-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathEvaluation.py
executable file
·289 lines (267 loc) · 15.3 KB
/
Evaluation.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
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import time
import warnings
from Regressor import MultiStart
import time
warnings.filterwarnings("ignore")
class Evaluation():
'''
Class for evaluation of the multistart curve fitting method for fitting the
parameters of predicted circuit.
'''
def __init__(self, freq_lb=-6, freq_ub=8):
'''
Args:
freq_lb (float): the logarithmic lower bound of the (later defind) frequency vector
freq_ub (float): the logarithmic upper bound of the (later defind) frequency vector
'''
self.Z_real_true = np.loadtxt('predictions/Test_Z_real')
self.Z_imag_true = np.loadtxt('predictions/Test_Z_imag')
self.Binary_Matrix_true = np.loadtxt('predictions/True_Binary')
self.degs_true = np.loadtxt('predictions/True_Deg') # RC Degree, indicates how many RCs in the circuit
self.w = np.logspace(freq_lb, freq_ub, 200) # creating the frequency vector
self.OPT = MultiStart() # calling the optimizer
def runPointsTest(self, test_set=100, N=300):
"""
Fucntion for running the points test for determing the optimal number of random points needed for the multi-start optimization method.
Args:
test_set (int): size of the test set (how many different circuits).
N (int): the maximum number of random points for intializing the optimizer.
Output:
1- saves the error data for each circuit for each number of points
2- plot the fitting error change with increasing the points
"""
error_array = np.zeros((test_set, N))
for i in range(test_set):
error_array[i], _ = self.OPT.multistart_min(self.w,
self.Z_real_true[i],
self.Z_imag_true[i],
N,
self.degs_true[i],
self.Binary_Matrix_true[i])
np.savetxt('Evaluation results/Starting_points_error', error_array)
err_90 = np.loadtxt('Evaluation results/Starting_points_error')
err_90 = err_90[:45]
N_times = np.arange(10, 300, 10)
err_list = []
for i in N_times:
err_list.append(np.nanmin(err_90[:, :i], 1))
df_boxes = pd.DataFrame(err_list).T
df_boxes.columns = N_times
plt.figure(figsize=(9, 9))
sns.boxplot(data=df_boxes, showfliers=False)
plt.xlabel('Number of Starting Points')
plt.ylabel('Relative Error')
plt.title('Multistart Evaluation')
plt.yscale('log')
plt.save()
plt.savefig('Evaluation results/Starting_points_error.png',
transparent=False, facecolor='white')
def final_evaluation(self,
Z_real_test_wide=None,
Z_imag_test_wide=None,
N_evaluation=30, test_length=300):
"""
Runs fitting tests on the predicted circuit and the true circuit.
Args:
Z_real_test_wide, Z_real_test_wide (numpy array): the real and imaginary spectra for testing,
N_evaluation (int): the number of the intialization points for the optimizer for each circuit fitting
test_length (int): the number of test circuits used for the evaluation.
Output:
1- saves fitting error of the circuits (true and predicted)
"""
degree_prediction = np.loadtxt('predictions/Pred_Deg')
binary_prediction = np.loadtxt('predictions/Pred_Binary')
top3 = np.loadtxt('predictions/Top3')
circuits_binary = np.loadtxt('predictions/CircuitsBinaryEncoding')
predication_error_evaluation = [] # appends error for predicted degree
predication_error_evaluation_2 = []
predication_error_evaluation_3 = []
True_error_evaluation = [] # appends error for true degree
deg_4_error = []
Z_real_test = np.array(Z_real_test_wide)
Z_imag_test = np.array(Z_imag_test_wide)
w = self.w
# topk_error=np.zeros((test_length,3))
time_array = np.zeros((test_length, 5))
true_circuit = []
pred_circuit = []
for i in range(test_length):
print(i)
pred_deg_i = degree_prediction[i].astype('int')
pred_binary_i = binary_prediction[i].astype('int')
pred1_time = time.time()
pred_error_array, pred_parameters_array, Z_mes, Z_fit_pred = self.OPT.multistart_min(w,
Z_real_test[i],
Z_imag_test[i],
N_evaluation,
pred_deg_i,
pred_binary_i,
Bounds=True)
pred1_time = time.time() - pred1_time
predication_error_evaluation.append(np.nanmin(pred_error_array))
pred_deg_i = (circuits_binary[top3[i, 1], 0] +2).astype('int')
pred_binary_i = (circuits_binary[top3[i, 1], 1:]).astype('int')
pred_binary_i = np.insert(pred_binary_i, 2, 0)
pred2_time = time.time()
pred_error_array, pred_parameters_array, Z_mes, Z_fit_pred = self.OPT.multistart_min(w,
Z_real_test[i],
Z_imag_test[i],
N_evaluation,
pred_deg_i,
pred_binary_i,
Bounds=True)
pred2_time =time.time() - pred2_time
predication_error_evaluation_2.append(np.nanmin(pred_error_array))
pred_deg_i = (circuits_binary[top3[i, 2], 0] +2).astype('int')
pred_binary_i = (circuits_binary[top3[i, 2], 1:]).astype('int')
pred_binary_i = np.insert(pred_binary_i, 2, 0)
pred3_time = time.time()
pred_error_array, pred_parameters_array, Z_mes, Z_fit_pred = self.OPT.multistart_min(w,
Z_real_test[i],
Z_imag_test[i],
N_evaluation,
pred_deg_i,
pred_binary_i,
Bounds=True)
pred3_time =time.time() - pred3_time
predication_error_evaluation_3.append(np.nanmin(pred_error_array))
true_deg_i = self.degs_true[i].astype('int')
true_binary_i = self.Binary_Matrix_true[i].astype('int')
true_time = time.time()
true_error_array, true_parameters_array,Z_mes,Z_fit_true = self.OPT.multistart_min(w,
Z_real_test[i],
Z_imag_test[i],
N_evaluation,
true_deg_i,
true_binary_i,
Bounds=True)
True_error_evaluation.append(np.nanmin(true_error_array))
true_time = time.time() - true_time
deg_4 = 4
binary_111 = np.array([1, 1, 0, 1, 1, 1])
UC_time = time.time()
error_array, true_parameters_array, Z_mes, Z_fit_U = self.OPT.multistart_min(w,
Z_real_test[i],
Z_imag_test[i],
N_evaluation,
deg_4,
binary_111,
Bounds=False)
deg_4_error.append(np.nanmin(error_array))
UC_time = time.time() - UC_time
true_circuit.append(np.append(true_deg_i, true_binary_i))
pred_circuit.append(np.append(pred_deg_i, pred_binary_i))
time_vec = np.array([pred1_time, pred2_time, pred3_time, true_time, UC_time])
time_array[i] = time_vec
np.savetxt('Evaluation results/Time array', time_array)
np.savetxt('Evaluation results/True circuit', true_circuit)
np.savetxt('Evaluation results/Predicted Ciruit', pred_circuit)
error_results = np.array((predication_error_evaluation,
predication_error_evaluation_2,
predication_error_evaluation_3,
True_error_evaluation, deg_4_error))
np.savetxt('Evaluation results/error_results', error_results.T)
def plot_results():
"""
plotting the error values of both the true and the predicted circuit against each other.
"""
Error_Results = np.loadtxt('Evaluation results/error_results').T
Time_results = np.loadtxt('Evaluation results/Time array').T
# pred_circuit = np.loadtxt('Evaluation results/Predicted Ciruit').T
# true_circuit = np.loadtxt('Evaluation results/True circuit').T
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 13))
points=[]
for i in range(len(Error_Results[0])):
if Error_Results[0][i]<=Error_Results[4][i]*1.2:
col = 'cornflowerblue'
label = 'Prediciotn is better'
else:
col='maroon'
# label='Prediciotn is worse'
point = ax1.scatter(Error_Results[0][i], Error_Results[4][i], c=col)
points.append(point)
ax1.set_title('Fitting Error of Predicted Circuit vs. Universal Circuit Fitting ')
ax1.grid(visible=True)
ax1.legend([points[0],points[9]],['Prediciotn is better', 'Prediciotn is worse'])
ax1.set_xlabel('Fitting Error of Predicted Circuit [%]')
ax1.set_ylabel('Fitting Error of Universal Circuit without reduction[%]')
ax1.set_xscale('log')
ax1.set_yscale('log')
# plt.savefig('PredvsU.png', transparent=False,facecolor='white')
# plt.show()
for i in range(len(Error_Results[0])):
if Error_Results[0][i] <= Error_Results[3][i]*1.2:
col = 'cornflowerblue'
else:
col = 'maroon'
ax2.scatter(Error_Results[0][i], Error_Results[3][i], c=col)
ax2.set_title('Fitting Error of Predicted Circuit vs. True Circuit Fitting ')
ax2.grid(visible=True)
ax2.legend([points[0], points[9]], ['Prediciotn is better/similar', 'Prediciotn is worse'])
ax2.set_xlabel('Fitting Error of Predicted Circuit [%]')
ax2.set_ylabel('Fitting Error of Ground Truth Circuit[%]')
ax2.set_xscale('log')
ax2.set_yscale('log')
fig.tight_layout(pad=8.0)
fig.savefig('Evaluation results/Top1Predscatter.png', transparent=False, facecolor='white')
fig.close()
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 13))
topk_error = Error_Results[:3].min(axis=0)
points = []
for i in range(len(Error_Results[0])):
if topk_error[i] <= Error_Results[4][i]*1.2:
col = 'cornflowerblue'
# label = 'Prediciotn is better'
else:
col = 'maroon'
# label='Prediciotn is worse'
point = ax1.scatter(topk_error[i], Error_Results[4][i], c=col)
points.append(point)
ax1.set_title('Fitting Error of Top-3 Predicted Circuit vs. Universal Circuit Fitting ')
ax1.grid(visible=True)
ax1.legend([points[0], points[2]], ['Prediciotn is better', 'Prediciotn is worse'])
ax1.set_xlabel('Fitting Error of Top-3 Predicted Circuit [%]')
ax1.set_ylabel('Fitting Error of Universal Circuit [%]')
ax1.set_xscale('log')
ax1.set_yscale('log')
# plt.savefig('PredvsU.png', transparent=False,facecolor='white')
# plt.show()
points=[]
for i in range(len(Error_Results[0])):
if topk_error[i]<=Error_Results[3][i]*1.2:
col = 'cornflowerblue'
else:
col='maroon'
point=ax2.scatter(topk_error[i], Error_Results[3][i], c=col)
points.append(point)
ax2.set_title('Fitting Error of Top-3 Predicted Circuit vs. True Circuit Fitting ')
ax2.grid(visible=True)
ax2.legend([points[0], points[2]], ['Prediciotn is better/similar', 'Prediciotn is worse'])
ax2.set_xlabel('Fitting Error of Predicted Circuit [%]')
ax2.set_ylabel('Fitting Error of Ground Truth Circuit[%]')
ax2.set_xscale('log')
ax2.set_yscale('log')
fig.tight_layout(pad=8.0)
fig.savefig('Evaluation results/Top3Predscatter.png', transparent=False, facecolor='white')
fig.close()
plt.figure(figsize=(9, 9))
sns.histplot(Time_results[0], kde=False)
sns.histplot(Time_results[-2], kde=False)
sns.histplot(Time_results[:3].sum(axis=0), kde=False, alpha=0.3)
plt.title('Distribuation of Fitting Time')
plt.xlabel('Time')
plt.legend(['Top-1 Prediction', 'Gound Truth', 'Top-3 Prediction'])
plt.close()
plt.figure(figsize=(9, 9))
plt.hist(Time_results[0], bins=50, alpha=0.5, color='coral')
plt.hist(Time_results[-2], bins=50, alpha=0.5, color='teal')
plt.title('Histogram of Fitting Time')
plt.xlabel('Time')
plt.legend(['Top-1 Prediction', 'Gound Truth', 'Top-3 Prediction'])
plt.grid()
plt.savefig('Evaluation results/TimeDist.png', transparent=False, facecolor='white')
plt.close()