A sparse matrix stores nonzero entries and their positions instead of allocating space for every cell. In recommender systems, user-item matrices are mostly empty because each user interacts with only a small share of items. section 2.2.1 section 3.3.4
For a matrix of 162,541 users and 59,047 movies, storing every cell would take about 56 GB; storing 25 million nonzero entries takes about 95 MB. CSR also supports efficient input to similarity calculations and ALS training. section 2.2.2 section 3.3.4
Listing 2.1 Creating the user-item matrix
user_ids = ratings['userId'].unique()
movie_ids = ratings['movieId'].unique()
user_to_idx = {uid: idx for idx, uid in enumerate(user_ids)} #A
movie_to_idx = {mid: idx for idx, mid in enumerate(movie_ids)} #A
idx_to_movie = {idx: mid for mid, idx in movie_to_idx.items()} #A
rows = [user_to_idx[uid] for uid in ratings['userId']] #B
cols = [movie_to_idx[mid] for mid in ratings['movieId']] #B
data = [1] * len(ratings) #B
user_item_matrix = csr_matrix(
(data, (rows, cols)),
shape=(len(user_ids), len(movie_ids))
) #CSee how interaction records, row and column indices, and matrix dimensions are assembled into a CSR user-item matrix. section 2.2.2




