-
Notifications
You must be signed in to change notification settings - Fork 0
/
gatefinderclass.py
218 lines (160 loc) · 7.59 KB
/
gatefinderclass.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
import numpy as np
from numba import jit
from time import time
from qiskit.quantum_info import DensityMatrix,Statevector,random_statevector
from cirq import is_hermitian,validate_density_matrix,unitary
from dmcx import testfinderhos,tracenumba
from gatematrixs import IdentityGate,cxgate,HadamardGate1Qbit,HadamardGate,XGate
from qiskit import QuantumCircuit
from qiskit.circuit.random import random_circuit
def is_pos_def(x):
return np.all(np.linalg.eigvals(x) > 0)
@jit(nopython = True)
def closest_unitary(A:np.ndarray):
#https://michaelgoerz.net/notes/finding-the-closest-unitary-for-a-given-matrix/ """
#V, __, Wh = scipy.linalg.svd(A)
V, __, Wh = np.linalg.svd(A, full_matrices=False)
#U = np.matrix(V.dot(Wh))
U = V.dot(Wh)
return U
def CompareRHOS(rhoA,rhoB,lindexs):
print("\n CompareRHOS dif = ",sum(np.abs(rhoA-rhoB)))
for i in range(len(lindexs)):
x = lindexs[i]
print(x,"dif",round(abs(rhoA[x]-rhoB[x]),5))
def getspacegrid(gridsize,cellsize,centerval):
# grid size must be odd (impar)
grid = np.zeros(gridsize,dtype=np.float32)
grid[0]=centerval
n=1
for k in range(1,gridsize,2):
grid[k] = centerval + n * cellsize
grid[k+1] = centerval - n * cellsize
n+=1
return grid
@jit(nopython = True)
def evolverho(rho:np.ndarray,gate : np.ndarray):
gate_transpose = gate.transpose()
# dot same as matmul but for numba
m = np.dot(gate, rho)
m = np.dot(m, np.conjugate(gate_transpose))
return m
class gatefinder:
# gridsize must be odd
def __init__(self,minval=-1,maxval=1,dims=6,gridsize=7):
self.dims = dims
self.minval=float(minval)
self.maxval=float(maxval)
self.gridsize=gridsize
self.resetminval=self.minval
self.resetmaxval=self.maxval
self.zoom=0
self.gate2x2=None
self.makegrid()
def makegrid(self):
self.grid=np.zeros([self.dims,self.gridsize],dtype=np.float32)
spacesize=self.maxval - self.minval
self.cellsize=spacesize/(self.gridsize-1)
for i in range(self.dims):
self.grid[i] =getspacegrid(self.gridsize,self.cellsize,0)
def reset(self):
self.zoom=0
self.minval=self.resetminval
self.maxval=self.resetmaxval
self.makegrid()
def search(self,rhobefore,vecsearch,distbreak=-1):
self.reset()
self.gate2x2=None
self.mindist=1000000.
self.searchrecursive(rhobefore,vecsearch,distbreak)
print("\nmindist search",self.mindist)
return self.gate2x2
@staticmethod
@jit(nopython = True)
def searchnumba1(rhobefore:np.ndarray, vectorfind: np.ndarray, n:int,grid: np.ndarray , distbreak : int):
I=IdentityGate()
gate2x2= np.zeros((2,2),dtype = np.complex128)
counter=0
best=[0,0,0,0,0,0]
mindist=100000.
for a in range(n):
for b in range(n):
for c in range(n):
for d in range(n):
for e in range(n):
for f in range(n):
gate2x2[0][0]=grid[0][a]
gate2x2[0][1]=grid[1][b] + grid[2][c]*1j
gate2x2[1][0]=grid[3][d] + grid[4][e]*1j
gate2x2[1][1]=grid[5][f]
gate2x2=closest_unitary(gate2x2)
# make 4x4 for qbit 0
gate4x4=np.kron(I,gate2x2)
# evolve rho
rho4x4 = evolverho(rhobefore,gate4x4)
counter+=1
dist = np.abs(rho4x4 - rhoafter).sum()
if dist<mindist:
best=[a,b,c,d,e,f]
mindist=dist
if distbreak > -1:
if mindist < distbreak:
return gate2x2,mindist,best
gate2x2[0][0]=grid[0][best[0]]
gate2x2[0][1]=grid[1][best[1]] + grid[2][best[2]]*1j
gate2x2[1][0]=grid[3][best[3]] + grid[4][best[4]]*1j
gate2x2[1][1]=grid[5][best[5]]
gate2x2=closest_unitary(gate2x2)
return gate2x2,mindist,best
def searchrecursive(self,rhobefore,vecsearch,distbreak):
# distbreak= -1 for search all space
for self.zoom in range(1,100,1):
gate2x2 ,mindist,best = self.searchnumba1(rhobefore,vecsearch, self.gridsize, self.grid,distbreak)
print(" ",self.zoom,")",round(mindist,5),end="")
if self.zoom%5==0: print("")
if mindist < self.mindist:
self.gate2x2=gate2x2
self.mindist=mindist
if self.mindist > distbreak:
oldcellsize=self.cellsize
self.cellsize *=0.9
SPACERIGHTLIMIT = 1.
SPACELEFTLIMIT= -1.
for i in range(self.dims):
idxgrid = best[i]
gridcenterval= self.grid[i][idxgrid] + oldcellsize/2
if gridcenterval>SPACERIGHTLIMIT-self.cellsize/2:
gridcenterval = SPACERIGHTLIMIT-self.cellsize/2
if gridcenterval<SPACELEFTLIMIT-self.cellsize/2:
gridcenterval = SPACELEFTLIMIT-self.cellsize/2
self.grid[i]=getspacegrid(self.gridsize,self.cellsize,gridcenterval)
else:
return
print("end search by zoom limit 100")
def testgate2x2(gate2x2,rhobefore,rhoafter):
print("gate2x2 hermitian = ",is_hermitian(gate2x2),"\ttrace=",np.trace(gate2x2),"\tis_pos_def",is_pos_def(gate2x2))
gate4x4=np.kron(IdentityGate(),gate2x2) # for qubit 0
resrho=evolverho(np.copy(rhobefore),gate4x4)
CompareRHOS(resrho.flatten(),rhoafter.flatten(),np.arange(16))
idxrho=1
#seed=5
seed=7
rhobefore,rhoafter = testfinderhos(seed=seed, idxrho=idxrho) # cx (0,1) get rho 1 = [0,2
##sv = random_statevector(4,1)
##rhobefore=DensityMatrix(sv).data
##rhoafter= evolverho(np.copy(rhobefore),cxgate(control=0))
print("seed =",seed,"idxrho = ",idxrho)
print("rhobefore hermitian = ",is_hermitian(rhobefore),"\ttrace=",np.trace(rhobefore),"\tis_pos_def",is_pos_def(rhobefore))
print("rhoafter hermitian = ",is_hermitian(rhoafter),"\ttrace=",np.trace(rhoafter),"\tis_pos_def",is_pos_def(rhoafter))
##try:
## validate_density_matrix(rhoafter,qid_shape=(2,2,))
##except ValueError as e:
## # is not positive semidefinite or not hermitian or trace not 1
## print('Failed validate_density_matrix: ' + str(e))
gf= gatefinder()
resgate2x2= gf.search(rhobefore,rhoafter,0.00001)
##resgate2x2 =np.array( [[1.00000000e+00+1.77448355e-17j, 9.01055114e-17+1.31939093e-16j],
## [1.11022302e-16+0.00000000e+00j, 1.00000000e+00+0.00000000e+00j]], dtype=np.complex128)
print("seed",seed,"idxrho",idxrho)
print("resgate2x2\n",resgate2x2)
testgate2x2(resgate2x2,rhobefore,rhoafter)