尝试使用pytorch的Dataset解压缩类中的numpy数组

Question

在Pytorch教程中，在运行代码时:

import torch
import torchvision
from torch.utils.data import Dataset
import numpy as np
class WineDataset(Dataset):

def __init__(self, transform=None):
# data loading
xy = np.loadtxt('./data/wine/wine.csv', delimiter = ",", dtype = np.float32, skiprows = 1)
self.n_samples = xy.shape[0]

# note that we do not convert to tensor here
self.x = xy[:,1:]
self.y = xy[:,[0]] 

self.transform = transform

def __getitem__(self, index):
#dataset[0]
sample = self.x[index], self.y[index]

if self.transform:
sample = self.transform(sample)

return sample

def __len__(self):
# len(dataset)
return self.n_samples
class ToTensor:
def __call__(self,sample):
inputs, targets = sample
return torch.from_numpy(inputs), torch.from_numpy(targets)

dataset = WineDataset(transform=ToTensor())
first_data = dataset[0]
features, labels = first_data
print(type(features), type(labels))

print(type(features), type(labels))

的输出表明变量features和labels确实是张量。但是如果是dataset = WineDataset(transform=None)，那么print(type(features), type(labels))会抛出以下错误:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-99-17180fb44f1a> in <module>
33 dataset = WineDataset(transform=None)
34 first_data = dataset[0]
---> 35 features, labels = first_data
36 print(type(features), type(labels))
TypeError: cannot unpack non-iterable NoneType object

根据本教程，在本例中，print的输出应该表明features和labels是numpy数组。

为什么我得到这个错误呢?提前感谢

Answer 1

我认为您应该尝试在变量上使用DataLoader()函数。一旦你完成了这些，你就可以遍历你的数据集并提取特征和标签。

您可以尝试以下操作:

dataset = WineDataset(transform=None)
data = DataLoader(dataset, batch_size=1, shuffle=True)
features, labels = next(iter(data))