-
Notifications
You must be signed in to change notification settings - Fork 0
/
SocialNetwork.py
586 lines (533 loc) · 22.3 KB
/
SocialNetwork.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
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
import csv
import math
import threading
from os.path import exists
import networkx as nx
from sklearn.cluster import KMeans
from collections import Counter
from dateutil.parser import parse as dateparse
# =====================================================================================================================
#
# Authors: Halie Eckert, Gavin Hulvey, Sydney Zuelzke
# Date: 11/3/2021
# Project: Spatial-Social Networks
#
# Purpose:
# SocialNetwork.py is the class object for social networks.
#
# =====================================================================================================================
class SocialNetwork:
def __init__(self, name, relFile=None, locFile=None, keyFile=None, keyMapFile=None, userDataFile=None, poiFile=None, **kwargs):
self.__name = name
self.networkX = nx.MultiGraph()
self.__rel = {}
self.__loc = {}
self.__userData = {}
self.clusterItems = {}
self.__keywordMap = {}
self.__keywordMapReverse = {}
self.__keywords = {}
self.__keywordTime = {}
self.__userPoiTime = {}
self.__flattenedRelData = [[], []]
self.__flattenedLocData = [[], []]
self.__chunkedLocData = []
threads = [threading.Thread(target=lambda: self.loadRel(path=relFile)),
threading.Thread(target=lambda: self.loadLoc(path=locFile)),
threading.Thread(target=lambda: self.loadUser(path=userDataFile)),
threading.Thread(target=lambda: self.loadKey(kPath=keyFile, mPath=keyMapFile)),
threading.Thread(target=lambda: self.loadPoi(path=poiFile))]
for thread in threads:
thread.start()
for thread in threads:
thread.join()
self.IDByLoc = self.IDByLoc()
self.flattenRelData()
self.flattenLocData()
self.chunkLocData()
# Read in user attributes from the given path
# "USER ID" {
# "username": value,
# "name": value,
# "email": value,
# "birthdate": value,
# "phone": value
# }
def loadUser(self, path=None):
if path is not None and exists(path):
dict = {}
with open(path, 'r') as csvfile:
reader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(reader)
for row in reader:
user_id = row[0]
user_dict = {}
user_dict["username"] = row[1]
user_dict["name"] = row[2]
user_dict["email"] = row[3]
user_dict["birthdate"] = row[4]
user_dict["phone"] = row[5]
dict[user_id] = user_dict
self.__userData = dict
else:
self.__userData = None
def getUserAttributes(self, user_id):
return self.__userData[user_id]
# Reads rel file from path. This is super awful, but it's the fastest way to do things. This is what it returns:
# dict = {
# "user_id":
# [
# [rel_user_id, weight],
# [rel_user_id, weight]
# ]
# }
# noinspection PyShadowingBuiltins
def loadRel(self, path=None):
if path is not None and exists(path):
dict = {}
with open(path, 'r') as csvfile:
reader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(reader)
for row in reader:
user_id = row[0]
rel_user_id = row[1]
weight = row[2]
if user_id in dict:
dict[user_id] = dict[user_id] + [[rel_user_id, weight]]
else:
dict[user_id] = [[rel_user_id, weight]]
self.networkX.add_edge(float(user_id), float(rel_user_id), weight=1)
self.__rel = dict
else:
self.__rel = None
def getUserRel(self, user):
return self.__rel[user]
# Reads loc file from path.
# dict = {
# "user_id":
# [
# [lat_pos, lon_pos],
# [lat_pos, lon_pos]
# ]
# }
# noinspection SpellCheckingInspection,PyShadowingBuiltins
def loadLoc(self, path=None):
if path is not None and exists(path):
dict = {}
with open(path, 'r') as csvfile:
reader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(reader)
for row in reader:
user_id = row[0]
lat_pos = row[1]
lon_pos = row[2]
if user_id in dict:
dict[user_id] = dict[user_id] + [lat_pos, lon_pos]
else:
dict[user_id] = [[lat_pos, lon_pos]]
self.networkX.add_node(float(user_id))
self.__loc = dict
else:
self.__loc = None
# Reads keyword files from path.
# keywordMap = {
# "keyword_id": "keyword",
# "keyword_id": "keyword"
# }
#
# keywords = {
# "user_id": [keyword_id, keyword_id],
# "user_id": [keyword_id]
# }
# noinspection SpellCheckingInspection,PyShadowingBuiltins
def loadKey(self, kPath=None, mPath=None):
if kPath is not None and mPath is not None and exists(kPath) and exists(mPath):
keywords = {}
keywordsReverse = {}
userKeywords = {}
userKeywordsTime = {}
# Gets key map
with open(mPath, 'r') as csvfile:
reader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(reader)
for row in reader:
keyword_id = row[0]
keyword = row[1]
if keyword_id in keywords:
raise Exception(f"Error: Duplicate value in {mPath}")
else:
keywords[keyword_id] = keyword
keywordsReverse[keyword] = keyword_id
# Gets user keywords
with open(kPath, 'r') as kfile:
reader = csv.reader(kfile, delimiter=',', quotechar='|')
next(reader)
for row in reader:
user_id = str(float(row[0]))
keyword_id = row[1]
if user_id in list(userKeywords.keys()):
userKeywords[user_id].append(keyword_id)
userKeywordsTime[user_id].append([row[2], row[3]])
else:
userKeywords[user_id] = [keyword_id]
userKeywordsTime[user_id] = ([row[2], row[3]])
self.__keywordMap = keywords
self.__keywordMapReverse = keywordsReverse
self.__keywords = userKeywords
self.__keywordTime = userKeywordsTime
else:
self.__keywordMap = None
self.__keywordMapReverse = None
self.__keywords = None
self.__keywordTime = None
def loadPoi(self, path=None):
if path is not None and exists(path):
dict = {}
with open(path, 'r') as csvfile:
reader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(reader)
for row in reader:
user = str(row[0]) # Convert user to string
poi = str(row[1]) # Convert poi to string
time = row[2]
if user in dict:
if poi in dict[user]:
dict[user][poi].append(time)
else:
dict[user][poi] = [time]
else:
dict[user] = {poi: [time]} # Create new dictionary entry
self.__userPoiTime = dict
else:
self.__userPoiTime = None
# Parses the rel data into instantly plottable lists. For example, lat is [startLat, endLat, None, startLat...]
# This also chunks the data for faster processing and dedicates x number of threads to storing that data.
def flattenRelData(self):
if self.__rel is not None and self.__loc is not None:
threads = []
total = len(self.__rel)
threadCount = 10
start = 0
end = math.floor(total / threadCount)
for x in range(1, threadCount + 1):
if x == threadCount + 1:
end = total
threads.append(threading.Thread(target=lambda s=start, e=end, y=x: self.parseRelDataChunk(s, e * y)))
start = end * x
for thread in threads:
thread.start()
# for thread in threads:
# thread.join()
# Parses a single chunk of data and adds it to the master list
# noinspection SpellCheckingInspection
def parseRelDataChunk(self, start, end):
# TODO: Include weight
lat = []
lon = []
relList = list(self.__rel)
for x in range(start, end):
y = relList[x]
rels = self.__rel[y]
if y in self.__loc:
locs = self.__loc[y]
for z in locs:
startLat = float(z[0])
startLon = float(z[1])
for a in rels:
if a[0] in self.__loc:
for b in range(0, len(self.__loc[a[0]])):
endLat = float(self.__loc[a[0]][b][0])
endLon = float(self.__loc[a[0]][b][1])
lat = lat + [startLat, endLat]
lon = lon + [startLon, endLon]
self.__flattenedRelData[0] = self.__flattenedRelData[0] + lat
self.__flattenedRelData[1] = self.__flattenedRelData[1] + lon
# Parses the loc data into instantly plottable lists. For example, lat is [lat, lon, lat, lon...]
# This also chunks the data for faster processing and dedicates x number of threads to storing that data.
def flattenLocData(self):
if self.__loc is not None:
threads = []
total = len(self.__loc)
threadCount = 10
start = 0
end = math.floor(total / threadCount)
for x in range(1, threadCount + 1):
if x == threadCount + 1:
end = total
threads.append(
threading.Thread(target=lambda s=start, e=end, y=x: self.parseLocDataChunk(s, e * y)))
start = end * x
for thread in threads:
thread.start()
# Parses a single chunk of data and adds it to the master list
# noinspection SpellCheckingInspection
def parseLocDataChunk(self, start, end):
lat = []
lon = []
locList = list(self.__loc)
for x in range(start, end):
y = locList[x]
locs = self.__loc[y]
for z in range(0, len(locs)):
lat = lat + [float(locs[z][0])]
lon = lon + [float(locs[z][1])]
self.__flattenedLocData[0] = self.__flattenedLocData[0] + lat
self.__flattenedLocData[1] = self.__flattenedLocData[1] + lon
# Chunks coords from [[lat, lat, lat...],[lon, lon, lon...]] to [[lat, lon], [lat lon]...]
def chunkLocData(self):
if self.__loc is not None:
threads = []
total = len(self.__loc)
threadCount = 10
start = 0
end = math.floor(total / threadCount)
for x in range(1, threadCount + 1):
if x == threadCount + 1:
end = total
threads.append(
threading.Thread(target=lambda s=start, e=end, y=x: self.parseChunkLocChunk(s, e * y)))
start = end * x
for thread in threads:
thread.start()
# for thread in threads:
# thread.join()
# Parses a single chunk of data and adds it to the master list
# noinspection SpellCheckingInspection
def parseChunkLocChunk(self, start, end):
coords = []
locList = list(self.__loc)
for x in range(start, end):
y = locList[x]
locs = self.__loc[y]
for z in range(0, len(locs)):
coords = coords + [[float(locs[z][0]), float(locs[z][1])]]
self.__chunkedLocData = self.__chunkedLocData + coords
def getFlattenedLocData(self):
return self.__flattenedLocData
def getChunkedLocData(self):
return self.__chunkedLocData
def IDByLoc(self):
temp = {}
for id in self.__loc:
for loc in self.__loc[id]:
temp[f'{loc}'] = id
return temp
def getIDByLoc(self, lat, lon):
return self.IDByLoc[f"['{lat}', '{lon}']"]
# Returns all keywords
def getKeywords(self):
return list(self.__keywordMap.values())
def getIDByKeyword(self, keyword):
return str(self.__keywordMapReverse[keyword])
def getKeywordByID(self, id):
return str(self.__keywordMap[id])
def getUsersWithKeywords(self, keywords):
matches = []
for user in list(self.__loc.keys()):
if keywords:
match = True
for keyword in keywords:
if user not in self.__keywords or keyword not in self.__keywords[user]:
match = False
break
if not match:
continue
matches.append(user)
else:
if user not in self.__keywords:
matches.append(user)
return matches
def getUser(self, userID):
return [userID, self.__loc[userID]]
def getUserKeywords(self, userID):
if userID in self.__keywords:
return self.__keywords[userID]
else:
return []
def getUserPoi(self, userID):
return self.__userPoiTime[userID].keys()
def getUserPoiInTime(self, userID, start, end):
pois = []
for poi in self.__userPoiTime[userID]:
for time in self.__userPoiTime[userID][poi]:
if dateparse(time) >= dateparse(start) and dateparse(time) <= dateparse(end):
pois.append(poi)
continue
return pois
def getUserKeywordsInTime(self, userID, start, end):
keywordsTemp = self.getUserKeywords(userID)
keywords = []
for keyword in keywordsTemp:
if dateparse(self.__keywordTime[userID][self.__keywords[userID].index(keyword)][0]) >= dateparse(start) and dateparse(self.__keywordTime[userID][self.__keywords[userID].index(keyword)][0]) <= dateparse(end):
keywords.append(keyword)
return keywords
def getUserPoiTime(self, userID, poi):
return self.__userPoiTime[userID][poi]
def getUserKeywordsTime(self, userID, keyword):
return self.__keywordTime[userID][self.__keywords[userID].index(keyword)]
def getUserRel(self, user):
res = []
try:
res = self.__rel[user]
except KeyError:
res = []
return res
def getUsers(self):
return list(self.__loc.keys())
def userLoc(self, userID):
return self.__loc[userID]
def numberOfHops(self, start, end):
hops = 0
try:
hops = nx.dijkstra_path_length(self.networkX, float(start), float(end))
except (nx.NetworkXNoPath, nx.NodeNotFound) as e:
hops = -1
#return self.networkX.number_of_edges(float(start), float(end))
return hops
def shortestPath(self, start, end):
path = []
try:
path = nx.dijkstra_path(self.networkX, float(start), float(end))
except (nx.NetworkXNoPath, nx.NodeNotFound) as e:
path = []
return path
def commonRelations(self, target, users):
result = []
for user in users:
if self.networkX.has_edge(user,target):
result.append(user)
return result
# Visualize the data
def visualize(self, snInst=None, rnInst=None):
if snInst is not None:
snInst.plot(self.__flattenedRelData[0], self.__flattenedRelData[1], connect='pairs', pen=(50, 50, 200, 10),
brush=(50, 50, 200, 100))
if rnInst is not None:
rnInst.plot(self.__flattenedLocData[0], self.__flattenedLocData[1], pen=None, symbol='o', symbolSize=2,
symbolPen=(50, 50, 200, 25), symbolBrush=(50, 50, 200, 175))
# Returns size for cluster icons so that clusters that contain fewer nodes are smaller
@staticmethod
def sizeSort(refs):
sizes = []
refsSorted = refs.copy()
refsSorted.sort()
for x in refs:
sizes += [((refsSorted.index(x) + 1) * (75 / len(refsSorted)))]
return sizes
def getSummaryClusters(self, n):
n = int(n)
if n < 1:
n = 10
# n_clusters is th number of nodes to plot
kmeans = KMeans(n_clusters=int(n))
chunkedData = self.getChunkedLocData()
kmeans.fit(chunkedData)
# Scales the nodes according to population
centers = kmeans.cluster_centers_
# Get items in clusters and put it into dictionary {'clusterid': [userid, userid...], ...}
self.clusterItems = {}
for i in range(0, len(chunkedData)):
label = kmeans.labels_[i]
userid = self.getIDByLoc(chunkedData[i][0], chunkedData[i][1])
if label in self.clusterItems:
self.clusterItems[label].append(userid)
else:
self.clusterItems[label] = [userid]
clusterStart = list(self.clusterItems.keys())
popSize = []
for x in self.clusterItems:
if isinstance(self.clusterItems[x], list):
popSize.append(len(self.clusterItems[x]))
relations = [[], []]
while len(clusterStart) != 1:
start = clusterStart[0]
for item in clusterStart:
if clusterStart[0] is not item:
relations[0] += [centers[start][0], centers[item][0]]
relations[1] += [centers[start][1], centers[item][1]]
# for user in clusterItems[start]:
# for user2 in clusterItems[start]:
# print(f" {user}")
clusterStart.pop(0)
ref = list(Counter(kmeans.labels_).values())
sizes = self.sizeSort(ref)
ids = list(self.clusterItems.keys())
return ids, centers, sizes, relations, popSize
# Return the cluster id for a given user
def getUserCluster(self, user):
for x in self.clusterItems:
if user in self.clusterItems[x]:
return x
return -1
def getClusterUsers(self, cluster):
return self.clusterItems[cluster]
def usersCommonKeyword(self, queryUser, k=1):
commonUsers = []
commonDetails = {}
if queryUser is not None:
users = self.getUsers()
for user in users:
if user is not queryUser:
common = list(set(self.getUserKeywords(user)).intersection(
self.getUserKeywords(queryUser)))
if len(common) > (k-1):
commonDetails[user] = common
commonUsers.append(user)
return commonUsers, commonDetails
def usersCommonPoi(self, queryUser, k=1):
commonUsers = []
commonDetails = {}
if queryUser is not None:
users = self.getUsers()
for user in users:
if user is not queryUser:
common = list(set(self.getUserPoi(user)).intersection(
self.getUserPoi(queryUser)))
if len(common) > (k-1):
commonDetails[user] = common
commonUsers.append(user)
return commonUsers, commonDetails
def usersCommonPoiTime(self, queryUser, poi, start, end, k=1):
commonUsers = []
commonDetails = {}
if queryUser is not None:
users = self.getUsers()
for user in users:
if user is not queryUser:
common = list(set(self.getUserPoiTime(user, poi)).intersection(
self.getUserPoiTime(queryUser, poi)))
if len(common) > (k-1):
commonDetails[user] = common
commonUsers.append(user)
return commonUsers, commonDetails
def usersWithinHops(self, queryUser, users, h=0):
withinHops = []
hopsDetails = {}
for user in users:
hops = self.numberOfHops(queryUser, user)
if h == 0:
withinHops.append(user)
hopsDetails[user] = hops
else:
if hops <= h and hops != -1:
withinHops.append(user)
hopsDetails[user] = hops
return withinHops, hopsDetails
# Returns users within d distance
def usersWithinDistance(self, road, query, users, d=2):
withinDistance = []
distDetails = {}
common = self.userLoc(query)
commonLoc = road.findNearest(common)
for user in users:
query = self.userLoc(user)
queryLoc = road.findNearest(query)
dist = road.realUserDistance(queryLoc, commonLoc)
if dist <= d:
withinDistance.append(user)
distDetails[user] = dist
return withinDistance, distDetails
def userKeywordTime(self, user, keyword):
keywords = self.getUserKeywords(user)
i = keywords.index(str(keyword))
return self.__keywordTime[user][i]