-
Notifications
You must be signed in to change notification settings - Fork 2
/
apples_plot_metrics.py
235 lines (224 loc) · 10.5 KB
/
apples_plot_metrics.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
# -*- coding: utf-8 -*-
import os
import json
import argparse
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker
import matplotlib.lines as mlines
import matplotlib.collections as mcollections
import seaborn as sns
import pandas as pd
IMPL = 'astra_cuda'
METRICS_PATH = '../learned_ct_reco_comparison_paper_results/metrics'
FIG_PATH = 'figures'
os.makedirs(FIG_PATH, exist_ok=True)
parser = argparse.ArgumentParser()
parser.add_argument('--methods', type=str, nargs='+', default=None)
parser.add_argument('--exclude_methods', type=str, nargs='+', default=[])
parser.add_argument('--save_formats', type=str, nargs='+',
default=['png', 'pdf'])
# default=[])
parser.add_argument('--plot_types', type=str, nargs='+',
default=[
'num_angles_on_x',
'noise_setting_on_x',
# 'noise_setting_and_num_angles_on_x_no_method_names',
'noise_setting_and_num_angles_on_x'
])
options = parser.parse_args()
methods = options.methods
if methods is None:
methods = ['learnedpd', 'fbpistaunet', 'fbpunet', 'fbpmsdnet', 'cinn', 'ictnet', 'tv', 'cgls', 'fbp']
exclude_methods = options.exclude_methods
save_formats = options.save_formats
plot_types = options.plot_types
method_list = [m for m in methods if m not in exclude_methods]
num_angles_list = [50, 10, 5, 2]
noise_setting_list = ['noisefree', 'gaussian_noise', 'scattering']
noise_setting_name_dict = {
'noisefree': 'Noise-free',
'gaussian_noise': 'Gaussian noise',
'scattering': 'Scattering'}
noise_setting_name_list = [
noise_setting_name_dict[noise_setting]
for noise_setting in noise_setting_list]
method_name_dict = {'learnedpd': 'Learned Primal-Dual',
'fbpistaunet': 'ISTA U-Net',
'fbpunet': 'U-Net',
'fbpmsdnet': 'MS-D-CNN',
'fbpunetpp': 'U-Net++',
'cinn': 'CINN',
'diptv': 'DIP + TV',
'ictnet': 'iCTU-Net',
'cgls': 'CGLS',
'tv': 'TV',
'fbp': 'FBP'}
def get_metrics(noise_setting, num_angles, method):
name = 'apples_{}_{:02d}_{}'.format(noise_setting, num_angles, method)
try:
with open(os.path.join(METRICS_PATH, '{}_metrics.json'.format(name)),
'r') as f:
metrics = json.load(f)
except FileNotFoundError:
metrics = {}
return metrics
aggregate_keys = [
'psnr_mean', 'psnr_std', 'psnr_min', 'psnr_max',
'ssim_mean', 'ssim_std', 'ssim_min', 'ssim_max',
'psnr_data_range_fixed_mean', 'psnr_data_range_fixed_std',
'psnr_data_range_fixed_min', 'psnr_data_range_fixed_max',
'ssim_data_range_fixed_mean', 'ssim_data_range_fixed_std',
'ssim_data_range_fixed_min', 'ssim_data_range_fixed_max']
records = [
{'noise_setting_name': noise_setting_name_dict[n],
'num_angles': a,
'method_name': method_name_dict[m],
**get_metrics(n, a, m).get('aggregates',
{k: np.nan for k in aggregate_keys})
}
for n in noise_setting_list for a in num_angles_list for m in method_list]
df = pd.DataFrame.from_records(records)
if 'num_angles_on_x' in plot_types:
for noise_setting in noise_setting_list:
fig, ax = plt.subplots(1, 2, figsize=(15, 5))
for method in method_list:
metrics_per_angle = [
get_metrics(noise_setting, num_angles, method)
for num_angles in num_angles_list]
psnr_per_angle = [m.get('aggregates', {}).get('psnr_mean', np.nan)
for m in metrics_per_angle]
ssim_per_angle = [m.get('aggregates', {}).get('ssim_mean', np.nan)
for m in metrics_per_angle]
fig.suptitle(noise_setting_name_dict[noise_setting])
ax[0].plot(num_angles_list, psnr_per_angle, 'D--',
label=method_name_dict[method])
ax[1].plot(num_angles_list, ssim_per_angle, 'D--',
label=method_name_dict[method])
ax[0].set_xscale('log')
ax[0].set_xticks(num_angles_list)
ax[0].get_xaxis().set_major_formatter(
matplotlib.ticker.ScalarFormatter())
ax[0].invert_xaxis()
ax[0].set_xlabel('Number of angles')
ax[0].set_ylabel('PSNR (dB)')
ax[0].grid()
ax[1].set_xscale('log')
ax[1].set_xticks(num_angles_list)
ax[1].get_xaxis().set_major_formatter(
matplotlib.ticker.ScalarFormatter())
ax[1].invert_xaxis()
ax[1].set_xlabel('Number of angles')
ax[1].set_ylabel('SSIM')
ax[1].grid()
ax[1].legend(loc='lower left')
if 'pdf' in save_formats:
filename = os.path.join(
FIG_PATH, 'apples_metrics_{}.pdf'.format(noise_setting))
fig.savefig(filename, bbox_inches='tight')
if 'png' in save_formats:
filename_png = os.path.join(
FIG_PATH, 'apples_metrics_{}.png'.format(noise_setting))
fig.savefig(filename_png, bbox_inches='tight', dpi=300)
if 'noise_setting_on_x' in plot_types:
for num_angles in num_angles_list:
fig, ax = plt.subplots(1, 2, figsize=(10, 3.5))
fig.subplots_adjust(wspace=0.3)
method_name_list = [method_name_dict[m] for m in method_list]
x, hue = np.meshgrid(noise_setting_name_list, method_name_list)
x, hue = x.ravel(), hue.ravel()
metrics_per_setting_per_method = [
get_metrics(noise_setting, num_angles, method)
for method in method_list for noise_setting in noise_setting_list]
y_psnr = [m.get('aggregates', {}).get('psnr_mean', np.nan)
for m in metrics_per_setting_per_method]
y_ssim = [m.get('aggregates', {}).get('ssim_mean', np.nan)
for m in metrics_per_setting_per_method]
fig.suptitle('{:d} angles'.format(num_angles))
sns.swarmplot(x, y_psnr, hue, ax=ax[0])
ax[0].set_ylabel('PSNR (dB)')
ax[0].get_legend().remove()
sns.swarmplot(x, y_ssim, hue, ax=ax[1])
ax[1].set_ylabel('SSIM')
ax[1].get_legend()._loc = 6 # center left
ax[1].get_legend().set_bbox_to_anchor((1.05, 0.5))
ax[1].get_legend().set_title('Method')
if 'pdf' in save_formats:
filename = os.path.join(
FIG_PATH, 'apples_metrics_{:02d}_angles.pdf'.format(
num_angles))
fig.savefig(filename, bbox_inches='tight')
if 'png' in save_formats:
filename_png = os.path.join(
FIG_PATH, 'apples_metrics_{:02d}_angles.png'.format(
num_angles))
fig.savefig(filename_png, bbox_inches='tight', dpi=300)
# if 'noise_setting_and_num_angles_on_x_no_method_names' in plot_types:
# # note: bad order of noise settings swarms due to invert_xaxis + dodge
# fig, ax = plt.subplots(1, 2, figsize=(15, 3.5))
# sns.swarmplot(x='num_angles', y='psnr_mean', hue='noise_setting_name', data=df, dodge=True, ax=ax[0])
# ax[0].invert_xaxis()
# ax[0].set_ylabel('PSNR (dB)')
# ax[0].get_legend().remove()
# sns.swarmplot(x='num_angles', y='ssim_mean', hue='noise_setting_name', data=df, dodge=True, ax=ax[1])
# ax[1].invert_xaxis()
# ax[1].set_ylabel('SSIM')
# ax[1].get_legend().set_title('')
# ax[1].get_legend()._loc = 6 # center left
# ax[1].get_legend().set_bbox_to_anchor((1.05, 0.5))
# if 'pdf' in save_formats:
# filename = os.path.join(
# FIG_PATH, 'apples_metrics_all_in_one_no_method_names.pdf')
# fig.savefig(filename, bbox_inches='tight')
# if 'png' in save_formats:
# filename_png = os.path.join(
# FIG_PATH, 'apples_metrics_all_in_one_no_method_names.png')
# fig.savefig(filename_png, bbox_inches='tight', dpi=300)
if 'noise_setting_and_num_angles_on_x' in plot_types:
fig, ax = plt.subplots(1, 2, figsize=(15, 3.5))
x_offsets = np.linspace(-.25, .25, len(noise_setting_list))
markers = ['P', 'o', 'X']
for i, n in enumerate(noise_setting_list):
df_noise_setting = df[df['noise_setting_name']==noise_setting_name_dict[n]]
children_old = ax[0].get_children()
sns.swarmplot(x='num_angles', y='psnr_mean', hue='method_name', data=df_noise_setting, dodge=False, marker=markers[i], ax=ax[0])
for c in ax[0].get_children():
if (isinstance(c, mcollections.PathCollection)
and c not in children_old):
c.set_offsets(c.get_offsets() - [x_offsets[i], 0])
children_old = ax[1].get_children()
sns.swarmplot(x='num_angles', y='ssim_mean', hue='method_name', data=df_noise_setting, dodge=False, marker=markers[i], ax=ax[1])
for c in ax[1].get_children():
if (isinstance(c, mcollections.PathCollection)
and c not in children_old):
c.set_offsets(c.get_offsets() - [x_offsets[i], 0])
if i == 0:
handles, labels = ax[1].get_legend_handles_labels()
ax[0].invert_xaxis()
ax[1].invert_xaxis()
ax[0].set_xlabel('Number of angles')
ax[1].set_xlabel('Number of angles')
ax[0].set_ylabel('PSNR (dB)')
ax[1].set_ylabel('SSIM')
ax[0].get_legend().remove()
method_handles, method_labels = handles[:len(method_list)], labels[:len(method_list)]
method_handles = [mlines.Line2D([], [], color=h.get_facecolors()[0], marker='s', linestyle='') for h in method_handles]
method_legend = ax[1].legend(method_handles, method_labels)
method_legend._loc = 2 # upper left
method_legend.set_bbox_to_anchor((1.05, 0.675))
method_legend.set_title('Method')
ax[1].add_artist(method_legend)
noise_setting_handles = [mlines.Line2D([], [], color='k', marker=markers[i], linestyle='') for i in range(len(noise_setting_list))]
noise_setting_labels = noise_setting_name_list
noise_setting_legend = ax[1].legend(noise_setting_handles, noise_setting_labels)
noise_setting_legend._loc = 3 # lower left
noise_setting_legend.set_bbox_to_anchor((1.05, 0.675))
noise_setting_legend.set_title('Noise setting')
if 'pdf' in save_formats:
filename = os.path.join(
FIG_PATH, 'apples_metrics_all_in_one.pdf')
fig.savefig(filename, bbox_inches='tight')
if 'png' in save_formats:
filename_png = os.path.join(
FIG_PATH, 'apples_metrics_all_in_one.png')
fig.savefig(filename_png, bbox_inches='tight', dpi=300)