如何在加载 MNIST 数据集时修复“没有这样的文件或目录”错误

Question

我已经从 yann.lecun.com 下载了 MNIST 训练图像和标签并解压了它们。我正在尝试使用此代码加载它们-

from mlxtend.data import loadlocal_mnist

features,labels = loadlocal_mnist(

    images_path='/python/mnist-files/train-images-idx3-ubyte',

    labels_path='/python/mnist-files/train-labels-idx1-ubyte')

但是，我收到此错误 -

Traceback (most recent call last):

  File "generateClassifier.py", line 12, in <module>

    labels_path='/python/mnist-files/train-labels-idx1-ubyte')

  File "/home/inglorion/.local/lib/python3.6/site- 

packages/mlxtend/data/local_mnist.py", line 36, in loadlocal_mnist

    with open(labels_path, 'rb') as lbpath:

FileNotFoundError: [Errno 2] No such file or directory: '/python/mnist- 

files/train-labels-idx1-ubyte'

该目录确实存在，并且文件名是正确的。我怎样才能解决这个问题？

编辑：我试着用同样的python-mnist套餐-

from mnist import MNIST

mndata = MNIST('/python/mnist-files')

features,labels = mndata.load_training()

我遇到了类似的错误-

Traceback (most recent call last):

  File "generateClassifier.py", line 11, in <module>

    features,labels = mndata.load_training()

  File "/home/inglorion/.local/lib/python3.6/site-packages/mnist/loader.py", 

line 126, in load_training

    os.path.join(self.path, self.train_lbl_fname))

  File "/home/inglorion/.local/lib/python3.6/site-packages/mnist/loader.py", 

line 247, in load

    with self.opener(path_lbl, 'rb') as file:

  File "/home/inglorion/.local/lib/python3.6/site-packages/mnist/loader.py", 

line 239, in opener

    return open(path_fn, *args, **kwargs)

FileNotFoundError: [Errno 2] No such file or directory: '/python/mnist- 

files/train-labels-idx1-ubyte'

错误似乎只与训练标签文件有关；我尝试重新下载该文件，但这并没有解决它。

编辑 2：根据要求，这里是输出ls -l /python/mnist-files-

total 46156

-rw-r--r-- 1 inglorion inglorion 47040016 Jul 21  2000 train-images-idx3- 

ubyte

-rw-r--r-- 1 inglorion inglorion    60008 Jul 21  2000 train-labels-idx1- 

ubyte

-rw-r--r-- 1 inglorion inglorion   147970 Feb  8 22:43 wget-log

-rw-r--r-- 1 inglorion inglorion      682 Feb  9 14:40 wget-log.1

Answer 1

对我来说，它有助于将文件重命名为train-images.idx3-ubyte而不是train-images-idx3-ubyte（-之后images更改为 a&nbsp;.）。

Answer 2

a /和~之间有区别。默认情况下，os.dir('/')将在'/'. 我猜你的文件 python 就在'~'你的主目录中。你可以试试这个：from os.path import expanduserhome = expanduser("~")+'/python/mnist-files'mndata = MNIST(home)features,labels = mndata.load_training()如果有帮助，请告诉我。

Answer 3

您可以尝试使用此代码。from tensorflow.examples.tutorials.mnist import input_datamnist = input_data.read_data_sets('MNIST_data', validation_size=0)我已经执行了代码，它工作正常！！希望你会发现它有帮助。