Open
Description
Based on @GaelVaroquaux's presentation on open-source, we have to say that it is very important to have standardized names and to use exactly the same variable to avoid extra cognitive load.
This image comes from his slides:

I feel like we are not doing this; the case of datasets is more obvious. We sometimes define subjects in some dataset objects, we define subject_list to get data using paradigm, and subjects for the function inside the dataset.
I am in favour of using subjects and putting the type that is a list or doing the cast.
Ideally, we would have some automatic way to do the checking and would be careful with the changes.
Metadata
Metadata
Assignees
Labels
No labels