def __init__(self, corpus_path, threshold=None, crossvalidate=True): self._logger = get_logger(self.__module__ + '.' + self.__class__.__name__) self._threshold = float('-inf') if threshold is None else threshold builder = self._get_classifier_builder( train_corpus_path=corpus_path, crossvalidate=crossvalidate, logger=self._logger ) self._classifier = get_cached(builder, get_model_path(corpus_path))
try: import cPickle as pickle except ImportError: # pragma: no cover import pickle import json import csv import os from items_classifier.utils.log import get_logger from items_classifier.configs.config import MODELS_DIR _logger = get_logger(__name__) def save_as_json(data, path): with open(path, 'w') as outfile: json.dump(data, outfile) def read_as_json(path): with open(path) as data_file: data = json.load(data_file) return data def save_json_as_csv(data, path): output = csv.writer(open(path, 'wb+')) output.writerow(data.keys()) # header row max_size = max(len(data[label]) for label in data) prepared_data = []