Sometimes the imdb.search_movie will fail in the case, AFAIK, of a movie title like Seven (a.k.a. Se7en). If we omit the a.k.a part on IMDb, it appears to work. However, an easier way is to use the links.csv from the MovieLens datasets. They provide a movieId <-> imdb_id mapping.
We can then use the imdbpy api to call imdb.get_movie(imdb_id).
If you'd like me to open a PR for this I can, but wanted to let you know. Not sure on the changes and how it would break for the 1M dataset, but it works for the 20M dataset AFAICT.
The basic change was changing the API call, the field dataclass, and then the MOVIELENS_MOVIES_FN_TEMPLATE to links from movies.
This change took missing posters from around 1000 to around 100.
Cheers and thanks for this handy repo!
Sometimes the
imdb.search_moviewill fail in the case, AFAIK, of a movie title likeSeven (a.k.a. Se7en). If we omit the a.k.a part on IMDb, it appears to work. However, an easier way is to use thelinks.csvfrom the MovieLens datasets. They provide amovieId<->imdb_idmapping.We can then use the
imdbpyapi to callimdb.get_movie(imdb_id).If you'd like me to open a PR for this I can, but wanted to let you know. Not sure on the changes and how it would break for the 1M dataset, but it works for the 20M dataset AFAICT.
The basic change was changing the API call, the
fielddataclass, and then theMOVIELENS_MOVIES_FN_TEMPLATEtolinksfrommovies.This change took missing posters from around 1000 to around 100.
Cheers and thanks for this handy repo!