DistributedLP-GNN/test_DLPGNN.py at main · NetSysOpt/DistributedLP-GNN · GitHub

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
from model import *
import torch
import gzip
import pickle
import os
import random
import torch
import torch.optim as optim
from model import DLPGNN
import argparse
torch.manual_seed(0)


parser = argparse.ArgumentParser()
parser.add_argument('-size','--size',type=int,default=100)#radius
parser.add_argument('-L','--L',type=int,default=5)#radius
args = parser.parse_args()
#inference
size=args.size
L=args.L


if torch.cuda.is_available():
	device = "cuda"

elif torch.backends.mps.is_available():
	device = "mps"
else:
	device = "cpu"


def max_product(M,p):
	m,n=M.shape

	P=torch.tile(p.unsqueeze(1), (1, m)).to(device)
	PM=torch.mul(M.to(device),torch.t(P).to(device))
	q=torch.max(PM,dim=1).values

	return q


#dateset seting:


print('==> Building model..')


net=DLPGNN(L=L,K=16,M=size,N=size).to(device)
model_name=f"best_DLPGNN_size{size}_num100_L5.pth"
net.load_state_dict(torch.load(model_name,map_location=torch.device(device)))
print(model_name)
net.eval()

#optimizer
criterion = nn.MSELoss()
optimizer = optim.Adam(net.parameters(), lr=0.001)
optimizer.zero_grad()
criterion = nn.MSELoss()
def test(size,task):
	R_primal=[]
	R_dual=[]
	train_loss=0
	if task==1:# test set
		training_data_len=100
		pointer=500
	else: # train set
		training_data_len=100
		pointer=0

	for iter in range(training_data_len):
		iter=iter+pointer
		with open(f'./instance/size_{size}/LPinstance_{size}_{iter}.pkl', 'rb') as f:
			data_list = pickle.load(f)
		# print(f"{iter}------------------------")
		A=torch.tensor(data_list[0],dtype=torch.float32).to(device)
		pred_primal,pred_dual=net(A)
		primal=torch.tensor(data_list[1],dtype=torch.float32).to(device)
		dual=torch.tensor(data_list[2],dtype=torch.float32).to(device)
		x_dot=resortation_1(A,pred_primal)
		y_dot=resortation_2(A,pred_dual)
		rp=abs(torch.sum(x_dot)-torch.sum(primal))/torch.sum(primal)
		rd=abs(torch.sum(y_dot)-torch.sum(dual))/torch.sum(dual)

		R_primal.append(rp)
		R_dual.append(rd)

	R_primal=torch.mean(torch.tensor(R_primal))
	R_dual=torch.mean(torch.tensor(R_dual))
	if task==1:
		print(f"test RP:{R_primal},test RD:{R_dual}")
	else:
		print(f"testing RP:{R_primal},training RD:{R_dual}")
	return R_primal,R_dual

def resortation_1(A,x):
	M,N=A.shape
	ones=torch.ones(N).to(device)
	zeros=torch.zeros(M).to(device)

	x=torch.max(zeros,torch.min(ones,x))
	for i in range(M):
		term=torch.sum(A[i]*x)
		if term >=1:
			nx=torch.where(A[i]!=0)
			x[nx]=x[nx]/term

	return x

def resortation_2(A,y):
	M,N=A.shape
	ones=torch.ones(M).to(device)
	eps=1e-5

	y=torch.max(ones*eps,torch.min(ones,y))
	for j in range(N):
		term=torch.sum(torch.t(A)[j]*y)
		if term <=1:
			nx=torch.where(torch.t(A)[j]!=0)
			y[nx]=y[nx]/term
			# if term==0:
			# 	return None
	return y


print("testing--------------")
task=1
testdata=test(size,task)
#TrainSet

print("training--------------")
task=0
testdata=test(size,task)


# quit()
# X = np.array([i for i in range(T)])

# plt.plot(X,epoch_loss_list ,color="g",label='Training Loss')

# plt.xlabel('epoch')
# plt.ylabel('epoch loss')
# plt.legend()
# plt.show()