-
Notifications
You must be signed in to change notification settings - Fork 4
/
match.py
199 lines (181 loc) · 6.35 KB
/
match.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
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
import zoneinfo
from django.contrib.auth import get_user_model
from django.db.models import Count
import numpy as np
from sklearn.neighbors import NearestNeighbors
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import LabelEncoder
from social.models import SocialProfile
from users.models import User, UserStatus
User = get_user_model()
"""
Todo
- use Preferred day, time, and timezone deviation to find closest match
"""
# substantial ML functionality lifted from connectdome's existing codebase and modified
def transform_variables_profile(social_profile):
"""Transforms the variables so that they work well on knn for 1:1"""
# idea_status
idea = 0
if social_profile[1] is None or social_profile[1] == "Open to exploring ideas.":
idea = 1
elif social_profile[1] == "Need people working on my idea.":
idea = 2
elif social_profile[1] == "Not open to exploring ideas.":
idea = 3
# video_call_friendly
vid = 0
if social_profile[2] is None or social_profile[2] is False:
vid = 0
elif social_profile[2] is True:
vid = 1
transformed_profile = [social_profile[0], idea, vid]
# raw_xp
xp = 0
if social_profile[3] is None or social_profile[3] < 3:
xp = 0
elif social_profile[3] < 6:
xp = 1
elif social_profile[3] < 10:
xp = 2
elif social_profile[3] < 15:
xp = 3
else:
xp = 4
transformed_profile.append(xp)
# location
location = social_profile[4]
if location == "":
transformed_profile.append(0)
elif location in [
"Australia",
"Canada",
"European Union",
"United States",
"United Kingdom",
"Japan",
]:
transformed_profile.append(1)
else:
transformed_profile.append(2)
# timezone
if social_profile[5] is not None:
le = LabelEncoder()
list_of_timezones = list(zoneinfo.available_timezones())
new_list = np.array(list_of_timezones, dtype=object)
if str(social_profile[5]).startswith("UTC"):
new_list = np.append(new_list, np.array(["UTC"]))
else:
new_list = np.append(
new_list, np.array([str(social_profile[5]).split("/")[0]])
)
le.fit(new_list)
transformed_profile.append(le.transform([str(social_profile[5]).split("/")[0]]))
else:
transformed_profile.append(0)
return transformed_profile
def get_indirect_match(user, number_of_matches_to_make=1):
pass
def get_one_one_match(user, number_of_matches_to_make=1):
indirect_matching = False
logged_in_social_profile = SocialProfile.objects.get(user=user)
base_profiles = (
SocialProfile.objects.exclude(user=user)
.filter(user__userstatus__approved=True)
.exclude(
available_always_off=True
) # for users who just logged in and haven't set their availability yet
.exclude(available_this_week=False)
.exclude(blocked=logged_in_social_profile)
.exclude(shadowed=logged_in_social_profile)
.exclude(circle=logged_in_social_profile)
.exclude(skipped=logged_in_social_profile)
.values_list(
"user",
"idea_status",
"video_call_friendly",
"raw_xp",
"location",
"timezone",
)
)
all_profiles = [
transform_variables_profile(social_profile)
for social_profile in list(base_profiles)
]
indirect_all_profiles = [
transform_variables_profile(social_profile)
for social_profile in list(
SocialProfile.objects.exclude(user=user)
.filter(user__userstatus__approved=True)
.exclude(
available_always_off=True
) # for users who just logged in and haven't set their availability yet
.filter(indirect_matching=True)
.annotate(num_matches=Count("matches_this_week"))
.filter(num_matches__lt=2)
.filter(num_matches__gt=0)
.exclude(blocked=logged_in_social_profile)
.exclude(shadowed=logged_in_social_profile)
.exclude(circle=logged_in_social_profile)
.exclude(skipped=logged_in_social_profile)
.values_list(
"user",
"idea_status",
"video_call_friendly",
"raw_xp",
"location",
"timezone",
)
)
]
try:
all_profiles_array = np.array(all_profiles, dtype=object)[:, 1:]
scaler = StandardScaler().fit(all_profiles_array)
all_profiles_array = scaler.fit_transform(all_profiles_array)
if len(all_profiles_array) < number_of_matches_to_make:
return []
nbrs = NearestNeighbors(
n_neighbors=number_of_matches_to_make,
).fit(all_profiles_array)
except IndexError:
try:
indirect_matching = True
indirect_all_profiles_array = np.array(indirect_all_profiles, dtype=object)[
:, 1:
]
scaler = StandardScaler().fit(indirect_all_profiles_array)
indirect_all_profiles_array = scaler.fit_transform(
indirect_all_profiles_array
)
if len(indirect_all_profiles_array) < number_of_matches_to_make:
return []
nbrs = NearestNeighbors(n_neighbors=number_of_matches_to_make).fit(
indirect_all_profiles_array
)
except IndexError:
return []
user = SocialProfile.objects.filter(user=user).values_list(
"user",
"idea_status",
"video_call_friendly",
"raw_xp",
"location",
"timezone",
)[0]
nbrs = nbrs.kneighbors(
scaler.transform([transform_variables_profile(user)[1:]]),
return_distance=False,
)
# could apply filters on neighbors
# todo: add filters from pref_dev_type and pref_timezone_deviation
user_list = []
for nbr in nbrs[0]:
if indirect_matching:
user = User.objects.get(id=indirect_all_profiles[nbr][0])
else:
user = User.objects.get(id=all_profiles[nbr][0])
user_list.append(user)
if indirect_matching:
return [indirect_all_profiles[nbr][0] for nbr in nbrs[0]]
return [all_profiles[nbr][0] for nbr in nbrs[0]]