dir, pid, features):
saved_vectorizers = self.get_vectorizers(model_base_dir, pid)
feature_names = [feature['name'] for feature in features]
feature_vectorizer_mapping = {
feature['name']: feature['vectorizer']['type']
for feature in features
}
self.preprocessors = dict()
indices, vocabs = self.create_vocabs(model_base_dir, pid,
feature_names)
self.FIELD_NAME = 'tokens'
for feature in feature_vectorizer_mapping:
self.preprocessors[feature] = create_preprocessors(
preprocessor_type=feature_vectorizer_mapping[feature],
feature=feature,
vectorizer=saved_vectorizers[feature],
index=indices[feature],
vocab=vocabs[feature])