-
Notifications
You must be signed in to change notification settings - Fork 8
/
graphing.py
355 lines (315 loc) · 17.6 KB
/
graphing.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
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
import torch
import torch.utils.data
import torchvision.transforms as transforms
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
from matplotlib import cm
import cv2
from PIL import Image
from dataHelper import DatasetFolder, DatasetFolder_WithReplacement
from load_run import accu, auroc, bootstrapped_test, calc_kappa
from highlighting_precision import hp
from vanilla_vgg import Vanilla_VGG
import re
import numpy as np
import os
import train_and_test as tnt
from sklearn.metrics import roc_curve, auc
from preprocess import mean, std, preprocess_input_function, undo_preprocess_input_function
import argparse
# line styles
linestyle_tuple = [
('solid', 'solid'),
('densely dashed', (0, (5, 1))),
('densely dotted', (0, (1, 1))),
('densely dashdotted', (0, (3, 1, 1, 1))),
('densely dashdotdotted', (0, (3, 1, 1, 1, 1, 1))),
('densely dashdashdotted', (0, (3, 1, 3, 1, 1, 1))),
('dashed', (0, (5, 2))),
('dotted', (0, (1, 2))),
('dashdotted', (0, (3, 2, 1, 2))),
('dashdotdotted', (0, (3, 2, 1, 2, 1, 2))),
('dashdashdotted', (0, (3, 2, 3, 2, 1, 2))),
# ('loosely dashed', (0, (5, 10))),
# ('loosely dashdotted', (0, (3, 10, 1, 10))),
# ('loosely dotted', (0, (1, 5))),
# ('dashdotdotted', (0, (3, 5, 1, 5, 1, 5))),
# ('loosely dashdotdotted', (0, (3, 10, 1, 10, 1, 10)))
]
marker_style = ["x", "+", "o", "s"]
def auroc_curves(test_dir, model_paths, model_labels, save_path, target_class, greyscale_support=True, kappa_vals=False, confusion=False, default_col=False):
# test loader set up
test_dataset = DatasetFolder(
test_dir,
augmentation=False,
loader=np.load,
extensions=("npy",),
transform=transforms.Compose([
torch.from_numpy,
])
)
num_classes=len(test_dataset.classes)
test_loader = torch.utils.data.DataLoader(
test_dataset, batch_size=100, shuffle=True,
num_workers=4, pin_memory=False)
# colors, ref: https://stackoverflow.com/a/33905962/7521428
start = 0.1
stop = 0.9
number_of_colors = len(model_labels)
cm_subsection = np.linspace(start, stop, number_of_colors)
colors = [ cm.nipy_spectral(x) for x in cm_subsection ]
#linestyles from global var
# linestyle_tuple = linestyle_tuple
plt.figure(figsize=(5.0, 3.6), dpi=200)
# plot a line for each model
for model_idx, model_path in enumerate(model_paths):
model = torch.load(model_path)
total_one_hot_label, total_output = [], []
if kappa_vals or confusion:
confusion_matrix = np.zeros((num_classes, num_classes))
for i, (image, label, patient_id) in enumerate(test_loader):
input = image.cuda()
target = label.cuda()
grad_req = torch.no_grad()
with grad_req:
if 'vanilla' in model_path:
output = model.forward(input)
else:
output, min_distances, _ = model.forward(input)
# one hot label for AUC
one_hot_label = np.zeros(shape=(len(target), num_classes))
for k in range(len(target)):
one_hot_label[k][target[k].item()] = 1
prob = torch.nn.functional.softmax(output, dim=1)
total_output.extend(prob.data.cpu().numpy())
total_one_hot_label.extend(one_hot_label)
# confusion matrix
if kappa_vals or confusion:
_, predicted = torch.max(output.data, 1)
for t_idx, t in enumerate(label):
confusion_matrix[predicted[t_idx]][t] += 1 #row is predicted, col is true
if kappa_vals:
kappa, oneclass_confusion_matrix = calc_kappa(confusion_matrix, target_class)
if confusion:
print('The confusion matrix for model {0} on class {1} is: \n{2}'.format(model_path, target_class, confusion_matrix))
if kappa_vals:
print('The kappa value for model {0} on class {1} is: {2}. Matrix: \n{3}'.format(model_path, target_class, kappa, oneclass_confusion_matrix))
total_output = np.array(total_output)
total_one_hot_label = np.array(total_one_hot_label)
# print(total_output[:5])
# Compute ROC curve and ROC area for each class
fpr = dict()
tpr = dict()
roc_auc = dict()
for i in range(num_classes):
fpr[i], tpr[i], _ = roc_curve(total_one_hot_label[:, i], total_output[:, i])
roc_auc[i] = auc(fpr[i], tpr[i])
print("Plotting model {} of {}.".format(model_idx+1, len(model_paths)))
lw = 1
label = '{} (area = {:.2f})'.format(model_labels[model_idx], roc_auc[target_class])
if (greyscale_support) and (not default_col):
plt.plot(fpr[target_class], tpr[target_class], color=colors[model_idx],
lw=lw, label=label, linestyle=linestyle_tuple[model_idx][1])
elif (not greyscale_support) and (not default_col):
plt.plot(fpr[target_class], tpr[target_class], color=colors[model_idx],
lw=lw, label=label, linestyle='solid')
elif (not greyscale_support) and (default_col):
plt.plot(fpr[target_class], tpr[target_class],
lw=lw, label=label, linestyle='solid')
else:
plt.plot(fpr[target_class], tpr[target_class],
lw=lw, label=label, linestyle=linestyle_tuple[model_idx][1])
plt.plot([0, 1], [0, 1], color='k', lw=lw, linestyle='--')
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.0])
plt.xlabel('FPR')
plt.ylabel('TPR')
# plt.title('AUC')
plt.legend(loc="lower right")
print('Saving to:', save_path)
plt.savefig(save_path, bbox_inches='tight', pad_inches=0)
def comparison_plot(test_dir, model_paths, legend_labels, plot_xs, save_path, greyscale_support=False, y_func=[(auroc, 'AUROC')]):
if len(y_func) == 1:
y_func = y_func[0]
assert len(model_paths) == len(legend_labels)
assert len(model_paths[0]) == len(plot_xs)
# colors, ref: https://stackoverflow.com/a/33905962/7521428
start = 0.1
stop = 0.9
number_of_colors = len(legend_labels)
cm_subsection = np.linspace(start, stop, number_of_colors)
colors = [ cm.nipy_spectral(x) for x in cm_subsection ]
#linestyles from global var
# linestyle_tuple = linestyle_tuple
# graphing
fig = plt.figure(figsize=(4, 3))
ax = fig.add_subplot(111)
for group_idx, model_group in enumerate(model_paths):
xs = [item+1 for item in range(len(plot_xs))]
ys = []
for model_path in model_group:
auroc_ = y_func[0](test_dir, model_path)
ys.append(auroc_)
if greyscale_support:
ax.plot(xs, ys, label=legend_labels[group_idx],
linewidth=1, linestyle=linestyle_tuple[group_idx][1], marker="None")
else:
ax.plot(xs, ys, label=legend_labels[group_idx],
linewidth=1.5, linestyle="solid", marker="None")
# ax.set_xscale('log')
# fig.canvas.draw()
ax.set_xticks(xs)
ax.set_xticklabels(plot_xs)
# ax.get_xaxis().set_major_formatter(mpl.ticker.ScalarFormatter()) # https://stackoverflow.com/questions/14530113/set-ticks-with-logarithmic-scale
# ax.set_xlim([0.5, 4.5])
ax.set_ylim([0.0, 1.0])
ax.set_xlabel('Fine annotation coefficient')
ax.set_ylabel(y_func[1])
# plt.title('AUC')
if len(legend_labels) > 1:
ax.legend(loc="lower right")
print('Saving to:', save_path)
plt.savefig(save_path, bbox_inches='tight', pad_inches=0, dpi=600)
else:
assert len(model_paths) == 1
assert len(legend_labels) == 1
assert len(model_paths[0]) == len(plot_xs)
# colors, ref: https://stackoverflow.com/a/33905962/7521428
start = 0.1
stop = 0.9
number_of_colors = len(y_func)
cm_subsection = np.linspace(start, stop, number_of_colors)
colors = [ cm.nipy_spectral(x) for x in cm_subsection ]
#linestyles from global var
# linestyle_tuple = linestyle_tuple
# graphing
fig = plt.figure(figsize=(4, 3))
ax = fig.add_subplot(111)
for y_func_idx, (y_function, y_function_name) in enumerate(y_func):
xs = [item+1 for item in range(len(plot_xs))]
ys = []
for group_idx, model_group in enumerate(model_paths):
for model_path in model_group:
auroc_ = y_function(test_dir, model_path)
ys.append(auroc_)
if greyscale_support:
ax.plot(xs, ys, label=y_function_name,
linewidth=1, linestyle=linestyle_tuple[y_func_idx][1], marker="None")
else:
ax.plot(xs, ys, label=y_function_name,
linewidth=1.5, linestyle="solid", marker="None")
# ax.set_xscale('log')
# fig.canvas.draw()
ax.set_xticks(xs)
ax.set_xticklabels(plot_xs)
# ax.get_xaxis().set_major_formatter(mpl.ticker.ScalarFormatter()) # https://stackoverflow.com/questions/14530113/set-ticks-with-logarithmic-scale
# ax.set_xlim([0.5, 4.5])
ax.set_ylim([0.0, 1.0])
ax.set_xlabel('Fine annotation coefficient')
ax.set_ylabel('AUROC / Lesion Activation Precision')
# plt.title('AUC')
ax.legend(loc="lower right")
print('Saving to:', save_path)
plt.savefig(save_path, bbox_inches='tight', pad_inches=0, dpi=600)
if __name__=="__main__":
plot = 4 # choose which plot you want
if plot==1:
comparison_plot(test_dir='/usr/xtmp/mammo/Lo1136i_with_fa/validation/',
model_paths=[["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.0_random=4/100_9push0.4500.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.001_random=4/90_1push0.8786.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.002_random=4/80_9push0.5317.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.005_random=4/50_5push0.5000.pth"],
["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.0_random=4/100_9push0.9430.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.001_random=4/100_9push0.8982.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.002_random=4/100_9push0.3970.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.005_random=4/100_9push0.5317.pth"],
["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.0_random=4/100_9push0.9245.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.001_random=4/100_9push0.9146.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.002_random=4/100_9push0.8573.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.005_random=4/100_9push0.4471.pth"]
],
legend_labels=["maxpool (top 0.5%)",
"top 2%",
"top 5%"
],
plot_xs=[0,
0.001,
0.002,
0.005
],
save_path="./visualizations/auroc_over_fa.png",
y_func=[(auroc, 'AUROC')]
)
if plot==2:
comparison_plot(test_dir='/usr/xtmp/mammo/Lo1136i_with_fa/validation/',
model_paths=[["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.0_random=4/100_9push0.4500.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.001_random=4/90_1push0.8786.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.002_random=4/80_9push0.5317.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.005_random=4/50_5push0.5000.pth"],
["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.0_random=4/100_9push0.9430.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.001_random=4/100_9push0.8982.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.002_random=4/100_9push0.3970.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.005_random=4/100_9push0.5317.pth"],
["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.0_random=4/100_9push0.9245.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.001_random=4/100_9push0.9146.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.002_random=4/100_9push0.8573.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.005_random=4/100_9push0.4471.pth"]
],
legend_labels=["maxpool (top 0.5%)",
"top 2%",
"top 5%"
],
plot_xs=[0,
0.001,
0.002,
0.005
],
save_path="./visualizations/hp_over_fa.png",
y_func=[(hp, 'Lesion Activation Precision')]
)
if plot==3:
comparison_plot(test_dir='/usr/xtmp/mammo/Lo1136i_finer/by_margin/validation/',
model_paths=[["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.0_random=4/100_9push0.4500.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.001_random=4/90_1push0.8786.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.002_random=4/80_9push0.5317.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=1_fa=0.005_random=4/50_5push0.5000.pth"],
["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.0_random=4/100_9push0.9430.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.001_random=4/100_9push0.8982.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.002_random=4/100_9push0.3970.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=3_fa=0.005_random=4/100_9push0.5317.pth"],
["/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.0_random=4/100_9push0.9245.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.001_random=4/100_9push0.9146.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.002_random=4/100_9push0.8573.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0112_topkk=9_fa=0.005_random=4/100_9push0.4471.pth"]
],
legend_labels=["maxpool (top 0.5%)",
"top 2%",
"top 5%"
],
plot_xs=[0,
0.001,
0.002,
0.005
],
save_path="./visualizations/finehp_over_fa.png",
y_func=[(hp, 'Fine Activation Precision')]
)
if plot==4:
for target_class in range(3):
auroc_curves(test_dir='/usr/xtmp/mammo/Lo1136i/test/',
model_paths=["/usr/xtmp/mammo/saved_models/vgg16/0129_pushonall_topkk=9_fa=0.001_random=4/50_4push0.9546.pth",
"/usr/xtmp/mammo/saved_models/vgg16/0125_topkk=1_fa=0.0_random=4/50_5push0.9209.pth",
"/usr/xtmp/mammo/saved_models/vanilla/0125_vanilla_3margin_vgg16_latent512_baseline3_random=4/0.9384582045743842_at_epoch_136"
],
model_labels=["IAIA-BL (our model)",
"Original ProtoPNet",
"Blackbox VGG-16"
],
save_path=f'./visualizations/roc_curves_ep50_defaultcol_class{target_class}.png',
target_class=target_class,
greyscale_support=False,
kappa_vals=True,
confusion=True,
default_col=True
)