Repository navigation
redundant estimation procedures in DB #1216
Description
In data/sql/estimation_procedure.sql:
(5, 1, '10% Holdout set', 'holdout', 1, NULL, 'false', 33, 'true', 'false', '2014-12-31 21:00:00'),
(6, 1, '33% Holdout set', 'holdout', 1, NULL, 'false', 33, 'true', 'false', '2014-12-31 21:00:00'),
Surely the first 33 ought to be 10?
These records (and possibly others) seem redundant (and the 10% case has also a wrong percentage field):
(27, 2, 'Test on Training Data', 'testontrainingdata', NULL, NULL, 'false', NULL, NULL, 'false', '2019-03-16 11:30:14'),
(29, 9, '10-fold Crossvalidation', 'crossvalidation', 1, 10, 'false', NULL, 'true', 'false', '2014-12-31 20:00:00'),
(30, 10, '10-fold Crossvalidation', 'crossvalidation', 1, 10, 'false', NULL, 'true', 'false', '2023-02-22 11:46:54'),
(31, 10, '5 times 2-fold Crossvalidation', 'crossvalidation', 5, 2, 'false', NULL, 'true', 'false', '2023-02-22 11:46:54'),
(32, 10, '10 times 10-fold Crossvalidation', 'crossvalidation', 10, 10, 'false', NULL, 'true', 'false', '2023-02-22 11:46:54'),
(33, 10, '10% Holdout set', 'holdout', 1, NULL, 'false', 33, 'true', 'false', '2023-02-22 11:46:54'),
(34, 10, '33% Holdout set', 'holdout', 1, NULL, 'false', 33, 'true', 'false', '2023-02-22 11:46:54'),
(35, 11, '33% Holdout set', 'holdout', 1, NULL, 'false', 33, 'true', 'false', '2023-06-15 16:34:54');
Also, dead link in https://github.com/openml/OpenML/blob/d2f1cab55d6bea95e9067c1a54af3a2b04b966ae/CONTRIBUTING.md
Activity
NandiniDhanrale commentedon Feb 22, 2026
Hi @degrave , I’d like to work on this issue.
My plan is to:
correct the percentage field for the “10% Holdout set” entries in data/sql/estimation_procedure.sql,
review and either fix or remove the redundant estimation procedure records (IDs 27, 29–35) as suggested, and
update the dead link in CONTRIBUTING.md to the current valid target.
Please let me know if there are any constraints on changing historical estimation procedures (e.g., for reproducibility) I should keep in mind.