Oscar Celma ocelma

ocelma commented on pull request ocelma/python-recsys#15
@ocelma

If you change the 'only_unknowns' parameter from False to True in the example, then you also need to update the results (the movies recommended). …

@ocelma
ocelma merged pull request ocelma/python-itunes#18
@ocelma
Add limit as a parameter for lookup method
4 commits with 5 additions and 5 deletions
ocelma commented on issue ocelma/python-recsys#14
@ocelma

I'm afraid I can't help you here. It's been a while since I tested it on Windows. No idea why it could be failing. Did you get any error message? O…

@ocelma
ocelma merged pull request ocelma/python-itunes#16
@ocelma
separate songs and music videos
2 commits with 40 additions and 7 deletions
ocelma commented on pull request ocelma/python-itunes#16
@ocelma

It looks good now. Did the tests pass?

ocelma commented on pull request ocelma/python-itunes#17
@ocelma

Yes, I think some IDs in the tests are wrong. I had to change it before. Feel free to fix that, as well :-) Then, I'm happy to merge and pull your…

ocelma commented on pull request ocelma/python-itunes#16
@ocelma

I will merge your code, if you keep the get_tracks method (so, do not rename it to get_songs, as it will break other's people code). Or, just do th…

ocelma commented on issue ocelma/python-recsys#13
@ocelma

1) Install python nosetests 2) Download Movielens 1M dataset ( from http://grouplens.org/datasets/movielens/ ) 3) Copy it inside test folder 4) Run…

ocelma commented on issue ocelma/python-recsys#12
@ocelma

I don't quite get your comments above. My advice here is that if you have items with only one (or zero) users and you want to include them in the r…

@ocelma
KeyError in svd.recommend()