Nicolas Hug comments

Results 570 comments of


                                            Nicolas Hug

Process gets killed while calculating similarity for 'jester' dataset

Did you manage to get it to work @xahiru ?

Speed up SVD

Maybe you can cache them using `joblib.memory`?

Speed up SVD

Unfortuntaely suprise does not support partial fitting. There are some discussions and potential workarounds in https://github.com/NicolasHug/Surprise/issues/208 though

Dead link

Thanks a lot for the report @seanhly , you're right, I guess it was updated to https://projects.scipy.org/scikits.html

Improve and add Spearman

what do you mean the rows of `yr`? `yr` is a dictionary so it doesn't have rows or columns. `yr` is equal to `ur` or `ir` (from the trainset object),...

Improve and add Spearman

Hey @MaFeg100 , sorry for the slow reply. I haven't looked in great details, but I think you are *probably* correct. Here is what I understand, let me know if...

Improve and add Spearman

Ok, thanks a lot for the benchmark. I agree that the computation time is well-reasonable, comparatively. I'll try to review it more in details soon. Instead of converting `yr`, would...

Slope One prediction formula does not make sense

Sorry I won't have time to dive deep into SlopeOne again, but you'll probably find some insights in the original paper https://arxiv.org/abs/cs/0702144 If the formula in the docs doesn't reflect...

get_neighbors() should not be implemented for algorithms that cant use it

It can be used by any algorithm that uses a similarity matrix and I feel that it's not worth having another base class just for that. But it's true that...

Predictions are not made for users that don't give rating to a specific item and build_anti_testset() is not helping

Sorry @RedLegend97 I'm not sure what you mean. I've you tried https://surprise.readthedocs.io/en/stable/FAQ.html#how-to-get-the-top-n-recommendations-for-each-user ?