chexpert/test.py at master · gokriznastic/chexpert · 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
import os

import matplotlib.pyplot as plt
import numpy as np
import sklearn.metrics as metrics
from sklearn.metrics import accuracy_score
from sklearn.metrics.ranking import roc_auc_score

import torch
import torch.backends.cudnn as cudnn
from dataloader import CheXpertDataset, getImagesLabels
from network import DenseNet121
from torch.utils.data import DataLoader
from utils import categorical
from viz import plot_ROC


def computeAUROC (pred, gt, num_classes):
    AUROC = []

    np_pred = pred.cpu().numpy()
    np_gt = gt.cpu().numpy()

    for i in range(num_classes):
        try:
            AUROC.append(roc_auc_score(np_gt[:, i], np_pred[:, i]))
        except ValueError:
            pass
    return AUROC

def test(test_loader, model, checkpoint, device, num_classes, class_names):
        cudnn.benchmark = True
        threshold = 0.45

        if checkpoint is not None:
            checkpoint = torch.load(checkpoint)
            model.load_state_dict(checkpoint['model_state_dict'])

        gt = torch.FloatTensor().to(device)
        pred = torch.FloatTensor().to(device)


        model.eval()
        with torch.no_grad():
            for i, (data, target) in enumerate(test_loader):

                target = target.to(device)
                gt = torch.cat((gt, target), 0).to(device)

                out = model(data)
                # print(out)

                # zeros = torch.zeros_like(target)
                # ones = torch.ones_like(target)
                # out = torch.where(out>threshold, ones, zeros)
                # out = torch.from_numpy(categorical(out.cpu().numpy())).float().to(device)
                pred = torch.cat((pred, out), 0)

        auroc_individual = computeAUROC(pred, gt, num_classes)
        auroc_mean = np.array(auroc_individual).mean()

        print ('\nmean AUROC:\t\t{}\n'.format(auroc_mean))

        for i in range(len(auroc_individual)):
            print (class_names[i], '\t\t', auroc_individual[i])

        return pred, gt

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

num_classes = 5
class_names = ['Atelectasis', 'Cardiomegaly', 'Consolidation', 'Edema', 'Pleural Effusion']

val_csv = 'CheXpert-v1.0-small/valid.csv'
# test_images, test_labels = getImagesLabels(val_csv, 'ones')
test_images, test_labels = getImagesLabels(val_csv, policy='both')


print('Testing on {} images\n'.format(len(test_images)))

test_dataset = CheXpertDataset(test_images, test_labels, test=True)
test_loader = DataLoader(test_dataset, batch_size=16, shuffle=False,  num_workers=24, pin_memory=True)

model = DenseNet121(num_classes, test=True)
model = torch.nn.DataParallel(model)

pred, gt = test(test_loader, model, 'checkpoints/22-10-2019-18:53:28_model.pt', device, num_classes, class_names) # U-ignore
plot_ROC(pred, gt, num_classes, class_names)

pred[pred>=0.5] = 1
pred[pred<0.5] = 0

print('\nAccuracy:\n')
for i in range(num_classes):
    acc = accuracy_score(gt.cpu()[:,i], pred.cpu()[:,i])
    print (class_names[i], '\t\t', acc)


# pred_zeros, gt_zeros = test(test_loader, model, '10-10-2019-15:33:21_model.pt', device, num_classes, class_names) # U-zeros
# pred_ones, gt_ones = test(test_loader, model, '10-10-2019-14:39:11_model.pt', device, num_classes, class_names) # U-ones

# fig_size = plt.rcParams["figure.figsize"]
# fig_size[0] = 20
# fig_size[1] = 4
# plt.rcParams["figure.figsize"] = fig_size
# plt.tight_layout()

# for i in range(num_classes):
#     fpr0, tpr0, threshold0 = metrics.roc_curve(gt_zeros.cpu()[:,i], pred_zeros.cpu()[:,i])
#     roc_auc0 = metrics.auc(fpr0, tpr0)
#     fpr1, tpr1, threshold1 = metrics.roc_curve(gt_ones.cpu()[:,i], pred_ones.cpu()[:,i])
#     roc_auc1 = metrics.auc(fpr1, tpr1)

#     plt.subplot(1, 5, i+1)

#     plt.title('ROC for: ' + class_names[i])
#     plt.plot(fpr0, tpr0, label = 'U-zeros: AUC = %0.2f' % roc_auc0)
#     plt.plot(fpr1, tpr1, label = 'U-ones: AUC = %0.2f' % roc_auc1)

#     plt.legend(loc = 'lower right')
#     plt.plot([0, 1], [0, 1],'r--')
#     plt.xlim([-0.01, 1])
#     plt.ylim([0, 1.01])
#     plt.ylabel('True Positive Rate')
#     plt.xlabel('False Positive Rate')

# # mng = plt.get_current_fig_manager()
# # mng.resize(*mng.window.maxsize())

# plt.savefig("ROC.png", dpi=1000)
# plt.show()