recsys_genres.py

import ast
from itertools import combinations

movies_df['genre_list'] = movies_df['genre'].apply(ast.literal_eval)
tf_genres = TfidfVectorizer(analyzer=lambda s: (c for i in range(1,4) for c in combinations(s, r=i)))
tfidf_genres = tf_genres.fit_transform(movies_df['genre_list'])

item_matrix_genres_trainset_loocv = get_item_matrix_with_inner_ids(tfidf_genres.todense(), movies_df, train_loocv)
cosine_sim_genres_trainset_loocv = cosine_similarity(item_matrix_genres_trainset_loocv,
                                                        item_matrix_genres_trainset_loocv)
item_matrix_genres_trainset = get_item_matrix_with_inner_ids(tfidf_genres.todense(), movies_df, trainset)
cosine_sim_genres_trainset = cosine_similarity(item_matrix_genres_trainset, item_matrix_genres_trainset)

get_algorithm_report(CustomSimKNNAlgorithm, trainset, testset, train_loocv, test_loocv, movies_df,
                     target_movie_id='movie_1', target_user_id='user_1', top_k=10,
                     algo_kwargs_trainset=dict(similarities=cosine_sim_genres_trainset, sim_options={'user_based': False}),
                     algo_kwargs_trainset_loocv=dict(similarities=cosine_sim_genres_trainset_loocv, sim_options={'user_based': False}))