ML_Projs/03_RS-UBCF.py at main · yasinsutoglu/ML_Projs · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
############################################
# User-Based Collaborative Filtering
#############################################

# Step 1: Preparing the Data Set
# Step 2: Determining the Movies Watched by the User to Make a Recommendation
# Step 3: Accessing the Data and IDs of Other Users Watching the Same Movies
# Step 4: Determining the Users with the Most Similar Behavior to the User to Make a Recommendation
# Step 5: Calculating the Weighted Average Recommendation Score
# Step 6: Functionalization of the Work

#############################################
# Step 1: Preparing the Data Set

import pandas as pd
pd.set_option('display.max_columns', None)
pd.set_option('display.width', 500)
pd.set_option('display.expand_frame_repr', False)

def create_user_movie_df():
    import pandas as pd

    movie = pd.read_csv('datasets/movie.csv')
    rating = pd.read_csv('datasets/rating.csv')
    df = movie.merge(rating, how="left", on="movieId")

    comment_counts = pd.DataFrame(df["title"].value_counts())
    rare_movies = comment_counts[comment_counts["title"] <= 1000].index
    common_movies = df[~df["title"].isin(rare_movies)]
    user_movie_df = common_movies.pivot_table(index=["userId"], columns=["title"], values="rating")

    return user_movie_df

user_movie_df = create_user_movie_df()

random_user = int(pd.Series(user_movie_df.index).sample(1, random_state=45).values)

#############################################
# Step 2: Determining the Movies Watched by the User to Make a Recommendation

print(random_user)
print(10 * "--")
print(user_movie_df)

random_user_df = user_movie_df[user_movie_df.index == random_user]

movies_watched = random_user_df.columns[random_user_df.notna().any()].tolist()

user_movie_df.loc[user_movie_df.index == random_user,
                  user_movie_df.columns == "Silence of the Lambs, The (1991)"]

print(10 * "--")
print(len(movies_watched))

#############################################
# Step 3: Accessing the Data and IDs of Other Users Watching the Same Movies


movies_watched_df = user_movie_df[movies_watched]

user_movie_count = movies_watched_df.T.notnull().sum()

user_movie_count = user_movie_count.reset_index()

user_movie_count.columns = ["userId", "movie_count"]

user_movie_count[user_movie_count["movie_count"] > 20].sort_values("movie_count", ascending=False)

# user_movie_count[user_movie_count["movie_count"] == 33].count()

users_same_movies = user_movie_count[user_movie_count["movie_count"] > 20]["userId"]

# ---more dynamic way to apply---
# users_same_movies = user_movie_count[user_movie_count["movie_count"] > perc]["userId"]
# perc = len(movies_watched) * 60 / 100

#############################################
# Step 4: Determining the Users with the Most Similar Behavior to the User to Make a Recommendation

# For this case, I will perform 3 steps:
# 1. I will bring together the data of X_named_user and other users.
# 2. I will create the correlation df.
# 3. I will find the most similar users (Top Users)

final_df = pd.concat([ movies_watched_df[movies_watched_df.index.isin(users_same_movies)] , random_user_df[movies_watched] ])

final_df.T.corr()

corr_df = final_df.T.corr().unstack().sort_values().drop_duplicates()

corr_df = pd.DataFrame(corr_df, columns=["corr"])

corr_df.index.names = ['user_id_1', 'user_id_2']

corr_df = corr_df.reset_index()


top_users = corr_df[ (corr_df["user_id_1"] == random_user) & (corr_df["corr"] >= 0.65) ][ ["user_id_2", "corr"] ].reset_index(drop=True)

top_users = top_users.sort_values(by='corr', ascending=False)

top_users.rename(columns={"user_id_2": "userId"}, inplace=True)

rating = pd.read_csv('datasets/rating.csv')
top_users_ratings = top_users.merge(rating[["userId", "movieId", "rating"]], how='inner')
top_users_ratings = top_users_ratings[top_users_ratings["userId"] != random_user]

#############################################
# Step 5: Calculating the Weighted Average Recommendation Score

top_users_ratings['weighted_rating'] = top_users_ratings['corr'] * top_users_ratings['rating']

# top_users_ratings.groupby('movieId').agg({"weighted_rating": "mean"})
recommendation_df = top_users_ratings.groupby('movieId').agg({"weighted_rating": "mean"})

recommendation_df = recommendation_df.reset_index()

# print(recommendation_df[["movieId"]].nunique())

# recommendation_df[recommendation_df["weighted_rating"] > 3.7]
movies_to_be_recommend = recommendation_df[recommendation_df["weighted_rating"] > 3.7].sort_values("weighted_rating", ascending=False)

movie = pd.read_csv('datasets/movie.csv')
movies_to_be_recommend.merge(movie[["movieId", "title"]])

#############################################
# Step 6: Functionalization All of the Work
#############################################

# def create_user_movie_df():
#     import pandas as pd

#     movie = pd.read_csv('datasets/movie.csv')
#     rating = pd.read_csv('datasets/rating.csv')
#     df = movie.merge(rating, how="left", on="movieId")

#     comment_counts = pd.DataFrame(df["title"].value_counts())
#     rare_movies = comment_counts[comment_counts["title"] <= 1000].index
#     common_movies = df[~df["title"].isin(rare_movies)]
#     user_movie_df = common_movies.pivot_table(index=["userId"], columns=["title"], values="rating")

#     return user_movie_df

# user_movie_df = create_user_movie_df()

# ------------------------
# def user_based_recommender(random_user, user_movie_df, ratio=60, cor_th=0.65, score=3.7):
#     import pandas as pd

#     random_user_df = user_movie_df[user_movie_df.index == random_user]

#     movies_watched = random_user_df.columns[random_user_df.notna().any()].tolist()
#     movies_watched_df = user_movie_df[movies_watched]
#     user_movie_count = movies_watched_df.T.notnull().sum()
#     user_movie_count = user_movie_count.reset_index()
#     user_movie_count.columns = ["userId", "movie_count"]

#     perc = len(movies_watched) * ratio / 100
#     users_same_movies = user_movie_count[user_movie_count["movie_count"] > perc]["userId"]

#     final_df = pd.concat([movies_watched_df[movies_watched_df.index.isin(users_same_movies)],
#                           random_user_df[movies_watched]])

#     corr_df = final_df.T.corr().unstack().sort_values().drop_duplicates()
#     corr_df = pd.DataFrame(corr_df, columns=["corr"])
#     corr_df.index.names = ['user_id_1', 'user_id_2']
#     corr_df = corr_df.reset_index()

#     top_users = corr_df[(corr_df["user_id_1"] == random_user) & (corr_df["corr"] >= cor_th)][
#         ["user_id_2", "corr"]].reset_index(drop=True)

#     top_users = top_users.sort_values(by='corr', ascending=False)
#     top_users.rename(columns={"user_id_2": "userId"}, inplace=True)
#     rating = pd.read_csv('datasets/movie_lens_dataset/rating.csv')
#     top_users_ratings = top_users.merge(rating[["userId", "movieId", "rating"]], how='inner')
#     top_users_ratings['weighted_rating'] = top_users_ratings['corr'] * top_users_ratings['rating']

#     recommendation_df = top_users_ratings.groupby('movieId').agg({"weighted_rating": "mean"})
#     recommendation_df = recommendation_df.reset_index()

#     movies_to_be_recommend = recommendation_df[recommendation_df["weighted_rating"] > score].sort_values("weighted_rating", ascending=False)
#     movie = pd.read_csv('datasets/movie_lens_dataset/movie.csv')

#     return movies_to_be_recommend.merge(movie[["movieId", "title"]])
# ------------------------
# random_user = int(pd.Series(user_movie_df.index).sample(1).values)
# user_based_recommender(random_user, user_movie_df) # fucntion calling