#### Description The task object contains the information regarding estimation parameters twice. Doing a small refactor would remove the duplicate information. #### Steps/Code to Reproduce ``` from pprint import pprint import openml task_id = 2 task = openml.tasks.get_task(task_id) pprint(vars(task)) ``` #### Result ``` {'class_labels': ['1', '2', '3', '4', '5', 'U'], 'cost_matrix': None, 'dataset_id': 1, 'estimation_parameters': {'number_folds': '2', 'number_repeats': '5', 'percentage': '', 'stratified_sampling': 'true'}, 'estimation_procedure': {'data_splits_url': 'https://test.openml.org/api_splits/get/2/Task_2_splits.arff', 'parameters': {'number_folds': '2', 'number_repeats': '5', 'percentage': '', 'stratified_sampling': 'true'}, 'type': 'crossvalidation'}, 'evaluation_measure': None, 'split': None, 'target_name': 'class', 'task_id': 2, 'task_type': 'Supervised Classification', 'task_type_id': 1} ```
Activity
PGijsbers commentedon Apr 15, 2019
I think this is outdated. Right now it looks like the actual data is not stored twice, but the same dictionary is referenced twice (e.g. here).
If it is just regarding the string representations I think this is duplicate with #653 or #75.
Feel free to re-open if I misunderstood.
ArlindKadra commentedon Apr 15, 2019
Referenced also in #656, addressed in #607