質問編集履歴
4
コードをすべて載せた
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -4,13 +4,25 @@
|
|
|
4
4
|
|
|
5
5
|
python module Rayを用いた関数の並列化処理を行いたいのですが、”TypeError: can't pickle _thread.lock objects”とエラーが出てしまいます。
|
|
6
6
|
|
|
7
|
+
エラーの原因はメソッドが直列化できないからであることはわかりました。
|
|
8
|
+
|
|
7
|
-
|
|
9
|
+
インスタンス変数を総て外に出して引数にしたクラスメソッドを作成し、そのクラスメソッドを並列化処理すればうまくいくことはわかったのですが、プログラムが非常に膨大であるため、関数構造を変えられません。
|
|
8
|
-
|
|
10
|
+
|
|
9
|
-
|
|
11
|
+
また、cloudpickleやpickleを使うとうまくいく？ような記述をwebで見つけたのですが、そもそもpickleがよくわからないため、うまくコードに起こせない状態です。
|
|
10
|
-
|
|
12
|
+
|
|
11
|
-
|
|
13
|
+
multiprocessingモジュールを使えば動いたので、並列処理できないプログラムでないことは確かなのですが、、、
|
|
12
|
-
|
|
14
|
+
|
|
15
|
+
|
|
16
|
+
|
|
13
|
-
|
|
17
|
+
もし何か解決策等あれば、ご教授よろしくおねがいします。
|
|
18
|
+
|
|
19
|
+
|
|
20
|
+
|
|
21
|
+
参考にしたwebページ
|
|
22
|
+
|
|
23
|
+
https://docs.ray.io/en/releases-0.8.5/serialization.html
|
|
24
|
+
|
|
25
|
+
https://docs.ray.io/en/master/serialization.html
|
|
14
26
|
|
|
15
27
|
|
|
16
28
|
|
|
@@ -20,7 +32,97 @@
|
|
|
20
32
|
|
|
21
33
|
```
|
|
22
34
|
|
|
23
|
-
|
|
35
|
+
TypeError: can't pickle _thread.lock objects
|
|
36
|
+
|
|
37
|
+
---------------------------------------------------------------------------
|
|
38
|
+
|
|
39
|
+
TypeError Traceback (most recent call last)
|
|
40
|
+
|
|
41
|
+
<ipython-input-13-9de0171f325e> in <module>
|
|
42
|
+
|
|
43
|
+
41
|
|
44
|
+
|
|
45
|
+
42
|
|
46
|
+
|
|
47
|
+
---> 43 results=[output.remote(EbNodB) for EbNodB in range(-3,5)]
|
|
48
|
+
|
|
49
|
+
44 print(ray.get(results))
|
|
50
|
+
|
|
51
|
+
45
|
|
52
|
+
|
|
53
|
+
|
|
54
|
+
|
|
55
|
+
<ipython-input-13-9de0171f325e> in <listcomp>(.0)
|
|
56
|
+
|
|
57
|
+
41
|
|
58
|
+
|
|
59
|
+
42
|
|
60
|
+
|
|
61
|
+
---> 43 results=[output.remote(EbNodB) for EbNodB in range(-3,5)]
|
|
62
|
+
|
|
63
|
+
44 print(ray.get(results))
|
|
64
|
+
|
|
65
|
+
45
|
|
66
|
+
|
|
67
|
+
|
|
68
|
+
|
|
69
|
+
~/.pyenv/versions/3.7.10/lib/python3.7/site-packages/ray/remote_function.py in _remote_proxy(*args, **kwargs)
|
|
70
|
+
|
|
71
|
+
97 @wraps(function)
|
|
72
|
+
|
|
73
|
+
98 def _remote_proxy(*args, **kwargs):
|
|
74
|
+
|
|
75
|
+
---> 99 return self._remote(args=args, kwargs=kwargs)
|
|
76
|
+
|
|
77
|
+
100
|
|
78
|
+
|
|
79
|
+
101 self.remote = _remote_proxy
|
|
80
|
+
|
|
81
|
+
|
|
82
|
+
|
|
83
|
+
~/.pyenv/versions/3.7.10/lib/python3.7/site-packages/ray/remote_function.py in _remote(self, args, kwargs, num_returns, num_cpus, num_gpus, memory, object_store_memory, accelerator_type, resources, max_retries, placement_group, placement_group_bundle_index, name)
|
|
84
|
+
|
|
85
|
+
175 # first driver. This is an argument for repickling the function,
|
|
86
|
+
|
|
87
|
+
176 # which we do here.
|
|
88
|
+
|
|
89
|
+
--> 177 self._pickled_function = pickle.dumps(self._function)
|
|
90
|
+
|
|
91
|
+
178
|
|
92
|
+
|
|
93
|
+
179 self._function_descriptor = PythonFunctionDescriptor.from_function(
|
|
94
|
+
|
|
95
|
+
|
|
96
|
+
|
|
97
|
+
~/.pyenv/versions/3.7.10/lib/python3.7/site-packages/ray/cloudpickle/cloudpickle_fast.py in dumps(obj, protocol, buffer_callback)
|
|
98
|
+
|
|
99
|
+
68 with io.BytesIO() as file:
|
|
100
|
+
|
|
101
|
+
69 cp = CloudPickler(file, protocol=protocol, buffer_callback=buffer_callback)
|
|
102
|
+
|
|
103
|
+
---> 70 cp.dump(obj)
|
|
104
|
+
|
|
105
|
+
71 return file.getvalue()
|
|
106
|
+
|
|
107
|
+
72
|
|
108
|
+
|
|
109
|
+
|
|
110
|
+
|
|
111
|
+
~/.pyenv/versions/3.7.10/lib/python3.7/site-packages/ray/cloudpickle/cloudpickle_fast.py in dump(self, obj)
|
|
112
|
+
|
|
113
|
+
654 def dump(self, obj):
|
|
114
|
+
|
|
115
|
+
655 try:
|
|
116
|
+
|
|
117
|
+
--> 656 return Pickler.dump(self, obj)
|
|
118
|
+
|
|
119
|
+
657 except RuntimeError as e:
|
|
120
|
+
|
|
121
|
+
658 if "recursion" in e.args[0]:
|
|
122
|
+
|
|
123
|
+
|
|
124
|
+
|
|
125
|
+
TypeError: can't pickle _thread.lock objects
|
|
24
126
|
|
|
25
127
|
```
|
|
26
128
|
|
|
@@ -32,93 +134,493 @@
|
|
|
32
134
|
|
|
33
135
|
```python
|
|
34
136
|
|
|
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
|
-
K=
|
|
76
|
-
|
|
77
|
-
|
|
78
|
-
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
|
|
82
|
-
|
|
83
|
-
|
|
84
|
-
|
|
85
|
-
|
|
86
|
-
|
|
87
|
-
|
|
88
|
-
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
|
|
92
|
-
|
|
93
|
-
|
|
94
|
-
|
|
95
|
-
|
|
137
|
+
#必要なライブラリ、定数
|
|
138
|
+
|
|
139
|
+
import sys
|
|
140
|
+
|
|
141
|
+
sys.path.append("../channel")
|
|
142
|
+
|
|
143
|
+
from AWGN import _AWGN
|
|
144
|
+
|
|
145
|
+
import numpy as np
|
|
146
|
+
|
|
147
|
+
import random
|
|
148
|
+
|
|
149
|
+
import time
|
|
150
|
+
|
|
151
|
+
import math
|
|
152
|
+
|
|
153
|
+
from decimal import *
|
|
154
|
+
|
|
155
|
+
import ray
|
|
156
|
+
|
|
157
|
+
random.seed(time.time())
|
|
158
|
+
|
|
159
|
+
|
|
160
|
+
|
|
161
|
+
ch=_AWGN()
|
|
162
|
+
|
|
163
|
+
|
|
164
|
+
|
|
165
|
+
class coding():
|
|
166
|
+
|
|
167
|
+
def __init__(self):
|
|
168
|
+
|
|
169
|
+
#super().__init__()
|
|
170
|
+
|
|
171
|
+
|
|
172
|
+
|
|
173
|
+
self.N=1024
|
|
174
|
+
|
|
175
|
+
self.R=0.5
|
|
176
|
+
|
|
177
|
+
self.K=math.floor(self.R*self.N)
|
|
178
|
+
|
|
179
|
+
self.design_SNR=4
|
|
180
|
+
|
|
181
|
+
|
|
182
|
+
|
|
183
|
+
#prepere constants
|
|
184
|
+
|
|
185
|
+
tmp2=np.log2(self.N)
|
|
186
|
+
|
|
187
|
+
self.itr_num=tmp2.astype(int)
|
|
188
|
+
|
|
189
|
+
self.frozen_bits,self.info_bits=\
|
|
190
|
+
|
|
191
|
+
self.Bhattacharyya_bounds()
|
|
192
|
+
|
|
193
|
+
|
|
194
|
+
|
|
195
|
+
self.Gres=self.make_H()
|
|
196
|
+
|
|
197
|
+
|
|
198
|
+
|
|
199
|
+
self.filename="polar_code_{}_{}".format(self.N,self.K)
|
|
200
|
+
|
|
201
|
+
|
|
202
|
+
|
|
203
|
+
class coding(coding):
|
|
204
|
+
|
|
205
|
+
#frozen_bitの選択
|
|
206
|
+
|
|
207
|
+
def Bhattacharyya_bounds(self):
|
|
208
|
+
|
|
209
|
+
E=np.zeros(1,dtype=np.float128)
|
|
210
|
+
|
|
211
|
+
E =Decimal('10') ** (Decimal(str(self.design_SNR)) / Decimal('10'))
|
|
212
|
+
|
|
213
|
+
itr_num=np.log2(N)
|
|
214
|
+
|
|
215
|
+
itr_num=itr_num.astype(int)
|
|
216
|
+
|
|
217
|
+
z=np.zeros(self.N,dtype=np.float128)
|
|
218
|
+
|
|
219
|
+
|
|
220
|
+
|
|
221
|
+
#10^10かけて計算する
|
|
222
|
+
|
|
223
|
+
|
|
224
|
+
|
|
225
|
+
z[0]=math.exp(Decimal('-1')*Decimal(str(E)))
|
|
226
|
+
|
|
227
|
+
|
|
228
|
+
|
|
229
|
+
#print("E=",np.exp(-E))
|
|
230
|
+
|
|
231
|
+
|
|
232
|
+
|
|
233
|
+
for j in range(1,itr_num+1):
|
|
234
|
+
|
|
235
|
+
tmp=2**(j)//2
|
|
236
|
+
|
|
237
|
+
|
|
238
|
+
|
|
239
|
+
for t in range(tmp):
|
|
240
|
+
|
|
241
|
+
T=z[t]
|
|
242
|
+
|
|
243
|
+
z[t]=Decimal('2')*Decimal(str(T))-Decimal(str(T))**Decimal('2')
|
|
244
|
+
|
|
245
|
+
z[tmp+t]=Decimal(str(T))**Decimal('2')
|
|
246
|
+
|
|
247
|
+
#print(z)
|
|
248
|
+
|
|
249
|
+
#np.savetxt("z",z)
|
|
250
|
+
|
|
251
|
+
tmp=self.indices_of_elements(z,N)
|
|
252
|
+
|
|
253
|
+
frozen_bits=tmp[:self.N-self.K]
|
|
254
|
+
|
|
255
|
+
info_bits=tmp[self.N-self.K:]
|
|
256
|
+
|
|
257
|
+
return np.sort(frozen_bits),np.sort(info_bits)
|
|
258
|
+
|
|
259
|
+
|
|
260
|
+
|
|
261
|
+
@staticmethod
|
|
262
|
+
|
|
263
|
+
def indices_of_elements(v,l):
|
|
264
|
+
|
|
265
|
+
tmp=np.argsort(v)[::-1]
|
|
266
|
+
|
|
267
|
+
#print(tmp)
|
|
268
|
+
|
|
269
|
+
res=tmp[0:l]
|
|
270
|
+
|
|
271
|
+
return res
|
|
272
|
+
|
|
273
|
+
|
|
274
|
+
|
|
275
|
+
class coding(coding):
|
|
276
|
+
|
|
277
|
+
@staticmethod
|
|
278
|
+
|
|
279
|
+
def tensordot(A):
|
|
280
|
+
|
|
281
|
+
tmp0=np.zeros((A.shape[0],A.shape[1]),dtype=np.int)
|
|
282
|
+
|
|
283
|
+
tmp1=np.append(A,tmp0,axis=1)
|
|
284
|
+
|
|
285
|
+
#print(tmp1)
|
|
286
|
+
|
|
287
|
+
tmp2=np.append(A,A,axis=1)
|
|
288
|
+
|
|
289
|
+
#print(tmp2)
|
|
290
|
+
|
|
291
|
+
tmp3=np.append(tmp1,tmp2,axis=0)
|
|
292
|
+
|
|
293
|
+
#print(tmp3)
|
|
294
|
+
|
|
295
|
+
return tmp3
|
|
296
|
+
|
|
297
|
+
|
|
298
|
+
|
|
299
|
+
def make_H(self):
|
|
300
|
+
|
|
301
|
+
G2=np.array([[1,0],[1,1]],dtype=np.int)
|
|
302
|
+
|
|
303
|
+
Gres=G2
|
|
304
|
+
|
|
305
|
+
for _ in range(self.itr_num-1):
|
|
306
|
+
|
|
307
|
+
#print(i)
|
|
308
|
+
|
|
309
|
+
Gres=self.tensordot(Gres)
|
|
310
|
+
|
|
311
|
+
return Gres
|
|
312
|
+
|
|
313
|
+
|
|
314
|
+
|
|
315
|
+
class encoding(coding):
|
|
316
|
+
|
|
317
|
+
#def __init__(self,N):
|
|
318
|
+
|
|
319
|
+
#super().__init__(N)
|
|
320
|
+
|
|
321
|
+
|
|
322
|
+
|
|
323
|
+
def generate_information(self):
|
|
324
|
+
|
|
325
|
+
#generate information
|
|
326
|
+
|
|
327
|
+
information=np.random.randint(0,2,self.K)
|
|
328
|
+
|
|
329
|
+
return information
|
|
330
|
+
|
|
331
|
+
|
|
332
|
+
|
|
333
|
+
class encoding(encoding):
|
|
334
|
+
|
|
335
|
+
def generate_U(self,information):
|
|
336
|
+
|
|
337
|
+
u_message=np.zeros(self.N)
|
|
338
|
+
|
|
339
|
+
u_message[self.info_bits]=information
|
|
340
|
+
|
|
341
|
+
return u_message
|
|
342
|
+
|
|
343
|
+
|
|
344
|
+
|
|
345
|
+
lass encoding(encoding):
|
|
346
|
+
|
|
347
|
+
def polar_encode(self):
|
|
348
|
+
|
|
349
|
+
information=self.generate_information()
|
|
350
|
+
|
|
351
|
+
u_message=self.generate_U(information)
|
|
352
|
+
|
|
353
|
+
codeword=(u_message@self.Gres)%2
|
|
354
|
+
|
|
355
|
+
return information,codeword
|
|
356
|
+
|
|
357
|
+
|
|
358
|
+
|
|
359
|
+
class decoding(coding):
|
|
360
|
+
|
|
361
|
+
n=0
|
|
362
|
+
|
|
363
|
+
EST_information=np.array([])
|
|
364
|
+
|
|
365
|
+
|
|
366
|
+
|
|
367
|
+
#def __init__(self,N):
|
|
368
|
+
|
|
369
|
+
#super().__init__(N)
|
|
370
|
+
|
|
371
|
+
|
|
372
|
+
|
|
373
|
+
@staticmethod
|
|
374
|
+
|
|
375
|
+
def chk(llr_1,llr_2):
|
|
376
|
+
|
|
377
|
+
CHECK_NODE_TANH_THRES=30
|
|
378
|
+
|
|
379
|
+
res=np.zeros(len(llr_1))
|
|
380
|
+
|
|
381
|
+
for i in range(len(res)):
|
|
382
|
+
|
|
383
|
+
|
|
384
|
+
|
|
385
|
+
if abs(llr_1[i]) > CHECK_NODE_TANH_THRES and abs(llr_2[i]) > CHECK_NODE_TANH_THRES:
|
|
386
|
+
|
|
387
|
+
if llr_1[i] * llr_2[i] > 0:
|
|
388
|
+
|
|
389
|
+
# If both LLRs are of one sign, we return the minimum of their absolute values.
|
|
390
|
+
|
|
391
|
+
res[i]=min(abs(llr_1[i]), abs(llr_2[i]))
|
|
392
|
+
|
|
393
|
+
else:
|
|
394
|
+
|
|
395
|
+
# Otherwise, we return an opposite to the minimum of their absolute values.
|
|
396
|
+
|
|
397
|
+
res[i]=-1 * min(abs(llr_1[i]), abs(llr_2[i]))
|
|
398
|
+
|
|
399
|
+
else:
|
|
400
|
+
|
|
401
|
+
res[i]= 2 * np.arctanh(np.tanh(llr_1[i] / 2, ) * np.tanh(llr_2[i] / 2))
|
|
402
|
+
|
|
403
|
+
return res
|
|
404
|
+
|
|
405
|
+
|
|
406
|
+
|
|
407
|
+
def SC_decoding(self,a):
|
|
408
|
+
|
|
409
|
+
#interior node operation
|
|
410
|
+
|
|
411
|
+
if a.shape[0]==1:
|
|
412
|
+
|
|
413
|
+
#frozen_bit or not
|
|
414
|
+
|
|
415
|
+
if np.any(self.frozen_bits==decoding.n):
|
|
416
|
+
|
|
417
|
+
tmp0=np.zeros(1)
|
|
418
|
+
|
|
419
|
+
elif a>=0:
|
|
420
|
+
|
|
421
|
+
tmp0=np.zeros(1)
|
|
422
|
+
|
|
423
|
+
elif a<0:
|
|
424
|
+
|
|
425
|
+
tmp0=np.ones(1)
|
|
426
|
+
|
|
427
|
+
else:
|
|
428
|
+
|
|
429
|
+
print("err!")
|
|
430
|
+
|
|
431
|
+
exit()
|
|
432
|
+
|
|
433
|
+
|
|
434
|
+
|
|
435
|
+
if np.any(self.info_bits==decoding.n):
|
|
436
|
+
|
|
437
|
+
decoding.EST_information=np.append(decoding.EST_information,a)
|
|
438
|
+
|
|
439
|
+
#print(decoding.n)
|
|
440
|
+
|
|
441
|
+
#print(t)
|
|
442
|
+
|
|
443
|
+
decoding.n+=1
|
|
444
|
+
|
|
445
|
+
#if t>=N:
|
|
446
|
+
|
|
447
|
+
#exit()
|
|
448
|
+
|
|
449
|
+
return tmp0
|
|
450
|
+
|
|
451
|
+
|
|
452
|
+
|
|
453
|
+
#step1 left input a output u1_hat
|
|
454
|
+
|
|
455
|
+
|
|
456
|
+
|
|
457
|
+
tmp1=np.split(a,2)
|
|
458
|
+
|
|
459
|
+
f_half_a=self.chk(tmp1[0],tmp1[1])
|
|
460
|
+
|
|
461
|
+
u1=self.SC_decoding(f_half_a)
|
|
462
|
+
|
|
463
|
+
|
|
464
|
+
|
|
465
|
+
#step2 right input a,u1_hat output u2_hat
|
|
466
|
+
|
|
467
|
+
tmp2=np.split(a,2)
|
|
468
|
+
|
|
469
|
+
g_half_a=tmp2[1]+(1-2*u1)*tmp2[0]
|
|
470
|
+
|
|
471
|
+
u2=self.SC_decoding(g_half_a)
|
|
96
472
|
|
|
97
473
|
|
|
98
474
|
|
|
99
|
-
|
|
100
|
-
|
|
101
|
-
re
|
|
102
|
-
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
|
|
106
|
-
|
|
107
|
-
|
|
108
|
-
|
|
109
|
-
|
|
110
|
-
|
|
111
|
-
|
|
112
|
-
|
|
113
|
-
|
|
114
|
-
|
|
115
|
-
o
|
|
116
|
-
|
|
117
|
-
|
|
118
|
-
|
|
119
|
-
|
|
120
|
-
|
|
121
|
-
p
|
|
475
|
+
#step3 up input u1,u2 output a_hat
|
|
476
|
+
|
|
477
|
+
res=np.concatenate([(u1+u2)%2,u2])
|
|
478
|
+
|
|
479
|
+
return res
|
|
480
|
+
|
|
481
|
+
|
|
482
|
+
|
|
483
|
+
class decoding(decoding):
|
|
484
|
+
|
|
485
|
+
def polar_decode(self,Lc):
|
|
486
|
+
|
|
487
|
+
#initialize class variable
|
|
488
|
+
|
|
489
|
+
decoding.n=0
|
|
490
|
+
|
|
491
|
+
decoding.EST_information=np.array([])
|
|
492
|
+
|
|
493
|
+
self.SC_decoding(Lc)
|
|
494
|
+
|
|
495
|
+
res=decoding.EST_information
|
|
496
|
+
|
|
497
|
+
res=-1*np.sign(res)
|
|
498
|
+
|
|
499
|
+
EST_information=(res+1)/2
|
|
500
|
+
|
|
501
|
+
|
|
502
|
+
|
|
503
|
+
return EST_information
|
|
504
|
+
|
|
505
|
+
|
|
506
|
+
|
|
507
|
+
class polar_code(encoding,decoding):
|
|
508
|
+
|
|
509
|
+
#def __init__(self,N):
|
|
510
|
+
|
|
511
|
+
#super().__init__(N)
|
|
512
|
+
|
|
513
|
+
|
|
514
|
+
|
|
515
|
+
def polar_code(self,EbNodB):
|
|
516
|
+
|
|
517
|
+
|
|
518
|
+
|
|
519
|
+
|
|
520
|
+
|
|
521
|
+
information,codeword=self.polar_encode()
|
|
522
|
+
|
|
523
|
+
|
|
524
|
+
|
|
525
|
+
Lc=-1*ch.generate_LLR(codeword,EbNodB)#デコーダが＋、ー逆になってしまうので-１をかける
|
|
526
|
+
|
|
527
|
+
|
|
528
|
+
|
|
529
|
+
EST_information=self.polar_decode(Lc)
|
|
530
|
+
|
|
531
|
+
|
|
532
|
+
|
|
533
|
+
return information,EST_information
|
|
534
|
+
|
|
535
|
+
|
|
536
|
+
|
|
537
|
+
if __name__=="__main__":
|
|
538
|
+
|
|
539
|
+
ray.init()
|
|
540
|
+
|
|
541
|
+
|
|
542
|
+
|
|
543
|
+
#N=512
|
|
544
|
+
|
|
545
|
+
#pc=polar_code()
|
|
546
|
+
|
|
547
|
+
|
|
548
|
+
|
|
549
|
+
#print(len(pc.info_bits))
|
|
550
|
+
|
|
551
|
+
#information,EST_information=pc.polar_code(100)
|
|
552
|
+
|
|
553
|
+
#print(len(information))
|
|
554
|
+
|
|
555
|
+
#print(len(EST_information))
|
|
556
|
+
|
|
557
|
+
#print(np.sum(information!=EST_information))
|
|
558
|
+
|
|
559
|
+
|
|
560
|
+
|
|
561
|
+
@ray.remote
|
|
562
|
+
|
|
563
|
+
def output(EbNodB):
|
|
564
|
+
|
|
565
|
+
count_err=0
|
|
566
|
+
|
|
567
|
+
count_all=0
|
|
568
|
+
|
|
569
|
+
count_berr=0
|
|
570
|
+
|
|
571
|
+
count_ball=0
|
|
572
|
+
|
|
573
|
+
MAX_ERR=8
|
|
574
|
+
|
|
575
|
+
|
|
576
|
+
|
|
577
|
+
while count_err<MAX_ERR:
|
|
578
|
+
|
|
579
|
+
|
|
580
|
+
|
|
581
|
+
pc=polar_code()
|
|
582
|
+
|
|
583
|
+
information,EST_information=pc.polar_code(EbNodB)
|
|
584
|
+
|
|
585
|
+
|
|
586
|
+
|
|
587
|
+
if np.any(information!=EST_information):#BLOCK error check
|
|
588
|
+
|
|
589
|
+
count_err+=1
|
|
590
|
+
|
|
591
|
+
|
|
592
|
+
|
|
593
|
+
count_all+=1
|
|
594
|
+
|
|
595
|
+
|
|
596
|
+
|
|
597
|
+
#calculate bit error rate
|
|
598
|
+
|
|
599
|
+
count_berr+=np.sum(information!=EST_information)
|
|
600
|
+
|
|
601
|
+
count_ball+=N
|
|
602
|
+
|
|
603
|
+
|
|
604
|
+
|
|
605
|
+
#print("\r","count_all=",count_all,",count_err=",count_err,"count_ball="\
|
|
606
|
+
|
|
607
|
+
#,count_ball,"count_berr=",count_berr,end="")
|
|
608
|
+
|
|
609
|
+
|
|
610
|
+
|
|
611
|
+
#print("\n")
|
|
612
|
+
|
|
613
|
+
#print("BER=",count_berr/count_ball)
|
|
614
|
+
|
|
615
|
+
return count_err,count_all,count_berr,count_all
|
|
616
|
+
|
|
617
|
+
|
|
618
|
+
|
|
619
|
+
|
|
620
|
+
|
|
621
|
+
results=[output.remote(EbNodB) for EbNodB in range(-3,5)]
|
|
622
|
+
|
|
623
|
+
print(ray.get(results))
|
|
122
624
|
|
|
123
625
|
```
|
|
124
626
|
|
|
@@ -144,4 +646,6 @@
|
|
|
144
646
|
|
|
145
647
|
ray 0.9.0
|
|
146
648
|
|
|
649
|
+
VScode のjupyter notebookにて記述
|
|
650
|
+
|
|
147
651
|
ここにより詳細な情報を記載してください。
|
3
誤字の修正
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -10,7 +10,7 @@
|
|
|
10
10
|
|
|
11
11
|
また、multiprocessing moduleを用いて記述したところ、うまくプログラムが動きました。
|
|
12
12
|
|
|
13
|
-
|
|
13
|
+
どこをどのように直せばよいのかわからないため質問いたしました。よろしくおねがいします。
|
|
14
14
|
|
|
15
15
|
|
|
16
16
|
|
2
追加の情報を入力
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -7,6 +7,8 @@
|
|
|
7
7
|
エラーの原因は関数内で別クラスのクラス内関数を引用しているからということはわかるのですが、なぜクラス内関数を引用するとエラーが出るのかがわかりません。
|
|
8
8
|
|
|
9
9
|
ちなみに、試しに(クラス内関数でなく)単なる関数を引用してコードを書いたところ、うまく動作しました。
|
|
10
|
+
|
|
11
|
+
また、multiprocessing moduleを用いて記述したところ、うまくプログラムが動きました。
|
|
10
12
|
|
|
11
13
|
そもそものクラス構造に問題がある気がするのですが、どのように直せばよいのかわからないため質問いたしました。よろしくおねがいします。
|
|
12
14
|
|
|
@@ -132,6 +134,8 @@
|
|
|
132
134
|
|
|
133
135
|
- この関数'output'をクラスに変更し、引用したいクラスを継承させても、同じエラーを吐いてしまう。
|
|
134
136
|
|
|
137
|
+
- multiprocessing module を用いたところ、うまく動いた
|
|
138
|
+
|
|
135
139
|
|
|
136
140
|
|
|
137
141
|
### 補足情報（FW/ツールのバージョンなど）
|
1
誤字の修正
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -102,7 +102,7 @@
|
|
|
102
102
|
|
|
103
103
|
#check
|
|
104
104
|
|
|
105
|
-
tc=turbo_code #クラスの呼び出し
|
|
105
|
+
tc=turbo_code() #クラスの呼び出し
|
|
106
106
|
|
|
107
107
|
|
|
108
108
|
|