ludovikcoba / rrecsys

rrecsys: Environment for Assessing Recommender Systems
https://cran.r-project.org/package=rrecsys
23 stars 11 forks source link

Implement with sparse matrix rather than matrix?? #3

Open jonnylee719 opened 8 years ago

jonnylee719 commented 8 years ago

Hey thank you for implementing this recommender system library. I have been using it to deal with some big dataset, e.g. 22531 users and 43968 items. To define the dataSet I just run out of memory since it requires a matrix. Would you be able to change it to a sparse matrix?

ludovikcoba commented 8 years ago

This is a valid comment! I am working in that direction and trying to define most elegant solution.

jonnylee719 commented 8 years ago

Thanks for your reply! Perfect!

jonnylee719 commented 8 years ago

Oh, forgot to mention, also be careful with vector operations while implementing with sparse matrix. At the recommenderlab CRAN package, it takes in sparseMatrix but at execution of the algorithms, memory runs out because vectors are too large. There is a sparseVector class as well as sparseMatrix class.