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
|
import os
import multiprocessing
import time
import numpy as np
import pyfits
import glob
import subprocess
import matplotlib.pyplot as plt
image = "30dor_256"
n_tests = 10
input_SNR = 30
def kernel_settings(kernel):
J = 4
oversample = 2
if kernel == "pswf":
J = 6
if kernel == "kb_interp":
oversample = 1.375
J = 5
if kernel == "box":
J = 1
if kernel == "kb_min":
oversample = 1.375
J = 5
if kernel == "kb_min4":
oversample = 1.375
J = 4
kernel = "kb_min"
if kernel == "kb_interp4":
oversample = 1.375
J = 4
kernel = "kb_interp"
return J, oversample, kernel
def run_test_padmm_reweighted((i, kernel, M_N_ratio, start_time, input_SNR, image)):
time.sleep(start_time)
J, oversample, kernel = kernel_settings(kernel)
os.system("screen -S "+ kernel + "_" + str(i) + " -d -m " +
"../build/cpp/example/padmm_reweighted_simulation " + kernel + " "
+ str(oversample) + " " +str(J) + " " +str(M_N_ratio) + " " + str(i) + " "+str(input_SNR)+ " " + str(image))
results_file = "../build/outputs/"+image+"_results_" + kernel + "_" + str(i) + ".txt"
while not os.path.exists(results_file):
time.sleep(1)
results = np.loadtxt(results_file, dtype = str)
SNR = results[0]
total_time = results[1]
os.system("rm " + results_file)
return [float(SNR), float(total_time)]
def run_test_padmm((i, kernel, M_N_ratio, start_time, input_SNR, image)):
time.sleep(start_time)
J, oversample, kernel = kernel_settings(kernel)
os.system("screen -S "+ kernel + "_" + str(i) + " -d -m " +
"../build/cpp/example/padmm_simulation padmm " + kernel + " "
+ str(oversample) + " " +str(J) + " " +str(M_N_ratio) + " " + str(i) + " "+str(input_SNR)+ " " + str(image))
results_file = "../build/outputs/"+image+"_results_" + kernel + "_" + str(i) + ".txt"
while not os.path.exists(results_file):
time.sleep(1)
results = np.loadtxt(results_file, dtype = str)
SNR = results[0]
total_time = results[1]
converged = results[2]
niters = results[3]
os.system("rm " + results_file)
return [float(SNR), float(total_time), int(converged), float(niters)]
def run_test_ms_clean((i, kernel, M_N_ratio, start_time, input_SNR, image)):
time.sleep(start_time)
J, oversample, kernel = kernel_settings(kernel)
os.system("screen -S "+ kernel + "_" + str(i) + " -d -m " +
"../build/cpp/example/padmm_simulation ms_clean " + kernel + " "
+ str(oversample) + " " +str(J) + " " +str(M_N_ratio) + " " + str(i) + " "+str(input_SNR)+ " " + str(image))
results_file = "../build/outputs/"+image+"_results_" + kernel + "_" + str(i) + ".txt"
while not os.path.exists(results_file):
time.sleep(1)
results = np.loadtxt(results_file, dtype = str)
SNR = results[0]
total_time = results[1]
converged = results[2]
niters = results[3]
os.system("rm " + results_file)
return [float(SNR), float(total_time), int(converged), float(niters)]
def run_test_clean((i, kernel, M_N_ratio, start_time, input_SNR, image)):
time.sleep(start_time)
J, oversample, kernel = kernel_settings(kernel)
os.system("screen -S "+ kernel + "_" + str(i) + " -d -m " +
"../build/cpp/example/padmm_simulation clean " + kernel + " "
+ str(oversample) + " " +str(J) + " " +str(M_N_ratio) + " " + str(i) + " "+str(input_SNR)+ " " + str(image))
results_file = "../build/outputs/"+image+"_results_" + kernel + "_" + str(i) + ".txt"
while not os.path.exists(results_file):
time.sleep(1)
results = np.loadtxt(results_file, dtype = str)
SNR = results[0]
total_time = results[1]
converged = results[2]
niters = results[3]
os.system("rm " + results_file)
return [float(SNR), float(total_time), int(converged), float(niters)]
def collect_data(args, results, M_N_ratios, kernel):
meantempSNR = []
errortempSNR = []
meantempTime = []
errortempTime = []
meantempIters = []
errortempIters = []
totaltempConverges = []
for m in M_N_ratios:
tempSNR = []
tempTime = []
tempIters = []
tempConverges = []
for i in range(len(args)):
if m == args[i][2]:
if args[i][1] == kernel:
tempSNR.append(results[i][0])
tempTime.append(results[i][1])
tempConverges.append(results[i][2])
tempIters.append(results[i][3])
tempSNR = np.array(tempSNR)
tempTime = np.array(tempTime)
tempConverges = np.array(tempConverges)
tempIters = np.array(tempIters)
meantempSNR.append(tempSNR.mean())
errortempSNR.append(tempSNR.std())
meantempTime.append(tempTime.mean())
errortempTime.append(tempTime.std())
meantempIters.append(tempIters.mean())
errortempIters.append(tempIters.std())
totaltempConverges.append(tempConverges.sum())
return meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges
def create_plots(args, results, M_N_ratios, name, kernels, colours,legend = []):
for k in range(len(kernels)):
meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges = collect_data(args, results, M_N_ratios, kernels[k])
plt.errorbar(M_N_ratios, meantempSNR, errortempSNR, fmt='', c = colours[k])
if len(legend) > 0:
plt.legend(legend)
plt.xlabel("M/N")
plt.ylabel("SNR, db")
plt.xlim(0, 2.2)
plt.ylim(5, 40)
plt.savefig(name + "_SNR_plot.pdf")
plt.clf()
for k in range(len(kernels)):
meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges = collect_data(args, results, M_N_ratios, kernels[k])
plt.errorbar(M_N_ratios, meantempTime, errortempTime, fmt='', c = colours[k])
plt.xlabel("M/N")
plt.ylabel("Time (seconds)")
plt.xlim(0, 2.2)
plt.savefig(name + "_Time_plot.pdf")
plt.clf()
for k in range(len(kernels)):
meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges = collect_data(args, results, M_N_ratios, kernels[k])
plt.errorbar(M_N_ratios, meantempIters, errortempIters, fmt='', c = colours[k])
plt.xlabel("M/N")
plt.ylabel("Iterations")
plt.xlim(0, 2.2)
plt.ylim(0, 110)
plt.savefig(name + "_Iterations_plot.pdf")
plt.clf()
for k in range(len(kernels)):
meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges = collect_data(args, results, M_N_ratios, kernels[k])
plt.scatter(M_N_ratios, totaltempConverges, c = colours[k])
plt.xlabel("M/N")
plt.ylabel("Number of converging tests")
plt.ylim(0, 11)
plt.xlim(0, 2.2)
plt.savefig(name + "_Converges_plot.pdf")
plt.clf()
for k in range(len(kernels)):
meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges = collect_data(args, results, M_N_ratios, kernels[k])
x = np.column_stack((meantempSNR, errortempSNR, meantempTime, errortempTime, meantempIters, errortempIters, totaltempConverges))
np.savetxt(name + '_' + kernels[k], x)
if __name__ == '__main__':
M_N_ratios = np.arange(1, 11) * 0.2
args = []
test_num = 0
#kernels = ["kb", "kb_interp", "pswf", "gauss", "box", "gauss_alt", "kb_min", "kb_min4"]#, "kb_interp4"]
kernels = ["kb", "pswf", "gauss", "gauss_alt", "box"]
total_tests = n_tests * len(kernels) * len(M_N_ratios)
for i in range(1, n_tests + 1):
for k in kernels:
for m in M_N_ratios:
test_num = test_num + 1
args.append((test_num, k, m, test_num * 1./ total_tests * 30., input_SNR, image))
print test_num
n_processors = multiprocessing.cpu_count() + 1
p = multiprocessing.Pool(min(n_processors, 1)) # Limiting the number of processes used to 40, otherwise it will cause problems with the user limit
#legend = ["Kaiser Bessel (KB)", "KB (Linear-interp, Min-oversample)", "PSWF", "Gaussian (Optimal)", "Box", "Gaussian (non-Optimal)", "KB (Min-oversample)", "KB4 (Min-oversample)"]#, "KB4 (Linear-interp, Min-oversample)"]
#colours = ['blue', 'red', 'black', 'green', 'magenta', 'cyan', 'yellow', "#800000"]#, "#808000"]
legend = ["Kaiser Bessel (KB)", "PSWF", "Gaussian (Optimal)", "Gaussian (non-Optimal)", "Box"]
colours = ['blue', 'black', 'green', 'cyan', 'magenta']
results = p.map(run_test_padmm, args)
create_plots(args, results, M_N_ratios, image + "_padmm_" + str(input_SNR) + "_", kernels, colours, legend)
print "PADMM Done!"
results = p.map(run_test_ms_clean, args)
create_plots(args, results, M_N_ratios, image + "_ms_clean" + str(input_SNR) + "_", kernels, colours)
print "MS CLEAN Done!"
results = p.map(run_test_clean, args)
create_plots(args, results, M_N_ratios, image + "_clean" + str(input_SNR) + "_", kernels, colours)
print "CLEAN Done!"
#results = p.map(run_test_padmm_reweighted, args)
#create_plots(args, results, M_N_ratios, image + "_padmm_reweighted" + str(input_SNR) + "_", kernels, colours, legend)
#print "PADMM REWEIGHTED Done!"
|
