data-driven_nonsmooth_optimization/plotting.py at master · aringh/data-driven_nonsmooth_optimization · GitHub

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
"""Code for creating the plot in the paper"""

import numpy as np
import matplotlib.pyplot as plt

save_path = 'Give a save path'


# -----------------------------------------------------------------------------
# Data
# -----------------------------------------------------------------------------
iter_count = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 15, 18, 21, 24, 27, 30,
              33, 36, 39, 42, 45, 50, 55, 60, 70, 80, 90, 100, 150, 200, 250,
              300]

# Results as average over 100 phantoms
loss_sidky = [18939.502158203126, 5025.5223461914065, 2067.0353845214845,
              1031.4667388916016, 623.06912780761718, 466.60300750732421,
              400.4218392944336, 361.47143035888672, 331.44562438964846,
              307.17602478027345, 288.30622772216799, 274.07745544433595,
              248.31353317260744, 234.24225967407227, 225.42493766784668,
              219.48006118774413, 215.23997070312501, 212.09245178222656,
              209.68343795776366, 207.79406005859374, 206.28261077880859,
              205.05378005981444, 204.04111808776855, 202.71044845581054,
              201.70333206176758, 200.92532341003417, 199.82694763183594,
              199.11397659301758, 198.63221626281739, 198.29623893737792,
              197.57292976379395, 197.37533462524414, 197.30535522460937,
              197.27619499206543]

loss_cp_const = [12886.935864257812, 3510.6859374999999, 1399.0583532714843,
                 659.00321716308599, 468.40173004150392, 438.02103546142575,
                 410.64037628173827, 363.5603674316406, 315.28353790283205,
                 279.62799102783202, 257.87679611206056, 245.26353256225585,
                 225.95082260131835, 216.28674453735351, 211.133060836792,
                 207.79802726745606, 205.5376668548584, 203.91241615295411,
                 202.69393508911133, 201.75682289123534, 201.02152137756349,
                 200.43507926940919, 199.96114624023437, 199.35437835693358,
                 198.91007316589355, 198.57754737854003, 198.12849472045897,
                 197.85297943115233, 197.67618385314941, 197.55838035583497,
                 197.32764778137206, 197.2742194366455, 197.25765716552735,
                 197.25151260375978]

loss_our_2x2 = [180977.49460937499, 103398.1903515625, 40974.235830078127,
                12934.823881835937, 3311.2684631347656, 877.31835174560547,
                369.17604156494139, 244.1925839996338, 222.44221405029296,
                221.42951957702635, 215.50117004394531, 210.15300270080567,
                205.31969169616698, 202.7324617767334, 201.18980865478517,
                200.17820907592773, 199.47986991882325, 198.98071517944337,
                198.61470878601074, 198.34082000732423, 198.13216163635255,
                197.97064826965331, 197.84404045104981, 197.68805557250977,
                197.57898353576661, 197.50089736938477, 197.40151451110839,
                197.34509552001953, 197.31142684936523, 197.29050506591796,
                197.25502494812011, 197.24878273010253, 197.2472582244873,
                197.24680015563965]

loss_cp = [177568.36874999999, 99800.789648437494, 39084.240976562498,
           12264.618803710937, 3101.2784545898439, 846.67608215332029,
           388.93958206176757, 252.46717437744141, 223.79262588500976,
           225.00771293640136, 217.26661735534668, 209.11574607849121,
           204.27016532897949, 201.70856002807616, 200.35189834594726,
           199.51293144226074, 198.95216743469237, 198.55739852905273,
           198.27155799865722, 198.0597243499756, 197.89971351623535,
           197.77689697265626, 197.68139686584473, 197.56472198486327,
           197.48413764953614, 197.4272787475586, 197.35644210815428,
           197.31780197143556, 197.2958839416504, 197.28319534301758,
           197.26755088806152, 197.26978858947754, 197.27350715637206,
           197.27663574218749]

loss_general_nm2 = [25133.611562499998, 9579.276791992188, 3766.6393432617188,
                    1675.0945513916015, 614.80189483642573, 388.86414230346679,
                    280.45288146972655, 232.67581756591798, 221.72526504516603,
                    217.27104461669921, 212.78144134521483, 208.80965476989746,
                    205.57998359680175, 201.87361968994139, 201.15900695800781,
                    200.31115478515625, 200.61713027954102, 201.36368919372558,
                    203.17898078918458, 206.33866371154784, 211.85435333251954,
                    221.37628929138182, 238.23620193481446, 303.99950210571291,
                    512.95106170654299, 1298.7509350585938, 21942.769462890625,
                    622258.85062499996, 20314128.739999998, 698223412.15999997,
                    44445109018312048.0, 3.4943773972890217e+24, 3.0364593370338254e+32,
                    np.inf]

loss_general_nm3 = [5526.8552441406246, 6227.8524243164065, 928.92287109375002,
                    532.07267700195314, 366.97947570800784, 279.29114532470703,
                    245.24093276977538, 222.62817131042482, 217.06319129943847,
                    212.15167015075684, 210.64053039550782, 207.48350738525392,
                    209.38224647521972, 212.69040100097655, 279.17968887329101,
                    508.82947814941406, 2091.5332922363282, 13776.14638671875,
                    152989.184453125, 2302716.473125, 39822794.369999997,
                    724085302.24000001, 13443981265.92, 1791730730598.3999,
                    243351063992729.59, 33512791449177948.0, 6.5539290926932872e+20,
                    1.3205202060865098e+25, 2.7184337183860469e+29, 5.6880178123594228e+33,
                    np.inf, np.nan, np.nan, np.nan]


normalize_values = True
normalization_value = 197.246657104  # Computed using CP solver with Sidky et.al parameters, 1000 iter.

if normalize_values:
    loss_sidky = [value - normalization_value for value in loss_sidky]
    loss_cp_const = [value - normalization_value for value in loss_cp_const]
    loss_our_2x2 = [value - normalization_value for value in loss_our_2x2]
    loss_cp = [value - normalization_value for value in loss_cp]
    loss_general_nm2 = [value - normalization_value for value in loss_general_nm2]
    loss_general_nm3 = [value - normalization_value for value in loss_general_nm3]


# -----------------------------------------------------------------------------
# Create the plot
# -----------------------------------------------------------------------------
# Create the big plot
fig, ax1 = plt.subplots()

lines = ax1.plot(iter_count, loss_sidky)
plt.setp(lines, color='xkcd:red', linestyle='-', linewidth=1.5)

lines = ax1.plot(iter_count, loss_cp_const)
plt.setp(lines, color='xkcd:green', linestyle='-', linewidth=1.5)

lines = ax1.plot(iter_count, loss_our_2x2)
plt.setp(lines, color='xkcd:sky blue', linestyle='-', linewidth=1.5)


lines = ax1.plot(iter_count, loss_cp)
plt.setp(lines, color='xkcd:orange', linestyle='-.', linewidth=1.5)

lines = ax1.plot(iter_count, loss_general_nm2)
plt.setp(lines, color='xkcd:purple', linestyle='-.', linewidth=1.5)

lines = ax1.plot(iter_count, loss_general_nm3)
plt.setp(lines, color='xkcd:black', linestyle='-.', linewidth=1.5)


ax1.axis([1, 300, 0.01, 10000])

plt.xscale('log')
plt.yscale('log')

plt.axvline(x=10, linestyle='--',  color='xkcd:dark gray', linewidth=0.8)

plt.rc('font', family='serif')

ax1.legend(['(i)', '(ii)', '(iii)', '(iv)', '(v)', '(vi)'])


# Create a small subplot
left, bottom, width, height = [0.2, 0.18, 0.25, 0.23]
ax2 = fig.add_axes([left, bottom, width, height])

from mpl_toolkits.axes_grid1.inset_locator import mark_inset
mark_inset(ax1, ax2, loc1=1, loc2=2, joinstyle='round')

lower_index = 6
upper_index = 15

lines = ax2.plot(iter_count[lower_index:upper_index], loss_sidky[lower_index:upper_index])
plt.setp(lines, color='xkcd:red', linestyle='-', linewidth=1.5)

lines = ax2.plot(iter_count[lower_index:upper_index], loss_cp_const[lower_index:upper_index])
plt.setp(lines, color='xkcd:green', linestyle='-', linewidth=1.5)

lines = ax2.plot(iter_count[lower_index:upper_index], loss_our_2x2[lower_index:upper_index])
plt.setp(lines, color='xkcd:sky blue', linestyle='-', linewidth=1.5)


lines = ax2.plot(iter_count[lower_index:upper_index], loss_cp[lower_index:upper_index])
plt.setp(lines, color='xkcd:orange', linestyle='-.', linewidth=1.5)

lines = ax2.plot(iter_count[lower_index:upper_index], loss_general_nm2[lower_index:upper_index])
plt.setp(lines, color='xkcd:purple', linestyle='-.', linewidth=1.5)

lines = ax2.plot(iter_count[lower_index:upper_index], loss_general_nm3[lower_index:upper_index])
plt.setp(lines, color='xkcd:black', linestyle='-.', linewidth=1.5)

plt.xscale('log')
plt.yscale('log')

plt.axvline(x=10, linestyle='--',  color='xkcd:dark gray', linewidth=0.8)


ax2.axis([7, 20, 5, 100])
ax2.xaxis.set_ticks([1], minor='True')


# Save the figure
fig.savefig(filename=save_path+'obj_fun_values', format='eps')