(self, dataset_dir, type='train', is_shuffle=False)
| 89 | return all_data_df |
| 90 | |
| 91 | def get_df(self, dataset_dir, type='train', is_shuffle=False): |
| 92 | file_df = pd.DataFrame() |
| 93 | file_dirs = [] |
| 94 | if type == 'train': |
| 95 | for i in self.config.DATA.TRAIN_LIST: |
| 96 | file_dirs.append("{}/train_{}/".format(dataset_dir, i)) |
| 97 | elif type == 'val': |
| 98 | for i in self.config.DATA.VAL_LIST: |
| 99 | file_dirs.append("{}/val_{}/".format(dataset_dir, i)) |
| 100 | elif type == 'test': |
| 101 | for i in self.config.DATA.TEST_LIST: |
| 102 | file_dirs.append("{}/test_{}/".format(dataset_dir, i)) |
| 103 | file_dirs.sort() |
| 104 | for index, folder in enumerate(file_dirs): |
| 105 | data_df = self.get_from_folder(folder, index) |
| 106 | file_df = pd.concat([file_df, data_df], ignore_index=True) |
| 107 | if is_shuffle: |
| 108 | file_df.sample(frac=1).reset_index(drop=True) |
| 109 | |
| 110 | return file_df |
| 111 | |
| 112 | def construct_query_dict(self, data_df, filename): |
| 113 | data_df.reset_index(drop=True, inplace=True) |
no test coverage detected