forked from FlxPr/Network_Science_Yelp_Dataset
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata_reader.py
178 lines (140 loc) · 6.37 KB
/
data_reader.py
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
import pandas as pd
from tqdm import tqdm
import json
from settings import file_names
from graph import make_friends_graph
from networkx import connected_components
def init_empty_lists(number_of_lists=4, list_length=100):
for i in range(number_of_lists):
yield [None] * list_length
def read_review_data(filtered_businesses: list = []):
"""
Reads the review data and returns it as a pandas DataFrame
The data file is 4GB. Might need to downsample it first (e.g. take only 1 country)
:return:
"""
filtered_businesses = dict.fromkeys(filtered_businesses, 1)
with open(file_names['review'], encoding='utf-8') as f:
line_count = len(f.readlines())
user_ids, business_ids, stars, dates, texts = init_empty_lists(5, line_count)
with open(file_names['review'], encoding='utf-8') as f:
for i, line in enumerate(tqdm(f, total=line_count)):
blob = json.loads(line)
if filtered_businesses is not None and filtered_businesses.get(blob['business_id']):
user_ids[i] = blob["user_id"]
business_ids[i] = blob["business_id"]
stars[i] = blob["stars"]
dates[i] = blob["date"]
texts[i] = blob["text"]
ratings = pd.DataFrame(
{"user_id": user_ids,
"business_id": business_ids,
"rating": stars,
"date": dates,
"text": texts}
).dropna(how='all')
return ratings
def read_business_data():
with open(file_names['business'], encoding='utf-8') as f:
line_count = len(f.readlines())
business_id, name, address, city = init_empty_lists(4, line_count)
state, postal_code, latitude, longitude = init_empty_lists(4, line_count)
stars, review_count, attributes, categories = init_empty_lists(4, line_count)
with open(file_names['business'], encoding='utf-8') as f:
for i, line in enumerate(tqdm(f, total=line_count)):
blob = json.loads(line)
business_id[i] = blob["business_id"]
name[i] = blob["name"]
address[i] = blob["address"]
city[i] = blob["city"]
state[i] = blob["state"]
postal_code[i] = blob["postal_code"]
latitude[i] = blob["latitude"]
longitude[i] = blob["longitude"]
stars[i] = blob["stars"]
review_count[i] = blob["review_count"]
attributes[i] = blob["attributes"]
categories[i] = blob["categories"]
df = pd.DataFrame({
'business_id': business_id,
'name': name,
'address': address,
'city': city,
'state': state,
'postal_code': postal_code,
'latitude': latitude,
'longitude': longitude,
'stars': stars,
'review_count': review_count,
'attributes': attributes,
'categories': categories
}).dropna(how='all')
return df
def filter_friend_list(friend_list: str, filtered_friends_dict):
friend_list = friend_list.split(", ")
friend_list = filter(lambda x: filtered_friends_dict.get(x), friend_list)
return ', '.join(friend_list)
def read_user_data(filtered_users_dict, parse_details=False):
with open(file_names['user'], encoding='utf-8') as f:
line_count = len(f.readlines())
with open(file_names['user'], encoding='utf-8') as f:
user_id, friends = [None] * line_count, [None] * line_count
for i, line in enumerate(tqdm(f, total=line_count)):
blob = json.loads(line)
if filtered_users_dict.get(blob["user_id"]):
user_id[i] = blob["user_id"]
friends[i] = blob["friends"]
if parse_details:
pass # TODO if useful, parse details about users:
# user_name += [blob["name"]]
# review_count += [blob["review_count"]]
# yelping_since += [blob["yelping_since"]]
# useful += [blob["yelping_since"]]
# funny += [blob["yelping_since"]]
# cool += [blob["text"]]
# elite += [blob["text"]]
df = pd.DataFrame({
'user_id': user_id,
'friends': friends
}).dropna(subset=['user_id'])
df.friends = df.friends.apply(lambda x: filter_friend_list(x, filtered_users_dict))
return df
def read_photo_data():
pass
def read_checkin_data():
pass
def read_tip_data():
pd.read_json(file_names['tip'], encoding='utf-8')
pass
def subset_toronto_data():
# Filter restaurants in Toronto
print('Filtering business data...')
business_df = read_business_data()
business_df = business_df.dropna(subset=['categories'])
business_df = business_df[(business_df.city == 'Toronto') & (business_df.categories.str.contains('Restaurant'))]
business_df.to_csv(file_names['toronto_businesses'], index=None)
# Filter reviews in Toronto
print('Filtering review data...')
reviews_df = read_review_data(filtered_businesses=business_df.business_id.unique())
reviews_df.to_csv(file_names['toronto_reviews'], index=None)
# Remove users with only one review in Toronto -- Cuts by half the number of users, by 10% the review count
review_count = reviews_df.user_id.value_counts()
reviews_df.user_id = reviews_df.user_id.apply(lambda user_id: user_id if review_count[user_id] > 1 else None)
reviews_df = reviews_df.dropna(subset=['user_id'])
# Save light version of reviews as csv for quick data loading
reviews_df.drop('text', axis=1).to_csv(file_names['toronto_reviews_without_text'])
print('Filtering user data...')
toronto_locals = dict.fromkeys(reviews_df.user_id.unique(), 1)
user_df = read_user_data(filtered_users_dict=toronto_locals)
user_df.to_csv(file_names['toronto_users'], index=None)
# Filter users who are in the largest connected component of the social network
social_network = make_friends_graph()
filtered_locals = max(connected_components(social_network), key=len)
user_df = user_df[user_df.user_id.isin(filtered_locals)]
user_df.to_csv(file_names['toronto_users'], index=None)
# Remove reviews from removed users
reviews_df = reviews_df[reviews_df.user_id.isin(filtered_locals)]
reviews_df.to_csv(file_names['toronto_reviews'], index=None)
reviews_df.drop('text', axis=1).to_csv(file_names['toronto_reviews_without_text'], index=None)
if __name__ == '__main__':
subset_toronto_data()