def generate_features()

in data/dataset.py [0:0]

• 19 lines of code
• 2 McCabe index (conditional complexity)

    def generate_features(self, out_file):

        data = self.train_data + self.val_data + self.test_data
        transform = imagenet_transform('test')
        feat_extractor = tmodels.resnet18(pretrained=True)
        feat_extractor.fc = nn.Sequential()
        feat_extractor.eval().cuda()

        image_feats = []
        image_files = []
        for chunk in tqdm.tqdm(
                utils.chunks(data, 512), total=len(data) // 512):
            files, attrs, objs = zip(*chunk)
            imgs = list(map(self.loader, files))
            imgs = list(map(transform, imgs))
            feats = feat_extractor(torch.stack(imgs, 0).cuda())
            image_feats.append(feats.data.cpu())
            image_files += files
        image_feats = torch.cat(image_feats, 0)
        print('features for %d images generated' % (len(image_files)))

        torch.save({'features': image_feats, 'files': image_files}, out_file)