Why doesn't plt.imshow() display the image?

Question

I have this code, copied from a tutorial:

import numpy as np
np.random.seed(123)
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Convolution2D, MaxPooling2D
from keras.utils import np_utils
from keras.datasets import mnist
(X_train,y_train),(X_test,y_test) = mnist.load_data()
print X_train.shape
from matplotlib import pyplot as plt
plt.imshow(X_train[0])

No image was displayed. Why not?

There doesn't appear to be anything wrong with the backend of matplotlib on my computer. I tested that like so:

import matplotlib.pyplot as plt

data = [[0, 0.25], [0.5, 0.75]]

fig, ax = plt.subplots()
im = ax.imshow(data, cmap=plt.get_cmap('hot'), interpolation='nearest',
               vmin=0, vmax=1)
fig.colorbar(im)
plt.show()

and was able to produce an image: enter image description here

I also tried printing X_train[0] and it looks right.

Try adding `plt.show()` at the end of your code snippet. – Marcin Możejko Mar 15 '17 at 14:21 — Marcin Możejko, Mar 15 '17 at 14:21
This problem often comes up when you copy code from Jupyter – arame3333 Jul 16 '20 at 09:37 — arame3333, Jul 16 '20 at 09:37

score 244 · Accepted Answer · answered Mar 15 '17 at 14:25

244

The solution was as simple as adding plt.show() at the end of the code snippet:

import numpy as np
np.random.seed(123)
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Convolution2D, MaxPooling2D
from keras.utils import np_utils
from keras.datasets import mnist
(X_train,y_train),(X_test,y_test) = mnist.load_data()
print X_train.shape
from matplotlib import pyplot as plt
plt.imshow(X_train[0])
plt.show()

answered Mar 15 '17 at 14:25

Marcin Możejko

39,542
10
109
120

7

This just avoid answering WHY while getting 168 upvote...Strange – Kuo Nov 26 '20 at 11:39
1

March 2021. The same thing is still missing in the tutorial https://www.tensorflow.org/tutorials/images/transfer_learning – pisoir Mar 24 '21 at 12:04
@Kuo The original code doesn't draw the image because it doesn't contain the line of code that was added. Simple. The title, of course, implies an expectation that `.imshow` itself would do this; but the only way to answer "why doesn't calling `.imshow` draw the image?" **literally** is "because it's not defined to do anything like that". The natural follow-on question is "well why do they call it `imshow` if it doesn't show an image?", but that's off topic for Stack Overflow; we aren't going to speculate on the thought process of the matplotlib authors. – Karl Knechtel Jan 10 '23 at 02:22

score 52 · Answer 2 · edited Sep 13 '19 at 10:28

52

plt.imshow just finishes drawing a picture instead of printing it. If you want to print the picture, you just need to add plt.show.

edited Sep 13 '19 at 10:28

shaik moeed

5,300
1
18
54

answered Jan 05 '18 at 03:18

adaxi

521
4
5

27

From that point of view, it would be better and less confusing to rename the "imshow" to "imdraw". Don't you agree? – Aram Paronikyan Aug 01 '19 at 13:55
4

@AramParonikyan matplotlib is layers upon layers of confusion and hellish design. One day people will move onto a better library (cv2 is even more confusing) – qwr Apr 22 '22 at 03:58

score 11 · Answer 3 · edited Apr 28 '20 at 16:47

plt.imshow displays the image on the axes, but if you need to display multiple images you use show() to finish the figure. The next example shows two figures:

import numpy as np
from keras.datasets import mnist
(X_train,y_train),(X_test,y_test) = mnist.load_data()
from matplotlib import pyplot as plt
plt.imshow(X_train[0])
plt.show()
plt.imshow(X_train[1])
plt.show()

In Google Colab, if you comment out the show() method from previous example just a single image will display (the later one connected with X_train[1]).

Here is the content from the help:

plt.show(*args, **kw)
        Display a figure.
        When running in ipython with its pylab mode, display all
        figures and return to the ipython prompt.

        In non-interactive mode, display all figures and block until
        the figures have been closed; in interactive mode it has no
        effect unless figures were created prior to a change from
        non-interactive to interactive mode (not recommended).  In
        that case it displays the figures but does not block.

        A single experimental keyword argument, *block*, may be
        set to True or False to override the blocking behavior
        described above.



plt.imshow(X, cmap=None, norm=None, aspect=None, interpolation=None, alpha=None, vmin=None, vmax=None, origin=None, extent=None, shape=None, filternorm=1, filterrad=4.0, imlim=None, resample=None, url=None, hold=None, data=None, **kwargs)
        Display an image on the axes.

Parameters
----------
X : array_like, shape (n, m) or (n, m, 3) or (n, m, 4)
    Display the image in `X` to current axes.  `X` may be an
    array or a PIL image. If `X` is an array, it
    can have the following shapes and types:

    - MxN -- values to be mapped (float or int)
    - MxNx3 -- RGB (float or uint8)
    - MxNx4 -- RGBA (float or uint8)

    The value for each component of MxNx3 and MxNx4 float arrays
    should be in the range 0.0 to 1.0. MxN arrays are mapped
    to colors based on the `norm` (mapping scalar to scalar)
    and the `cmap` (mapping the normed scalar to a color).

Why doesn't plt.imshow() display the image?

3 Answers3

Linked

Related