Skip to content

Conversation

@paoloRais
Copy link
Contributor

Modified predict so that the representations are computed only once for each user and item.

The same idea could be applied on predict_rank (I wanted to get your opinion first though).

I also added a method predict _score that accepts the same input as predict_rank, but is based on predict by unrolling the matrix as an array.

…score with same input as predict_rank, tests
@paoloRais
Copy link
Contributor Author

The appveyor fail is really strange, I wonder if it is due to the testing using Windows, since on Linux I can't replicate it.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant