-
Notifications
You must be signed in to change notification settings - Fork 0
/
tgbr.py
456 lines (386 loc) · 12.8 KB
/
tgbr.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
#!/usr/bin/env python
#-*-coding:utf-8 -*-
import anydbm
import re
import math
from myzodb import MyZODB, transaction
class TGBR:
MusicSet = {}
InvertedList = {}
#gview有向图模型
class gview:
def __init__(self,tagMat):
self.nodes = {}
self.edges = {}
for tags in tagMat:
for i in range(len(tags)):
if self.nodes.has_key(tags[i]):
self.nodes[tags[i]] += 1
else:
self.nodes[tags[i]] = 1
if i == len(tags)-1:continue
key = tags[i]+','+tags[i+1]
if self.edges.has_key(key):
self.edges[key] += 1
else:
self.edges[key] = 1
#合并有向图,用于聚类中心的计算
def merge(self,mainGraph,subGraph):
for i in subGraph.nodes:
if mainGraph.nodes.has_key(i):
mainGraph.nodes[i] += subGraph.nodes[i]
else:
mainGraph.nodes[i] = subGraph.nodes[i]
for j in subGraph.edges:
if mainGraph.edges.has_key(j):
mainGraph.edges[j] += subGraph.edges[j]
else:
mainGraph.edges[j] = subGraph.edges[j]
return mainGraph
def __init__(self):
self.uri2name = anydbm.open('E:\\sam_work\\pagerank\\pr\\uri2name.db','r')
self.yc = anydbm.open('E:\\sam_work\\pagerank\\pr\\yanchang.db','r')
self.id2name = anydbm.open('id2name.db','r')
#计算两个有向图的同构度
def CalSim(self,graph1,graph2):
#计算子图与父图的同构度
def Calfai(self,fview_nodes,fview_edges,gview):
nodes_fview = 0
for node in fview_nodes:
nodes_fview += float(gview.nodes[node])
if len(gview.nodes) == 0:nodes_gview = 1
else:nodes_gview = float(sum(gview.nodes.values()))
edges_fview = 0
for edge in fview_edges:
edges_fview += float(gview.edges[edge])
if len(gview.edges) == 0:edges_gview = 1
else:edges_gview = float(sum(gview.edges.values()))
nodes = float(len(fview_nodes))
edges = float(len(fview_edges))
outvalue = nodes*(nodes_fview/nodes_gview) + edges*(edges_fview/edges_gview)
return outvalue
#发现子图
def Findfview():
fview_nodes = []
fview_edges = []
for i in graph1.edges:
if graph2.edges.has_key(i):
fview_edges.append(i)
for j in graph1.nodes:
if graph2.nodes.has_key(j):
fview_nodes.append(j)
return fview_edges,fview_nodes
fview_edges,fview_nodes = Findfview()
outvalue1 = Calfai(self,fview_nodes,fview_edges,graph1)
outvalue2 = Calfai(self,fview_nodes,fview_edges,graph2)
outvalue = math.sqrt(outvalue1**2 + outvalue2**2)
return outvalue
#从MODB中读取聚类的数据
def loadingMODB(self,DBFilePath):
db = MyZODB(DBFilePath)
dbrootN = db.dbroot
for t in dbrootN['InvertedList']:
if t == '':continue
temphash = dbrootN['InvertedList'][t]
keys = sorted(temphash.iteritems(),key=lambda temphash:temphash[1],reverse=True)
c = 0
temphash2 = {}
for j in keys:
c += 1
if c > 1000:break
temphash2[j[0]] = j[1]
self.InvertedList[t] = temphash2
for sl in dbrootN['gviews']:
self.MusicSet[sl] = self.gview([])
for node in dbrootN['gviews'][sl]['nodes']:
self.MusicSet[sl].nodes[node] = dbrootN['gviews'][sl]['nodes'][node]
for edge in dbrootN['gviews'][sl]['edges']:
self.MusicSet[sl].edges[edge] = dbrootN['gviews'][sl]['edges'][edge]
transaction.commit()
db.close()
return self.MusicSet,self.InvertedList
#模型构建
def traning(self,db):
#读入歌单数据
def LoadingData(self,testnum):
self.gUserBase = {}
self.gMusicBase = {}
totalTags = {}
c = 0
for i in db:
c += 1
if c >= testnum:break
tempmat = re.compile(r'\|').split(db[i])
if len(tempmat[1]) == '':continue
author = tempmat[0]
tags = re.compile(',').split(tempmat[1])
if len(tags) == 0:continue
for t in tags:totalTags[t] = ''
#print ','.join(tags)
songs = re.compile(',').split(tempmat[2])
if author != 'null':
if self.gUserBase.has_key(author):
self.gUserBase[author].append(tags)
else:
self.gUserBase[author] = []
self.gUserBase[author].append(tags)
for song in songs:
if self.gMusicBase.has_key(song):
self.gMusicBase[song].append(tags)
else:
self.gMusicBase[song] = []
self.gMusicBase[song].append(tags)
print 'initial process completed! ',len(self.gUserBase),'users,',len(self.gMusicBase),'musics! ',len(totalTags),'tags'
#创建标签序列集合
def GetGraphSet(self):
self.gSetToUser = {}
self.gSetToMusic = {}
#gSetToUser
for i in self.gUserBase:
tagsMat = self.gUserBase[i]
self.gSetToUser[i] = self.gview(tagsMat)
for j in self.gMusicBase:
tagsMat = self.gMusicBase[j]
self.gSetToMusic[j] = self.gview(tagsMat)
#聚类
def Kmeans(self,dataset,d):
#随机初始分配
def RandomSep(self):
keys = dataset.keys()
Len = len(keys)/d
GraphGroup = []
for i in range(d):
subGraphGroup = []
startIndex = 0+i*Len
stopIndex = (i+1)*Len-1
if stopIndex > len(keys):
stopIndex = len(keys)
for j in keys[startIndex:stopIndex]:
subGraphGroup.append(dataset[j])
GraphGroup.append(subGraphGroup)
return GraphGroup
#计算聚类中心
def CalCentroidView(self,GraphGroup):
CentroidGroup = []
for subgroup in GraphGroup:
centGraph = self.gview([])
for graph in subgroup:
centGraph.merge(centGraph,graph)
CentroidGroup.append(centGraph)
for index in range(len(CentroidGroup)):
Centroid = CentroidGroup[index]
h = len(GraphGroup[index])
for node in Centroid.nodes:
Centroid.nodes[node] = math.ceil(float(Centroid.nodes[node])/h)
for edge in Centroid.edges:
Centroid.edges[edge] = math.ceil(float(Centroid.edges[edge])/h)
return CentroidGroup
#判断前后两次聚类是否有变化
def CompareMat(self,mat1,mat2,c):
judgeCount = 0
threshold = 1*len(mat1)/10
for index1 in range(len(mat1)):
tempjudgeCount = 0
for index2 in range(len(mat1[index1])):
if not mat1[index1][index2] in mat2[index1]:
tempjudgeCount += 1
if tempjudgeCount > len(mat1[index1])/4:
judgeCount += 1
break
#for i in range(len(mat2)):
# print 'Group:',i,' len:',len(mat2[i])
if judgeCount > threshold:
print 'iteration:',c,' difference:',judgeCount,'/',threshold,'/',len(mat1)
return False
elif judgeCount :
print 'iteration:',c,' difference:',judgeCount,'/',threshold,'/',len(mat1)
return True
#存储聚类最终结果
def saveKmeans(self,group,centr):
db = MyZODB('Data.fs')
dbrootN = db.dbroot
#temphash = {}
for index in range(len(centr)):
dbrootN[index] = {}
dbrootN[index]['Centroid'] = {}
dbrootN[index]['Centroid']['nodes'] = {}
dbrootN[index]['Centroid']['edges'] = {}
for node in centr[index].nodes:
dbrootN[index]['Centroid']['nodes'][node] = centr[index].nodes[node]
for edge in centr[index].edges:
dbrootN[index]['Centroid']['edges'][edge] = centr[index].edges[edge]
dbrootN[index]['GraphGroup'] = {}
for gviewIndex in range(len(group[index])):
dbrootN[index]['GraphGroup'][gviewIndex] = {}
dbrootN[index]['GraphGroup'][gviewIndex]['nodes'] = {}
dbrootN[index]['GraphGroup'][gviewIndex]['edges'] = {}
for node in group[index][gviewIndex].nodes:
dbrootN[index]['GraphGroup'][gviewIndex]['nodes'][node] = group[index][gviewIndex].nodes[node]
for edge in group[index][gviewIndex].edges:
dbrootN[index]['GraphGroup'][gviewIndex]['edges'][edge] = group[index][gviewIndex].edges[edge]
#dbrootN = temphash
#print dbrootN
transaction.commit()
db.close()
#循环聚类
def iteration(self):
GraphGroup = RandomSep(self)
CentroidGroup = CalCentroidView(self,GraphGroup)
c = 0
while True:
c += 1
if c > 100:
saveKmeans(self,GraphGroup_new,CentroidGroup)
return GraphGroup,CentroidGroup
GraphGroup_new = []
for i in range(len(GraphGroup)):
GraphGroup_new.append([])
for groupIndex in range(len(GraphGroup)):
subgroup = GraphGroup[groupIndex]
for gviewIndex in range(len(subgroup)):
similarity = []
for Centroid in CentroidGroup:
simivalue = self.CalSim(subgroup[gviewIndex],Centroid)
similarity.append(simivalue)
maxValueIndex = similarity.index(max(similarity))
GraphGroup_new[maxValueIndex].append(subgroup[gviewIndex])
#print subgroup[gviewIndex],' result:',groupIndex,'->',maxValueIndex
bl = CompareMat(self,GraphGroup,GraphGroup_new,c)
if bl:
saveKmeans(self,GraphGroup_new,CentroidGroup)
return GraphGroup_new,CentroidGroup
else:
CentroidGroup = CalCentroidView(self,GraphGroup_new)
GraphGroup = GraphGroup_new
continue
self.GraphGroup,self.CentroidGroup = iteration(self)
#建立倒排表
def CreatInvertedList(self):
'''
def saveData(self):
db = MyZODB('Data.fs')
dbrootN = db.dbroot
dbrootN['gviews'] = {}
dbrootN['InvertedList'] = {}
for g in self.gSetToMusic:
dbrootN['gviews'][g] = {}
dbrootN['gviews'][g]['nodes'] = {}
dbrootN['gviews'][g]['edges'] = {}
for node in self.gSetToMusic[g].nodes:
dbrootN['gviews'][g]['nodes'][node] = self.gSetToMusic[g].nodes[node]
for edge in self.gSetToMusic[g].edges:
dbrootN['gviews'][g]['edges'][edge] = self.gSetToMusic[g].edges[edge]
for t in self.InvertedList:
dbrootN['InvertedList'][t] = self.InvertedList[t]
transaction.commit()
db.close()
'''
def saveData(self):
db = MyZODB('Data2.fs')
dbrootN = db.dbroot
dbrootN['gviews'] = {}
dbrootN['InvertedList'] = {}
for g in self.gSetToMusic:
dbrootN['gviews'][g] = self.gSetToMusic[g]
for t in self.InvertedList:
dbrootN['InvertedList'][t] = self.InvertedList[t]
transaction.commit()
db.close()
for sl in self.gSetToMusic:
for node in self.gSetToMusic[sl].nodes:
if self.InvertedList.has_key(node):
temphash = self.InvertedList[node]
temphash[sl] = self.gSetToMusic[sl].nodes[node]
self.InvertedList[node] = temphash
else:
temphash = {}
temphash[sl] = self.gSetToMusic[sl].nodes[node]
self.InvertedList[node] = temphash
saveData(self)
LoadingData(self,33333)
GetGraphSet(self)
#Kmeans(self,self.gSetToMusic,1000)
CreatInvertedList(self)
#查询
def search(self,tags,SongNum,MusicSet,InvertedList):
def hashInCommon(hash1,hash2):
temphash = {}
if len(hash1) == 0:return hash2
if len(hash2) == 0:return hash1
if len(hash1) > len(hash2):
for i in hash2:
if hash1.has_key(i):
temphash[i] = ''
else:
for i in hash1:
if hash2.has_key(i):
temphash[i] = ''
return temphash
def parse(key):
try:
tempstr = self.yc[key]
except:
return ' ',' ',' '
s = re.compile(r'<song:(.*?)>').findall(tempstr)
al = re.compile(r'<album:(.*?)>').findall(tempstr)
ar = re.compile(r'<artist:(.*?)>').findall(tempstr)
try:
tempmat = []
for i in s:
tempmat.append(self.uri2name[i])
song = '/'.join(tempmat)
except:
song = ' '
try:
#if self.querry in al:
# album = self.querry
#else:
tempmat = []
for i in al:
tempmat.append(self.uri2name[i])
#album = '/'.join(tempmat)
album = tempmat[0]
except:
album = ' '
try:
#if self.querry in ar:
# artist = self.querry
#else:
tempmat = []
for i in ar:
tempmat.append(self.uri2name[i])
artist = tempmat[0]
except:
artist = ' '
return song,album,artist
myhash = {}
tag2 = []
for i in tags:
i = i.decode('utf-8').encode('gbk')
if InvertedList.has_key(i):
tag2.append(i)
myhash = hashInCommon(myhash,InvertedList[i])
print len(myhash)
print 'tags:',','.join(tag2)
Usergview = self.gview([tag2])
scorehash = {}
for i in myhash:
scorehash[i] = self.CalSim(Usergview,MusicSet[i])
c = 0
keys = sorted(scorehash.iteritems(),key=lambda scorehash:scorehash[1],reverse=True)
for i in keys:
if i[0] == '':continue
c += 1
if c > SongNum:break
#song,album,artist = parse(i[0])
song = self.id2name[i[0]]
print c,':',song,i[1]
#print ','.join([tag+str(MusicSet[i[0]].nodes[tag]) for tag in MusicSet[i[0]].nodes])
#print '-------------------------------'
t = TGBR()
db = anydbm.open('songlistDB_baidu.db','r')
#t.traning(db)
MusicSet,InvertedList = t.loadingMODB('Data.fs')
#print InvertedList
tags = ['感动','夜晚','怀旧','经典',]
t.search(tags,10,MusicSet,InvertedList)