pytorch，數據管理

# Dataset responsible for manipulating data for training as well as training tests.
class DatasetMNIST(torch.utils.data.Dataset):
    def __init__(self, data, transform=None):
        self.data = data
        self.transform = transform

    def __len__(self):
        return len(self.data)

    def __getitem__(self, index):
        item = self.data.iloc[index]

        image = item[1:].values.astype(np.uint8).reshape((28, 28))
        label = item[0]

        if self.transform is not None:
            image = self.transform(image)

        return image, label



BATCH_SIZE = 32
VALID_SIZE = 0.15 # percentage of data for validation

transform_train = transforms.Compose([
    transforms.ToPILImage(),
   # transforms.RandomRotation(0, 0.5),
    transforms.ToTensor(),
    transforms.Normalize(mean=(0.5,), std=(0.5,))
])

transform_valid = transforms.Compose([
    transforms.ToPILImage(),
    transforms.ToTensor(),
    transforms.Normalize(mean=(0.5,), std=(0.5,))
])

# Importing data that will be used for training and validation
dataset = pd.read_csv('./data/train/train.csv')

# Creating datasets for training and validation
train_data = DatasetMNIST(dataset, transform=transform_train)
valid_data = DatasetMNIST(dataset, transform=transform_valid)