-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrandomwalk.py
executable file
·124 lines (113 loc) · 3.72 KB
/
randomwalk.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
from numba import cuda, float64
import scipy.io as sci
import numpy as np
import time as tim
import sys
from numba.cuda.random import create_xoroshiro128p_states, xoroshiro128p_uniform_float64
@cuda.jit(device=True)
def get_node(p, cn, rng_state):
prob = p
currnode = cn
rng_states = rng_state
thread_id = cuda.grid(1)
r = cuda.random.xoroshiro128p_uniform_float64(rng_states,thread_id)
upto = 0
c=0
while c < 116:
#print(upto)
if (upto + prob[(currnode)*116+c]) >= r:
currnode = c
break
upto += prob[(currnode)*116+c]
c += 1
return currnode
@cuda.jit(device=True)
def loop(p, t, so, ta, stepp, ste, rng_state, reg):
source = so -1
targ = ta -1
regions = reg
steps = stepp
rng_states = rng_state
prob = p
time = t
delay = 0;
currnode = source
prevnode = -1;
step = 0
if source != targ:
while currnode != targ:
step += 1
currnode = get_node(prob,currnode, rng_states)
if prevnode != -1:
delay += time[(prevnode)*116+currnode]
if step>13456:
step = 13456
break
prevnode = currnode
ste[0]= delay
return step
@cuda.jit
def random_walk(prob, time, regions, steps, delay, ste, rng_states):
targ = cuda.threadIdx.x+1
source = cuda.blockIdx.x + 1
p = prob
t = time
r=regions
ids = (source-1)*116 + targ-1
c = cuda.local.array(shape=4,dtype=float64)
steps[ids] = loop(p, t, source,targ,steps, c, rng_states,r)
delay[ids] = c[0]
#print("source: ", source, " target: ", targ, " steps: ", steps[ids], " delay: ")
cuda.syncthreads()
start_time = tim.time()
cuda.config.ENABLE_CUDASIM = 1
name = sys.argv[1]
s = sci.loadmat(name+"_prob.mat")
s2 = sci.loadmat(name+"_time.mat")
prob = np.matrix(s[name+"_prob"], dtype=np.float64)
time = np.matrix(s2[name+"_time"], dtype=np.float64)
p = np.array(prob)
p = p.ravel()
t = np.array(time)
t = t.ravel()
regions = np.array(range(1,117), dtype=np.uint8)
cuda.select_device(0)
stream = cuda.stream()
steps = np.zeros((116*116,), dtype=np.uint64,)
ste = np.zeros((116*116,), dtype=np.float64,)
tt = np.zeros((116*116,), dtype=np.uint64,)
stream = cuda.stream()
g_prob = cuda.to_device(p,stream=stream)
g_time = cuda.to_device(t,stream=stream)
g_reg = cuda.to_device(regions,stream=stream)
g_ste = cuda.to_device(ste,stream=stream)
rng_states = create_xoroshiro128p_states(13456, seed=int(tim.time()))
g_steps = cuda.to_device(steps,stream=stream)
g_del = cuda.to_device(tt,stream=stream)
ranwalk = np.empty((10000,116,116), dtype=g_steps.dtype)
delay = np.empty((10000,116,116), dtype=g_steps.dtype)
output = np.empty(shape=g_steps.shape, dtype=g_steps.dtype)
delay_o = np.empty(shape=g_steps.shape, dtype=g_steps.dtype)
for i in range(10000):
random_walk[116, 116](g_prob, g_time, g_reg, g_steps, g_del, g_ste, rng_states)
print(i)
#print("g_steps size:", g_steps.size, " output size: ", output.size)
cuda.cudadrv.driver.Context.synchronize(cuda.current_context())
#time.sleep(0.7)
#print("synchronized")
output = g_steps.copy_to_host(stream=stream)
delay_o = g_del.copy_to_host(stream=stream)
#print("ovde zaglaviv")
sd = output
sc = delay_o
ranwalk[i,:,:] = sd.reshape((116,116))
delay[i,:,:] = sc.reshape((116,116))
del rng_states
#del g_steps
rng_states = create_xoroshiro128p_states(116*116, seed=np.uint64(tim.time()))
#g_steps = cuda.to_device(steps)
sci.savemat(name +'_randomwalk_steps.mat', {'ranwalk':ranwalk})
sci.savemat(name +'_randomwalk_time.mat', {'delay':delay})
cuda.close()
elapsed_time = tim.time() - start_time
print("elapsed time: ", elapsed_time)