def blending(): # saved = True if args.saved != 0 else False saved = args.saved test_file1 = ROOT_DIR + "/data/test.txt" test_file2 = ROOT_DIR + "/data/test_predict_aspect_ensemble.txt" test_texts, test_aspects = load_ab_test(test_file1, test_file2) # print(test_aspects) word2index = pickle.load(open(ROOT_DIR + "/data/vocabulary.pkl", 'rb')) f_dict = ROOT_DIR + "/dataset/polarity.json" polarity_list, polarity_dict = parse_json(f_dict) f_dict2 = ROOT_DIR + "/dataset/attribute.json" attr_list, attr_dict = parse_json(f_dict2) paths = args.test_dir.split('#') models_files = [] for path in paths: models_files.append([ os.path.join(path, f) for f in os.listdir(path) if os.path.isfile(os.path.join(path, f)) ]) test_data = Data3((test_texts, None, test_aspects), word2index, polarity_dict, args, target_dict=attr_dict) if args.use_elmo != 0: test_elmo = load_elmo(test_texts) test_data.add_feature(test_elmo) x_test = [] for dir, checkpoints_per_model in zip(paths, models_files): print(dir, checkpoints_per_model) if saved == 1: oof_test = load_oof_test(dir) else: clfs = checkpoints_per_model oof_test = get_oof_test(clfs, test_data) x_test.append(oof_test) x_test = np.stack(x_test, axis=1) print(x_test) print(x_test.shape) test_predict = np.mean(x_test, axis=1) fw = codecs.open(ROOT_DIR + "/data/test_predict_polarity_ensemble.txt", 'w', encoding='utf-8') for j, prob in enumerate(test_predict): polarity = np.argmax(prob) - 1 fw.write(test_aspects[j] + ',' + str(polarity) + '\n') time_stamp = time.asctime().replace(':', '_').split() fw.close() shutil.copy2( ROOT_DIR + "/data/test_predict_polarity_ensemble.txt", ROOT_DIR + "/data/backup/test_predict_polarity_ensemble_%s.txt" % time_stamp)
def test(): # model = Classifier() test_file1 = ROOT_DIR + "/attribute_level/data/attribute_test.txt" test_file2 = ROOT_DIR + "/attribute_level/test_predict.txt" test_texts, test_aspects = load_ab_test(test_file1, test_file2) f_w2v = ROOT_DIR + "/embedding/embedding_all_merge_300.txt" W, word2index = load_w2v(f_w2v) f_dict1 = ROOT_DIR + "/dataset/polarity.json" f_dict2 = ROOT_DIR + "/dataset/attribute.json" polarity_list, polarity_dict = parse_json(f_dict1) attr_list, attr_dict = parse_json(f_dict2) assert len(test_texts) == len(test_aspects) files = ["checkpoint_HEAT_0.7189.pt", "checkpoint_HEAT_0.7062.pt"] predicts = [] for check_point in files: predict = [] classifier = torch.load(check_point) for text, aspect in zip(test_texts, test_aspects): if aspect != '': if aspect is None: print("error") test_data = Data3(([text], [None], [aspect]), word2index, polarity_dict, args, target_dict=attr_dict) test_predict = train_single.predict(classifier, test_data, args) assert len(test_predict) == 1 polarity = str(test_predict[0].item() - 1) else: print(aspect) print(text) polarity = '0' # fw.write(aspect+','+polarity+'\n') predict.append(aspect + ',' + polarity) predicts.append(predict) print(len(predicts)) print(len(predicts[0])) fw = codecs.open("test_predict_polarity_ensemble.txt", 'w', encoding='utf-8') for j in range(len(predicts[0])): votes = [predicts[i][j] for i in range(len(predicts))] voted = Counter(votes).most_common(1) fw.write(voted + '\n')
def stacking(): # saved = True if args.saved != 0 else False saved = args.saved f_train = ROOT_DIR + "/data/train.txt" test_file1 = ROOT_DIR + "/data/test.txt" test_file2 = ROOT_DIR + "/data/test_predict_aspect_ensemble.txt" test_texts, test_aspects = load_ab_test(test_file1, test_file2) # print(test_aspects) fo = load_abp_raw(f_train) word2index = pickle.load(open(ROOT_DIR + "/data/vocabulary.pkl", 'rb')) f_dict = ROOT_DIR + "/dataset/polarity.json" polarity_list, polarity_dict = parse_json(f_dict) f_dict2 = ROOT_DIR + "/dataset/attribute.json" attr_list, attr_dict = parse_json(f_dict2) paths = args.test_dir.split('#') models_files = [] for path in paths: path = BASE_DIR + '/' + path models_files.append([ os.path.join(path, f) for f in os.listdir(path) if os.path.isfile(os.path.join(path, f)) ]) test_data = Data3((test_texts, None, test_aspects), word2index, polarity_dict, args, target_dict=attr_dict) if args.use_elmo != 0: test_elmo = load_elmo(test_texts) test_data.add_feature(test_elmo) x_train = [] y_train = [] x_test = [] for dir, checkpoints_per_model in zip(paths, models_files): print(dir, checkpoints_per_model) dir = BASE_DIR + '/' + dir if saved == 1: oof_train, oof_train_y, oof_test = load_oof_dir(dir) else: print(checkpoints_per_model) NFOLDS = len(checkpoints_per_model) print(NFOLDS) # assert NFOLDS == args.folds clfs = [None for i in range(NFOLDS)] for cp in checkpoints_per_model: fold = int(cp.replace('_', '.').split('.')[-2]) clfs[fold - 1] = cp if saved == 2: oof_train, oof_train_y, oof_test = load_oof( clfs, fo, test_data, word2index, polarity_dict=polarity_dict, attr_dict=attr_dict) elif saved == 3: oof_train, oof_train_y, oof_test = load_oof3( clfs, fo, test_data, word2index, polarity_dict=polarity_dict, attr_dict=attr_dict) elif saved == 0: oof_train, oof_train_y, oof_test = get_oof( clfs, fo, test_data, word2index, polarity_dict=polarity_dict, attr_dict=attr_dict) else: print("saved error, [0:3]") exit(-1) x_train.append(oof_train) oof_train_y = oof_train_y.reshape(oof_train_y.shape[0], ) if y_train == []: y_train = oof_train_y else: assert (y_train == oof_train_y).all() x_test.append(oof_test) x_train = np.concatenate(x_train, axis=1) x_test = np.concatenate(x_test, axis=1) y_train = np.asarray(y_train).reshape((len(y_train), )) meta_clf = LogisticRegression() meta_clf.fit(x_train, y_train) test_predict = meta_clf.predict_proba(x_test) fw = codecs.open(ROOT_DIR + "/data/test_predict_polarity_ensemble.txt", 'w', encoding='utf-8') for j, prob in enumerate(test_predict): polarity = np.argmax(prob) - 1 fw.write(test_aspects[j] + ',' + str(polarity) + '\n') time_stamp = time.asctime().replace(':', '_').split() fw.close() shutil.copy2( ROOT_DIR + "/data/test_predict_polarity_ensemble.txt", ROOT_DIR + "/data/backup/test_predict_polarity_ensemble_%s.txt" % time_stamp)