26
votes

I know this question has been asked before but I can't seem to get mine to work.

import numpy as np


def load_dataset():
    def download(filename, source="http://yaan.lecun.com/exdb/mnist/"):
        print ("Downloading ",filename)
        import urllib
        urllib.urlretrieve(source+filename,filename)

    import gzip
    
    def load_mnist_images(filename):
        if not os.path.exists(filename):
            download(filename)
        with gzip.open(filename,"rb") as f:
            data=np.frombuffer(f.read(), np.uint8, offset=16)
            
            data = data.reshape(-1,1,28,28)
            
            return data/np.float32(256)

        def load_mnist_labels(filename):
            if not os.path.exists(filename):
                download(filename)
            with gzip.open(filename,"rb") as f:
                data = np.frombuffer(f.read(), np.uint8, offset=8)
            return data

        X_train = load_mnist_images("train-images-idx3-ubyte.gz")
        y_train = load_mnist_labels("train-labels-idx1-ubyte.gz")
        X_test = load_mnist_images("t10k-images-idx3-ubyte.gz")
        y_test = load_mnist_labels("t10k-labels-idx1-ubyte.gz")

        return X_train, y_train, X_test, y_test


X_train, y_train, X_test, y_test = load_dataset()


import matplotlib
matplotlib.use("TkAgg")

import matplotlib.pyplot as plt
plt.show(plt.imshow(X_train[3][0]))

This is the error I am getting:

Traceback (most recent call last):
  File "C:\Users\nehad\Desktop\Neha\Non-School\Python\Handwritten Digits 
Recognition.py", line 38, in <module>
    X_train, y_train, X_test, y_test = load_dataset()
TypeError: cannot unpack non-iterable NoneType object

I am new to machine learning. Did I just miss something simple? I am trying a Handwritten Digit Recognition project for my school Science Exhibition.

4
load_dataset() doesn't return anything, so by default, it returns None, which you can't unpack. Maybe you're thinking about doing return load_mnist_images("""some-filename""");? But then you'd need a file name.TrebledJ

4 Answers

19
votes

I think your X_train, y_train, X_test, y_test are defined inside your load_mnist_imagesfunction, and are thus not defined for your load_dataset function.

You should de-indent your 5 lines from X_train = ... to return X_train, ... and your code might work better then.

16
votes

You get this error when you perform a multiple assignment to None (which is of NoneType). For instance:

X_train, y_train, X_test, y_test = None

TypeError: cannot unpack non-iterable NoneType object

So if you get this, the error is most likely that the right-hand part of the assignment is not what you expected (it's nothing).

2
votes

You get that error because load_dataset() returns None.

run this and you get the same error:

X_train, y_train, X_test, y_test = None
1
votes

Splitting/breaking multiple assignment to None resolves this issue.

e.g. This DOESN'T work:

test_set_raw, test_set_transformed, train_set_raw, train_set_transformed = None

This DOES work (no error):

test_set_raw = None
test_set_transformed = None
train_set_raw = None
train_set_transformed = None