-
Notifications
You must be signed in to change notification settings - Fork 0
/
Data_Ins.py
123 lines (99 loc) · 3.56 KB
/
Data_Ins.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
import time
import torch
import torch.multiprocessing as mp
from functools import partial
from utils import mymkdir
def pbc(x, L):
return x % L
def l2c(x, y, L):
x, y = pbc(x, L), pbc(y, L)
return x + y * L
def c2l(n, L):
assert n < L ** 2
return n % L, n // L
def Ham(k, L):
H = torch.zeros((L ** 2, L ** 2), dtype=torch.complex64)
for x in range(L):
for y in range(L):
n = l2c(x, y, L)
nx = l2c(x + 1, y, L)
ny = l2c(x, y + 1, L)
n1 = l2c(x + 1, y + 1, L)
n2 = l2c(x + 1, y - 1, L)
H[nx, n] = H[nx, n] + (-1) ** y
H[n, nx] = H[n, nx] + (-1) ** y
H[ny, n] = H[ny, n] + 1. + (-1) ** y * (1 - k)
H[n, ny] = H[n, ny] + 1. + (-1) ** y * (1 - k)
H[n1, n] = H[n1, n] + 1j * (-1) ** y * k / 2
H[n, n1] = H[n, n1] - 1j * (-1) ** y * k / 2
H[n2, n] = H[n2, n] - 1j * (-1) ** y * k / 2
H[n, n2] = H[n, n2] + 1j * (-1) ** y * k / 2
return H
def nnn_mask(L):
m = torch.eye(L ** 2, dtype=torch.complex64)
for x in range(L):
for y in range(L):
n = l2c(x, y, L)
nx = l2c(x + 1, y, L)
ny = l2c(x, y + 1, L)
n1 = l2c(x + 1, y + 1, L)
n2 = l2c(x + 1, y - 1, L)
m[nx, n] = m[nx, n] + 1.
m[n, nx] = m[n, nx] + 1.
m[ny, n] = m[ny, n] + 1.
m[n, ny] = m[n, ny] + 1.
m[n1, n] = m[n1, n] + 1.
m[n, n1] = m[n, n1] + 1.
m[n2, n] = m[n2, n] + 1.
m[n, n2] = m[n, n2] + 1.
return m
def genData(amount, L):
torch.manual_seed(int(time.time() * 1e16) % (2 ** 31 - 1))
Hs, labels = [], []
for i in range(int(amount / 2)):
H = Ham(0.1 + 0.4 * torch.rand(1), L)
Hs.append(H)
labels.append(0)
H = Ham(0.5 + 0.5 * torch.rand(1), L)
Hs.append(H)
labels.append(1)
Hs = torch.stack(Hs, dim=0) # (amount, L ** 2, L ** 2)
labels = torch.tensor(labels) # (amount,)
return Hs, labels
if __name__ == "__main__":
import os, warnings
warnings.filterwarnings('ignore')
os.environ['OMP_NUM_THREADS'] = '1'
os.environ['OPENBLAS_NUM_THREADS'] = '1'
os.environ['MKL_NUM_THREADS'] = '1'
os.environ['VECLIB_MAXIMUM_THREADS'] = '1'
os.environ['NUMEXPR_NUM_THREADS'] = '1'
torch.set_num_threads(1)
L = 12
TYPE = 'train'
amount, processors = 400, 50 # total_amount = amount * processors
# torch.manual_seed(0)
path = 'datasets/Ins_{}'.format(L)
mymkdir(path)
path = '{}/{}'.format(path, TYPE)
mymkdir(path)
t = time.time()
Hs, labels = [], []
mp.set_start_method('fork', force=True)
pool = mp.Pool(processes=processors)
res = pool.imap(partial(genData, L=L), amount * torch.ones(processors, dtype=torch.int32))
for (h, l) in res:
Hs.append(h)
labels.append(l)
pool.close()
pool.join()
Hs = torch.cat(Hs, dim=0).unsqueeze(1) # (amount * processors, 1, L ** 2, L ** 2)
labels = torch.cat(labels, dim=0) # (amount * processors,)
delta_t = time.time() - t
print(delta_t, '\n', L, Hs.shape, labels.shape)
f = open('{}/info.txt'.format(path), 'w')
f.write('time={}\nL={}\ndataset.shape={}\nlabels.shape={}'.format(delta_t, L, Hs.shape, labels.shape))
f.close()
# save
torch.save(Hs, '{}/dataset.pt'.format(path)) # (amount * processors, 1, L ** 2, L ** 2)
torch.save(labels.long(), '{}/labels.pt'.format(path)) # (amount * processors,)