# return top-n movies
movie_ids = [
full_dataset.to_raw_iid(i)
for i in heapq.nlargest(k, candidates, key=candidates.get)
]
return self.movies.get_movie_by_movie_ids(movie_ids)
if __name__ == '__main__':
from movies_recommender.Recommender import test_recommendation
from movies_recommender.RecommenderUserBased import RecommenderUserBased
from movies_analyzer.RecommendationDataset import RecommendationDataSet
from movies_analyzer.Movies import Movies
movies = Movies()
recommendation_dataset = RecommendationDataSet(movies=movies)
recommender = RecommenderUserBased(movies)
assert recommender.__module__[:len('movies_recommender.'
)] == 'movies_recommender.'
test_recommendation(recommender,
recommendation_dataset,
example_items=['arek', 'mateusz'],
anti_test=False)
""" For test only
%load_ext autoreload
%autoreload 2
from filmweb_integrator.fwimdbmerge.filmweb import Filmweb
from filmweb_integrator.fwimdbmerge.merger import Merger, get_json_df
from tqdm import tqdm
from movies_analyzer.RecommendationDataset import RecommendationDataSet
from movies_analyzer.Movies import Movies
from movies_analyzer.Imdb import get_imdb_movie
# download all images from imdb ased on movielens, this is very long query
# for 9000 could take about 6 hours, for all imdb (500,000) could take about 2 weeks.
recommendation_dataset = RecommendationDataSet(movies=Movies())
# all imdb movies
for tmbdid in tqdm(recommendation_dataset.movies.data.index):
tmbdid = tmbdid.replace('tt', '')
get_imdb_movie(tmbdid)