PROBLEM: jupyter notebook are not purely reproducible
Executed jupyter notebook hold the execution time and other random values:
"id": "c618109c",
"iopub.execute_input": "2022-01-03T16:00:42.843642Z",
"iopub.status.busy": "2022-01-03T16:00:42.843242Z",
"iopub.status.idle": "2022-01-03T16:00:43.475618Z",
"shell.execute_reply": "2022-01-03T16:00:43.475259Z"
As such two execution of the exact same notebook will lead to different HASH of the results.
I see two solution:
- provide a way to exclude some files from the Hashing
- provide an option to automatically remove execution times from jupyter notebook.
- according to: stackoverflow the only way would be by hand with a user defined script.
Edited by Yori Fournier