#环境准备
import numpy as np              # numpy数组库
import math                     # 数学运算库
import matplotlib.pyplot as plt # 画图库

import torch             # torch基础库
import torchvision.datasets as dataset  #公开数据集的下载和管理
import torchvision.transforms as transforms  #公开数据集的预处理库,格式转换
import torchvision.utils as utils 
import torch.utils.data as data_utils  #对数据集进行分批加载的工具集

print("Hello World")
print(torch.__version__)
print(torch.cuda.is_available())

3.3 数据集下载与导入

如果本地没有数据集，会自动远程下载

#2-1 准备数据集
train_data = dataset.CIFAR10 (root = "cifar10",
                           train = True,
                           transform = transforms.ToTensor(),
                           download = True)

#2-1 准备数据集
test_data = dataset.MNIST(root = "cifar10",
                           train = False,
                           transform = transforms.ToTensor(),
                           download = True)

print(train_data)
print("size=", len(train_data))
print("")
print(test_data)
print("size=", len(test_data))

Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to cifar10\\cifar-10-python.tar.gz
Failed download. Trying https -> http instead. Downloading http://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to cifar10\\cifar-10-python.tar.gz

100.0%

Extracting cifar10\\cifar-10-python.tar.gz to cifar10

1.1%

Downloading http://183.207.33.38:9011/yann.lecun.com/c3pr90ntc0td/exdb/mnist/train-images-idx3-ubyte.gz to cifar10\\MNIST\\raw\\train-images-idx3-ubyte.gz

100.0%

Extracting cifar10\\MNIST\\raw\\train-images-idx3-ubyte.gz to cifar10\\MNIST\\raw

102.8%

Downloading http://183.207.33.42:9011/yann.lecun.com/c3pr90ntc0td/exdb/mnist/train-labels-idx1-ubyte.gz to cifar10\\MNIST\\raw\\train-labels-idx1-ubyte.gz
Extracting cifar10\\MNIST\\raw\\train-labels-idx1-ubyte.gz to cifar10\\MNIST\\raw

5.0%

Downloading http://183.207.33.38:9011/yann.lecun.com/c3pr90ntc0td/exdb/mnist/t10k-images-idx3-ubyte.gz to cifar10\\MNIST\\raw\\t10k-images-idx3-ubyte.gz

100.0%

Extracting cifar10\\MNIST\\raw\\t10k-images-idx3-ubyte.gz to cifar10\\MNIST\\raw
Downloading http://183.207.33.42:9011/yann.lecun.com/c3pr90ntc0td/exdb/mnist/t10k-labels-idx1-ubyte.gz to cifar10\\MNIST\\raw\\t10k-labels-idx1-ubyte.gz

112.7%

Extracting cifar10\\MNIST\\raw\\t10k-labels-idx1-ubyte.gz to cifar10\\MNIST\\raw
Processing...

C:\\ProgramData\\Anaconda3\\envs\\pytorch1.8_py3.8\\lib\\site-packages\\torchvision\\datasets\\mnist.py:479: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at  ..\\torch\\csrc\\utils\\tensor_numpy.cpp:143.)
  return torch.from_numpy(parsed.astype(m[2], copy=False)).view(*s)

Done!
Dataset CIFAR10
    Number of datapoints: 50000
    Root location: cifar10
    Split: Train
    StandardTransform
Transform: ToTensor()
size= 50000

Dataset MNIST
    Number of datapoints: 10000
    Root location: cifar10
    Split: Test
    StandardTransform
Transform: ToTensor()
size= 10000

3.4 显示单张样本图片

#原图不叠加噪声
#获取一张图片数据
print("原始Pytorch图片")
image, label = train_data[2]
print("torch image shape:", image.shape)
print("torch image label:", label)

print("\\n通道转换后的Numpy图片")
image = image.numpy().transpose(1,2,0)  #交换维度，从GBR换成RGB
print("numpy image shape:", image.shape)
print("numpy image label:", label)

plt.imshow(image)
plt.show()

3.5 启动loader对象

# 批量数据读取
train_loader = data_utils.DataLoader(dataset = train_data,
                                  batch_size = 8,
                                  shuffle = True)

test_loader = data_utils.DataLoader(dataset = test_data,
                                  batch_size = 8,
                                  shuffle = True)

print(train_loader)
print(test_loader)
print(len(train_data), len(train_data)/8)
print(len(test_data),  len(test_data)/8)

<torch.utils.data.dataloader.DataLoader object at 0x0000012C3BBA85E0>
<torch.utils.data.dataloader.DataLoader object at 0x0000012C3BBA8F40>
50000 6250.0
10000 1250.0

3.6 显示批量图片

pytorch对图片的格式定义与Numpy对图片的格式定义是不一样的。

因此需要通过transpose（）进行维度的变换。

#显示一个batch图片
print("获取一个batch组图片")
imgs, labels = next(iter(train_loader))
print(imgs.shape)
print(labels.shape)
print(labels.size()[0])

print("\\n合并成一张三通道灰度图片")
images = utils.make_grid(imgs, nrow = 4)
print(images.shape)
print(labels.shape)

print("\\n转换成imshow格式")
images = images.numpy().transpose(1,2,0) 
print(images.shape)
print(labels.shape)

print("\\n显示图片")
plt.imshow(images)
plt.show()

获取一个batch组图片
torch.Size([8, 3, 32, 32])
torch.Size([8])
8

合并成一张三通道灰度图片
torch.Size([3, 70, 138])
torch.Size([8])

转换成imshow格式
(70, 138, 3)
torch.Size([8])

显示图片

作者主页(文火冰糖的硅基工坊)：文火冰糖（王文兵）的博客_文火冰糖的硅基工坊_CSDN博客

本文网址：https://blog.csdn.net/HiWangWenBing/article/details/121055970

以上是关于[Pytorch系列-33]：数据集 - torchvision与CIFAR10详解的主要内容，如果未能解决你的问题，请参考以下文章

PyTorch: 张量的拼接切分索引

[Pytorch系列-36]：数据集 - torchvision与ImageNet数据集

[Pytorch系列-37]：工具集 - torchvision库详解(数据集数据预处理模型)

如何创建图神经网络数据集？（pytorch 几何）

[Pytorch系列-45]：卷积神经网络 - 用GPU训练AlexNet+CIFAR10数据集

[Pytorch系列-46]：卷积神经网络 - 用GPU训练ResNet+CIFAR100数据集