Hi all, Is there a reason we don't have caching support in

PartitionDataset

? Image running an expensive computation but in the middle an error occurs and re-run is needed. I would assume having a logic to resume where we left off would be quite handy instead of starting from all over again. Specially in the case of return

dict of Callable

for kedro to invoke. I can certainly override this but i was wondering if there was a special reason why we don't have this yet

Someone will reply to you shortly. In the meantime, this might help:

Interested in this one!

It would be great to create a feature request for this, as it would need design

hmm really? I thought it was somehow simple like adding cache: True parameter to check if file exists before we call the user function but i'm most likely missing lots of other cases 😄 In any case i'll open a request ticket for this

I think I say design because we have a few features associated with it, this might make sense as a wider piece.

https://github.com/kedro-org/kedro-plugins/issues/974 Opened it here.