-
Notifications
You must be signed in to change notification settings - Fork 1
/
QAOA_with_encoding.py
696 lines (512 loc) · 26.3 KB
/
QAOA_with_encoding.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
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
# -*- coding: utf-8 -*-
"""
Created on Thu Apr 14 17:44:00 2022
@author: burak
"""
# import pennylane as qml
import numpy as np
import networkx as nx
from scipy.linalg import expm
import time
import qiskit
from QAutoencoder.Utils import *
from qiskit.algorithms.optimizers import SPSA
from copy import deepcopy
# %% Problem Parameters
n = 3 # number of nodes in the graph
N = n**2 # number of qubits to represent the one-hot-encoding of TSP
G = nx.Graph()
for i in range(n):
G.add_node(i)
G.add_edge(0,1, weight = 50)
G.add_edge(0,2, weight = 2)
G.add_edge(1,2, weight = 2)
nx.draw(G, with_labels=True, alpha=0.8)
adj = nx.adjacency_matrix(G).todense()
edges = list(G.edges)
# %%
# 2 Backends, one(qasm) for simulating the circuit
# the other (unitary) is for debugging purposes
backend = qiskit.BasicAer.get_backend('qasm_simulator')
unitary_backend = qiskit.BasicAer.get_backend('unitary_simulator')
def QAOA(edges, adj, params = np.random.rand(6)*np.pi / 10, initial_state = None, decode = True, layers = 1):
'''
Building the QAOA Ansatz, that includes U_Hc and encoding scheme,
first e^(-i Hc beta) applies, which only shifts the phases, since the eigenvectors
are all in the computational basis state. Then the encoder maps those into the latent space.
After swapping the trash states with ancilla qubits, and decoding again, we acquire only the states
that obeys the constraints
Returns
-------
circ : qiskit circuit
'''
global n, N
'''
i.e when the input is 98, output is 001000000, the 1 being the 6th qubit!
the right most one is the 0th qubit, left most is the 8th qubit
graph3_dict = {'001010100': '000000000',
'001100010': '000000001',
'010001100': '000000010',
'010100001': '000000011',
'100001010': '000000100',
'100010001': '000000101'}
'''
ancillas = 6
c = qiskit.ClassicalRegister(3) # stores the measurements
if(decode == False):
c = qiskit.ClassicalRegister(3) # stores the measurements
q = qiskit.QuantumRegister(N+ ancillas)
circ = qiskit.QuantumCircuit(q,c)
beta = params[:layers]
gamma = params[layers:]
def initializePlusState():
# initial_state = np.ones((2**N)) / (np.sqrt(2) **N )
initial_state = np.zeros((2**N))
initial_state[266] = 1
circ.initialize(initial_state , [q[i] for i in range(N)])
# def initializer(initial_state):
# # b = np.ones((2**N)) / (np.sqrt(2) **N ) # equal superposition
# if(initial_state is None):
# # initial_state = np.zeros((2**N), dtype = np.complex128)
# # initial_state[266] = 1 # or the statevector for one of the feasible solutions
# initial_state = np.ones((2**N)) / (np.sqrt(2) **N ) # equal superposition
# circ.initialize(initial_state , [q[i] for i in range(N)])
# else:
# if(stateVectorCheck(initial_state) == False):
# raise('state vector is not valid')
# circ.initialize(initial_state , [q[i] for i in range(N)])
N_index = N - 1 #since qiskit has a different ordering (LSB for the first qubit), we subtract the index from this number
def costHamiltonian(l):
for edge in edges:
u, v = edge
for step in range(n - 1):
q_1 = (u * n) + step
q_2 = (v * n) + step + 1 # (+1 because we want that node visited in the next step)
q_1_rev = (u * n) + step + 1 # (reverse path from above)
q_2_rev = (v * n) + step
q_1 = (N - 1) - q_1
q_2 = (N - 1) - q_2
q_1_rev = (N - 1) - q_1_rev
q_2_rev = (N - 1) - q_2_rev
circ.rzz(2 * beta[l] * (adj[u,v]) , q_1 , q_2)
circ.rzz(2 * beta[l] * (adj[u,v]) , q_1_rev, q_2_rev) # exponential of the cost Hamiltonian
def mixerHamiltonian(l):
for i in range(N):
circ.rx( gamma[l] , i) # exponential of the mixer Hamiltonian
# Encoder Circuit
def encoderCircuit(n):
if(n == 2):
circ.cnot(3,0)
circ.x(3)
circ.cnot(3,2)
circ.cnot(3,1)
circ.x(3)
elif (n == 3):
circ.x( N_index - 0)
circ.x( N_index - 1)
circ.toffoli( N_index - 0, N_index - 1, N_index - 2)
circ.x( N_index - 0)
circ.x( N_index - 1)
circ.toffoli( N_index - 0, N_index -4, N_index - 2)
circ.toffoli( N_index - 1, N_index - 3, N_index - 2)
circ.toffoli( N_index - 3, N_index -7, N_index - 2)
circ.toffoli( N_index - 0, N_index - 2, N_index -4)
circ.toffoli( N_index - 0, N_index - 2, N_index -8)
circ.toffoli( N_index - 1, N_index - 2, N_index - 3)
circ.toffoli( N_index - 1, N_index - 2, N_index -8)
# Applying C_inv_toffoli of apply_CinvToffoli(bitstring, 2, 0, 1, 3)
circ.cnot(N_index - 2, N_index - 0)
circ.cnot(N_index - 2, N_index - 1)
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 3)
circ.cnot(N_index - 2, N_index - 1)
circ.cnot(N_index - 2, N_index - 0)
circ.cnot(N_index - 2, N_index - 0)
circ.cnot(N_index - 2, N_index - 1)
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 7)
circ.cnot(N_index - 2, N_index - 1)
circ.cnot(N_index - 2, N_index - 0)
circ.x(N_index - 2)
circ.toffoli( N_index - 0, N_index - 2, N_index -5 )
circ.toffoli( N_index - 0, N_index - 2, N_index -7 )
circ.toffoli( N_index - 1, N_index - 2, N_index -5 )
circ.toffoli( N_index - 1, N_index - 2, N_index -6 )
circ.cnot(N_index - 2, N_index - 0)
circ.cnot(N_index - 2, N_index - 1)
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 4)
circ.cnot(N_index - 2, N_index - 1)
circ.cnot(N_index - 2, N_index - 0)
circ.cnot(N_index - 2, N_index - 0)
circ.cnot(N_index - 2, N_index - 1)
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 6)
circ.cnot(N_index - 2, N_index - 1)
circ.cnot(N_index - 2, N_index - 0)
circ.x(N_index - 2)
# Swapping the Least Sig. Qubit with MSQ
for i in range(n):
circ.swap((N_index - 2)+ i, i)
elif(n == 4):
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2], N_index - 3)
circ.x(N_index - 2)
circ.x(N_index - 1)
circ.x(N_index - 0) # reset 3rd qubit, it will hold 1>2?
circ.toffoli(N_index - 4, N_index - 9, N_index - 3)
circ.toffoli(N_index - 4, N_index - 10, N_index - 3)
circ.toffoli(N_index - 4, N_index - 11, N_index - 3)
circ.toffoli(N_index - 5, N_index - 10, N_index - 3)
circ.toffoli(N_index - 5, N_index - 11, N_index - 3)
circ.toffoli(N_index - 6, N_index - 11, N_index - 3) # 3rd qubit= 1>2 ?
circ.x(N_index - 8)
circ.x(N_index - 9)
circ.x(N_index - 10)
circ.mcx([N_index - 8, N_index - 9, N_index - 10], N_index - 11)
circ.x(N_index - 8)
circ.x(N_index - 9)
circ.x(N_index - 10) # reset 11th qubit, it will hold 1>3 ?
circ.toffoli(N_index - 4, N_index - 13, N_index - 11)
circ.toffoli(N_index - 4, N_index - 14, N_index - 11)
circ.toffoli(N_index - 4, N_index - 15, N_index - 11)
circ.toffoli(N_index - 5, N_index - 14, N_index - 11)
circ.toffoli(N_index - 5, N_index - 15, N_index - 11)
circ.toffoli(N_index - 6, N_index - 15, N_index - 11) # 11th qubit= 1>3 ? (later to be swapped with 4th)
circ.x(N_index - 4)
circ.x(N_index - 6)
circ.x(N_index - 7)
circ.mcx([N_index - 4, N_index - 6, N_index - 7], N_index - 5)
circ.x(N_index - 4)
circ.x(N_index - 6)
circ.x(N_index - 7) # reset 5th qubit, it will hold 2>3 ?
circ.toffoli(N_index - 8, N_index - 13, N_index - 5)
circ.toffoli(N_index - 8, N_index - 14, N_index - 5)
circ.toffoli(N_index - 8, N_index - 15, N_index - 5)
circ.toffoli(N_index - 9, N_index - 14, N_index - 5)
circ.toffoli(N_index - 9, N_index - 15, N_index - 5)
circ.toffoli(N_index - 10, N_index - 15, N_index - 5) # 5th qubit= 2>3 ?
# encoding is complete, now we have to reset the rest of the qubits
# 111
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
# circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
#last step of 111
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 14)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
circ.x(N_index - 2)
circ.x(N_index - 1)
circ.x(N_index - 0)
# reverse 3 , 011
circ.x(N_index - 3)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 8)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 15)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 8)
# circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
#last step of 011
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 14)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 8)
# circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
circ.x(N_index - 2)
circ.x(N_index - 1)
circ.x(N_index - 0)
#reverse 11, 001
circ.x(N_index - 11)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 14)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 14)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 8)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 13)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 8)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
#last step of 001
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 13)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 8)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.x(N_index - 2)
circ.x(N_index - 1)
circ.x(N_index - 0)
#reverse 5 000
circ.x(N_index - 5)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 13)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 12)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 12)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
#last step of 000
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 12)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.x(N_index - 2)
circ.x(N_index - 1)
# unreverse 3, 100
circ.x(N_index - 3)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 13)
# circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 12)
# circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 12)
# circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
# circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
#last step of 100
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 12)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
# circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 6)
circ.x(N_index - 2)
circ.x(N_index - 1)
# unreverse 11, 110
circ.x(N_index - 11)
circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 14)
# circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
# circ.mcx([N_index - 0, N_index - 3, N_index - 11, N_index - 5], N_index - 7)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 14)
# circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 1, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 13)
# circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 9)
circ.mcx([N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
#last step of 110
circ.x(N_index - 0)
circ.x(N_index - 1)
circ.x(N_index - 2)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 13)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 10)
circ.mcx([N_index - 0, N_index - 1, N_index - 2, N_index - 3, N_index - 11, N_index - 5], N_index - 4)
circ.x(N_index - 2)
circ.x(N_index - 1)
# unreverse 5
circ.x(N_index - 5)
# 101 and 010 are not valid since 1>2 & 2>3 & 3>1 is not a valid logical assignment
circ.swap(N_index - 11,N_index - 4)
def discardQubits(n = 3):
if(n == 2):
circ.reset(2)
circ.reset(1)
circ.reset(0)
if(n==3):
for i in range(ancillas):
circ.swap(N + i, N_index - i)
def decoderCircuit(n = 3):
if(n == 2):
circ.x(3).inverse
circ.cnot(3,2).inverse
circ.cnot(3,1).inverse
circ.x(3).inverse
circ.cnot(3,0).inverse
if(n == 3):
for i in range(n):
circ.swap((N_index - 2)+ i, i).inverse()
circ.x(N_index - 2).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 6).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 4).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.toffoli(N_index - 1, N_index - 2, N_index - 6 ).inverse
circ.toffoli(N_index - 1, N_index - 2, N_index - 5 ).inverse
circ.toffoli(N_index - 0, N_index - 2, N_index - 7 ).inverse
circ.toffoli(N_index - 0, N_index - 2, N_index - 5 ).inverse
circ.x(N_index - 2).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 7).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.mcx([N_index - 2, N_index - 0, N_index - 1], N_index - 3).inverse
circ.cnot(N_index - 2, N_index - 1).inverse
circ.cnot(N_index - 2, N_index - 0).inverse
circ.toffoli(N_index - 1, N_index - 2, N_index - 8).inverse
circ.toffoli(N_index - 1, N_index - 2, N_index - 3).inverse
circ.toffoli(N_index - 0, N_index - 2, N_index - 8).inverse
circ.toffoli(N_index - 0, N_index - 2, N_index - 4).inverse
circ.toffoli( N_index - 3, N_index - 7, N_index - 2).inverse
circ.toffoli( N_index - 1, N_index - 3, N_index - 2).inverse
circ.toffoli( N_index - 0, N_index - 4, N_index - 2).inverse
circ.x(N_index - 1).inverse
circ.x(N_index - 0).inverse
circ.toffoli( N_index - 0, N_index - 1, N_index - 2 ).inverse
circ.x( N_index - 1 ).inverse
circ.x( N_index - 0 ).inverse
initializePlusState()
for l in range(layers):
costHamiltonian(l)
mixerHamiltonian(l)
encoderCircuit(n)
if(decode):
discardQubits(n)
# decoderCircuit(n)
if(decode == False):
circ.measure(q[:3],c)
else:
circ.measure(q[:3],c[:3])
return circ
# %% Loss function and other utils
def majorityVote(result):
# Get the majority vote as the output of the circuit
max_count = 0
out = None
for _ , (res, count )in enumerate(result.get_counts().items()):
if(res == '110000110' or res == '111110000'):
continue
if(count > max_count):
out = res
max_count = count
if(out == None):
raise('error in measurement')
return out
#one-hot-vectors and their embeddings
graph_dict = {'001010100': '000000000',
'001100010': '000000001',
'010001100': '000000010',
'010100001': '000000011',
'100001010': '000000100',
'100010001': '000000101'}
# false positives: '110000110' -> gets encoded as '110' and
# 111110000 -> gets encoded as '111'
iteration_count = 0
out = 0
def lossFunction(counts):
global cost_dict
global iteration_count
iteration_count += 1
loss = 0
count_total = 0
for idx, (keys, vals) in enumerate(counts.items()):
count_total += vals
loss += (cost_dict[keys] * vals)
loss /= count_total
if(iteration_count % 10 == 0):
print(loss)
return loss
def createPathCosts(graph3_dict, adj, n = 3):
cost = np.array([100,100,100,100,100,100,100,100])
cost_dict = {'111': 100,
'101': 52,
'001': 4,
'011': 52,
'100': 4,
'000': 52,
'010': 52,
'110': 100}
for idx, (keys, vals) in enumerate(graph3_dict.items()):
path = findPath(keys)
loss = adj[path[0], path[1]] + adj[path[1], path[2]]
cost[idx] = loss
return cost, cost_dict
cost, cost_dict = createPathCosts(graph_dict, adj, n = 3)
# %%
loss_history = []
running_loss_history = []
running_loss = 0
run_index= 0
def getExpVal(edges,adj):
backend = qiskit.Aer.get_backend('qasm_simulator')
backend.shots = 1024
def execute_circ(params):
global out
global loss_history
circuit = QAOA(edges, adj, params, decode = True)
counts = backend.run(circuit, seed_simulator = 10, shots = 1024).result()
out = deepcopy(counts)
loss_val = lossFunction(counts.get_counts())
loss_history.append(loss_val)
return loss_val
return execute_circ
init_params = np.random.rand(6)*np.pi / 10
exp = getExpVal(edges, adj)
print(exp(init_params))
# %% Optimization
start = time.time()
optimizer = SPSA(maxiter=1000)
optimized_parameters, final_cost, number_of_circuit_calls = optimizer.optimize(6, exp, initial_point = init_params )
end = time.time()
print('Optimization terminated within {:.3f} seconds'.format(end-start))
print('The final cost: {}'.format(exp(optimized_parameters)))
print(out)
# %%
circuit = QAOA(edges, adj, init_params, decode = True)
backend = qiskit.Aer.get_backend('qasm_simulator')
counts = backend.run(circuit, seed_simulator = 10, shots = 1024).result()
counts.get_counts()
exp = getExpVal(edges, adj)
exp(optimized_parameters)
running_loss= 0
for idx, l in enumerate(loss_history):
running_loss += l
running_loss_history.append(running_loss / (idx + 1))
from matplotlib import pyplot as plt
plt.plot(running_loss_history)
# %% Unit Tests
# sys.argv = ['Tests\\unitTest.py','n=3']
# execfile('Tests\\unitTest.py')
# %%
# for i in range(10):
# random_idx = np.random.randint(512)
# initial_state_out = np.zeros((2**N), dtype = np.complex128)
# initial_state_out[random_idx] = 1 # or the statevector for one of the feasible solutions
# start = time.time()
# circ = QAOA(edges, adj,initial_state = initial_state_out, decode = True)
# job = backend.run(qiskit.transpile(circ, backend))
# result = job.result()
# end = time.time()
# print('{}th Circuit run within {:.3f} seconds'.format(i ,end-start))
# # print('Expected Output: {}'.format(vals))
# out = majorityVote(result)
# print('Output: {}'.format(out))
# print('****')
# betas = np.random.rand(3)
# for idx, (keys, vals) in enumerate(graph_dict.items()):
# state = bit2int(keys)
# initial_state_out = np.zeros((2**N), dtype = np.complex128)
# initial_state_out[state] = 1 # or the statevector for one of the feasible solutions
# start = time.time()
# # circ = QAOA(edges, adj,initial_state = initial_state_out, decode = True)
# circ = QAOA(edges, adj, betas)
# job = backend.run(qiskit.transpile(circ, backend))
# result = job.result()
# end = time.time()
# print('{}th Circuit run within {:.3f} seconds'.format(idx ,end-start))
# print('Expected Output: {}'.format(vals))
# out = majorityVote(result)
# print('Output: {}'.format(out))
# print('****')