-
Notifications
You must be signed in to change notification settings - Fork 2
/
Main.py
352 lines (303 loc) · 14.5 KB
/
Main.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
import torch as t
import Utils.TimeLogger as logger
from Utils.TimeLogger import log
from Params import args
from Model import Model, RandomMaskSubgraphs, LocalGraph
from DataHandler import DataHandler
import numpy as np
import pickle
from Utils.Utils import calcRegLoss, contrastLoss, pairPredict
import os
import torch
from layers import *
from view_learner import *
from torch_sparse import SparseTensor
from torch.utils.data.dataloader import default_collate
os.environ["CUDA_VISIBLE_DEVICES"] = '1'
class Coach:
def __init__(self, handler):
self.handler = handler
print('USER', args.user, 'ITEM', args.item)
print('NUM OF INTERACTIONS', self.handler.trnLoader.dataset.__len__())
self.metrics = dict()
mets = ['Loss', 'preLoss', 'Recall', 'NDCG']
for met in mets:
self.metrics['Train' + met] = list()
self.metrics['Test' + met] = list()
print()
self.graph_learner = GraphLearner(input_size=args.latdim, hidden_size=args.latdim,
graph_type=args.graph_type, top_k=args.top_k,
epsilon=args.epsilon, num_pers=args.num_per, metric_type=args.graph_metric_type,
feature_denoise=args.feature_denoise, device=args.gpu)
self.backbone = args.backbone
if self.backbone == "GCN":
self.backbone_gnn = myGCN(args, in_dim=args.latdim, out_dim=args.IB_size*2,
hidden_dim=args.latdim).cuda()
elif self.backbone == "GIN":
self.backbone_gnn = myGIN(args, in_dim=args.latdim, out_dim=args.IB_size*2,
hidden_dim=args.latdim).cuda()
elif self.backbone == "GAT":
self.backbone_gnn = myGAT(args, in_dim=args.latdim, out_dim=args.IB_size*2,
hidden_dim=args.latdim).cuda()
elif self.backbone == "mixhop":
self.backbone_gnn = MixHopNetwork(args, feature_number = args.latdim, class_number = args.IB_size*2)
self.view_learner = ViewLearner(self.backbone_gnn, mlp_edge_model_dim = args.latdim)
self.IB_size = args.IB_size
def makePrint(self, name, ep, reses, save):
ret = 'Epoch %d/%d, %s: ' % (ep, args.epoch, name)
for metric in reses:
val = reses[metric]
ret += '%s = %.4f, ' % (metric, val)
tem = name + metric
if save and tem in self.metrics:
self.metrics[tem].append(val)
ret = ret[:-2] + ' '
return ret
def run(self):
self.prepareModel()
log('Model Prepared')
if args.load_model != None:
self.loadModel()
stloc = len(self.metrics['TrainLoss']) * args.tstEpoch - (args.tstEpoch - 1)
else:
stloc = 0
log('Model Initialized')
import time
start_time = time.time()
for ep in range(stloc, args.epoch):
tstFlag = (ep % args.tstEpoch == 0)
reses = self.trainEpoch()
log(self.makePrint('Train', ep, reses, tstFlag))
if tstFlag:
reses = self.testEpoch()
log(self.makePrint('Test', ep, reses, tstFlag))
self.saveHistory()
print("------training time------:", (time.time()-start_time)/args.epoch)
reses = self.testEpoch()
log(self.makePrint('Test', args.epoch, reses, True))
self.saveHistory()
def learn_graph(self, node_features, edge_index):
# print("node_features:", node_features[:10])
# print("edge_index:", edge_index[:10])
new_feature, new_adj = self.graph_learner(node_features, edge_index)
return new_feature, new_adj
def prepareModel(self):
self.model = Model().cuda()
self.opt = t.optim.Adam(self.model.parameters(), lr=args.lr, weight_decay=0)
self.masker = RandomMaskSubgraphs()
self.sampler = LocalGraph()
def reparametrize_n(self, mu, std, n=1):
eps = Variable(std.data.new(std.size()).normal_())
return mu + eps * std
def transsparse(self, mat, edge_index, s):
# mat = self.handler.normalizeAdj(mat.cpu().detach().numpy())
# print("mat:", mat.size())
# println()
# idxs = edge_index[:,int(0.2*(edge_index.size()[1])):]
# vals = mat[int(0.2*(mat.size()[0])):]
idxs = edge_index
vals = mat
# print("&&&:", idxs[:,20000:20300], vals[20000:20300])
# println()
shape = torch.Size((s, s))
new_adj = torch.sparse.FloatTensor(idxs, vals, shape).cuda()
return new_adj
def sim_loss(self, x1, x2):
simi_loss = F.l1_loss(x1, x2)
simi_loss = torch.exp(1 - simi_loss)
return simi_loss
def calc_loss(self, x, x_aug, temperature=0.2, sym=True):
# x and x_aug shape -> Batch x proj_hidden_dim
batch_size,_ = x.size()
# print("batch_size:", batch_size)
# print("x:", x.size())
x_abs = x.norm(dim=1)
x_aug_abs = x_aug.norm(dim=1)
sim_matrix = torch.einsum('ik,jk->ij', x, x_aug) / torch.einsum('i,j->ij', x_abs, x_aug_abs)
sim_matrix = torch.exp(sim_matrix / temperature)
pos_sim = sim_matrix[range(batch_size), range(batch_size)]
if sym:
loss_0 = pos_sim / (sim_matrix.sum(dim=0) - pos_sim)
loss_1 = pos_sim / (sim_matrix.sum(dim=1) - pos_sim)
loss_0 = - torch.log(loss_0).mean()
loss_1 = - torch.log(loss_1).mean()
loss = (loss_0 + loss_1)/2.0
else:
loss_1 = pos_sim / (sim_matrix.sum(dim=1) - pos_sim)
loss_1 = - torch.log(loss_1).mean()
return loss_1
return loss
def trainEpoch(self):
trnLoader = self.handler.trnLoader
trnLoader.dataset.negSampling()
epLoss, epPreLoss = 0, 0
steps = trnLoader.dataset.__len__() // args.batch
for i, tem in enumerate(trnLoader):
# kl_loss = []
adjlist = {}
adjlist[0] = self.handler.torchBiAdj
# adjlist[0] = self.handler.allOneAdj
shape = self.handler.torchBiAdj.size()
o_edge_index, new_edge_attr = self.handler.torchBiAdj._indices(), self.handler.torchBiAdj._values()
# print(new_edge_index.size())
# println()
ofea = self.model.getEgoEmbeds(adjlist, 1)
# number = int(0.25*(o_edge_index.size()[1]))
number = int(100000)
rdmUsrs = t.randint(args.user, [number])#ancs
rdmItms1 = t.randint_like(rdmUsrs, args.item)
new_idxs = default_collate([rdmUsrs,rdmItms1])
new_vals = t.tensor([0.05]*number)
new_graphs_list = []
node_embs = []
for j in range(args.gen):
# new_feature, new_adj = self.learn_graph(node_features=ofea, edge_index = o_edge_index)
# new_adj = self.transsparse(new_adj, o_edge_index, ofea.size()[0])
# com_adj = add_new + new_adj
# # new_edge_index, new_edge_attr = new_adj._indices(), new_adj._values()
# new_edge_index, new_edge_attr = com_adj._indices(), com_adj._values()
# adjlist[1] = com_adj
# new_graph = Data(x=new_feature, edge_index=new_edge_index, edge_attr=new_edge_attr)
# new_graphs_list.append(new_graph)
# loader = DataLoader(new_graphs_list, batch_size=len(new_graphs_list))
# batch_data = next(iter(loader))
# node_embs, _ = self.backbone_gnn(batch_data.x, batch_data.edge_index)
# another generator to generate graph
add_new = t.sparse.FloatTensor(new_idxs, new_vals, shape).cuda()
ant_node, ant_adj = self.view_learner(ofea, o_edge_index, self.handler.torchBiAdj)
new_adjs = self.transsparse(ant_adj, o_edge_index, ofea.size()[0])
com_adj_ant = new_adjs + add_new
# com_adj_ant = new_adjs
new_edge_index, new_edge_attr = com_adj_ant._indices(), com_adj_ant._values()
adjlist[j+1] = com_adj_ant
node_embs.append(ant_node)
node_embs = t.mean(t.stack(node_embs, 0), dim=0)
# print("node_embs:", node_embs.size())
# pritnln()
mu = node_embs[:, :self.IB_size]
std = F.softplus(node_embs[:, self.IB_size:]-self.IB_size, beta=1)
num_sample = 2
new_node_embs = self.reparametrize_n(mu, std, num_sample)
# pos_emb = new_node_embs[::2 ]
# neg_emb = new_node_embs[1::2]
klloss = -0.5 * (1 + 2 * std.log() - mu.pow(2) - std.pow(2)).sum(1).mean().div(math.log(2))
# simloss = self.sim_loss(pos_emb, neg_emb)
ancs, poss, negs = tem
ancs = ancs.long().cuda()
poss = poss.long().cuda()
negs = negs.long().cuda()
# usrEmbeds, itmEmbeds, usrEmbeds1, itmEmbeds1, usrEmbeds2, itmEmbeds2 = self.model(self.handler.torchBiAdj, args.keepRate)
usrEmbeds, itmEmbeds, usrEmbeds1, itmEmbeds1, usrEmbeds2, itmEmbeds2 = self.model(adjlist, args.keepRate, 3)
ancEmbeds = usrEmbeds[ancs]
posEmbeds = itmEmbeds[poss]
negEmbeds = itmEmbeds[negs]
simloss_u1 = self.sim_loss(usrEmbeds1, usrEmbeds2)
simloss_e1 = self.sim_loss(itmEmbeds1, itmEmbeds2)
clLoss = (contrastLoss(usrEmbeds1, usrEmbeds2, ancs, args.temp) + contrastLoss(itmEmbeds1, itmEmbeds2, poss, args.temp)) * args.ssl_reg
scoreDiff = pairPredict(ancEmbeds, posEmbeds, negEmbeds)
bprLoss = - (scoreDiff.sigmoid()+ 1e-8).log().sum() / args.batch
regLoss = calcRegLoss(self.model) * args.reg
# loss = bprLoss + regLoss + clLoss + 0.00001*klloss + 60*simloss_u1 + 60*simloss_e1
# loss = bprLoss + regLoss + clLoss+ 0.00001*klloss
loss = bprLoss + regLoss + clLoss + 0.00001*klloss
epLoss += loss.item()
epPreLoss += bprLoss.item()
self.opt.zero_grad()
# loss.backward(retain_graph=True)
loss.backward()
if loss != loss:
raise Exception('NaN in loss, crack!')
pritnln()
# torch.nn.utils.clip_grad_norm_(self.model.parameters(), max_norm=10, norm_type=2)
# print("back done")
self.opt.step()
log('Step %d/%d: loss = %.3f, regLoss = %.3f ' % (i, steps, loss, regLoss), save=False, oneline=True)
ret = dict()
ret['Loss'] = epLoss / steps
ret['preLoss'] = epPreLoss / steps
return ret
def testEpoch(self):
tstLoader = self.handler.tstLoader
epRecall, epNdcg = [0] * 2
epRecall2, epNdcg2 = [0] * 2
i = 0
num = tstLoader.dataset.__len__()
steps = num // args.tstBat
for usr, trnMask in tstLoader:
i += 1
usr = usr.long().cuda()
trnMask = trnMask.cuda()
adjlist = {}
adjlist[0] = self.handler.torchBiAdj
usrEmbeds, itmEmbeds = self.model(adjlist, 1.0)
allPreds = t.mm(usrEmbeds[usr], t.transpose(itmEmbeds, 1, 0)) * (1 - trnMask) - trnMask * 1e8
if i == steps:
tmp = nn.Linear(256, 32).cuda()(allPreds.T.float()).float().detach().cpu().numpy()
print("prediction results dimension:", tmp.shape)
import pickle
file=open(r"../Models/cluster_ours.pickle","wb")
pickle.dump(tmp,file) #storing_list
file.close()
# println()
_, topLocs = t.topk(allPreds, args.topk)
_, topLocs2 = t.topk(allPreds, args.topk2)
recall, ndcg = self.calcRes(topLocs.cpu().numpy(), self.handler.tstLoader.dataset.tstLocs, usr)
recall2, ndcg2 = self.calcRes(topLocs2.cpu().numpy(), self.handler.tstLoader.dataset.tstLocs, usr)
epRecall += recall
epNdcg += ndcg
epRecall2 += recall2
epNdcg2 += ndcg2
# log('Steps %d/%d: recall = %.2f, ndcg = %.2f ' % (i, steps, recall, ndcg), save=False, oneline=True)
log('Steps %d/%d: recall = %.2f, ndcg = %.2f, recall40 = %.2f, ndcg40 = %.2f ' % (i, steps, recall, ndcg, recall2, ndcg2), save=False, oneline=True)
ret = dict()
ret['Recall'] = epRecall / num
ret['NDCG'] = epNdcg / num
ret['Recall2'] = epRecall2 / num
ret['NDCG2'] = epNdcg2 / num
return ret
def calcRes(self, topLocs, tstLocs, batIds):
assert topLocs.shape[0] == len(batIds)
allRecall = allNdcg = 0
recallBig = 0
ndcgBig =0
for i in range(len(batIds)):
temTopLocs = list(topLocs[i])
temTstLocs = tstLocs[batIds[i]]
tstNum = len(temTstLocs)
maxDcg = np.sum([np.reciprocal(np.log2(loc + 2)) for loc in range(min(tstNum, args.topk))])
recall = dcg = 0
for val in temTstLocs:
if val in temTopLocs:
recall += 1
dcg += np.reciprocal(np.log2(temTopLocs.index(val) + 2))
recall = recall / tstNum
ndcg = dcg / maxDcg
allRecall += recall
allNdcg += ndcg
return allRecall, allNdcg
def saveHistory(self):
if args.epoch == 0:
return
with open('../History/' + args.save_path + '.his', 'wb') as fs:
pickle.dump(self.metrics, fs)
content = {
'model': self.model,
}
t.save(content, '../Models/' + args.save_path + '.mod')
log('Model Saved: %s' % args.save_path)
def loadModel(self):
ckp = t.load('../Models/' + args.load_model + '.mod')
self.model = ckp['model']
self.opt = t.optim.Adam(self.model.parameters(), lr=args.lr, weight_decay=0)
with open('../History/' + args.load_model + '.his', 'rb') as fs:
self.metrics = pickle.load(fs)
log('Model Loaded')
if __name__ == '__main__':
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu
logger.saveDefault = True
log('Start')
handler = DataHandler()
handler.LoadData()
log('Load Data')
coach = Coach(handler)
coach.run()