Open JoeZiminski opened 5 months ago
See https://github.com/neuroinformatics-unit/movement/issues/24 and the linked articles to make an informed decision on PyDantic
vs attrs
. They both do very similar things, but int he end I went with attrs
for movement
.
Thanks @niksirbi I think it makes sense to follow your lead, also good for consistency across NIU repos.
It would be nice to properly validate all inputs to public functions. This could be done following the movement way.
Currently this is done ad-hoc in datashuttle e.g. #375. This should be reworked to be centralised in a argument-validation module.