How can I save 3D array results to a 4D array in Python/numpy?

Question

I am reading the information about 32 X 32 RGB images. So it is a 3D array with shape (32,32,3) third dimension holds the color R, G and B

Now, I want to read 50 such images and make a array of these images. So I decide to make a 4D array which will have dimension (50, 32, 32, 3) here 50 in first dimension is the number of images and the second, third and fourth dimension are the dimension of the image which are (32, 32, 3)

I tried doing it using concatenate but I get errors. Is there any way to do this?

jakevdp · Accepted Answer · 2015-11-03 12:10:17Z

13

You need to add an axis before concatenation, e.g.

import numpy as np
arrs = [np.random.random((32, 32, 3))
        for i in range(50)]

res = np.concatenate([arr[np.newaxis] for arr in arrs])
res.shape
# (50, 32, 32, 3)

Edit: alternatively, in this case you can simply call np.array on your list of arrays:

res = np.array(arrs)
res.shape
# (50, 32, 32, 3)

answered Nov 3, 2015 at 12:10

jakevdp

88.5k15 gold badges140 silver badges177 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

user3051460 Over a year ago

Great job. How to achieve (32, 32, 3, 50)

jakevdp Over a year ago

@user3051460: try np.stack(arrs, axis=3)

Shaoor Ahmad Qureshi · Accepted Answer · 2018-03-07 10:46:59Z

2

more easier would be to just add another axis in your Images and append them

    old_image = np.ones((100,100,3))
    new_image = np.ones((100,100,3,1))    
    # lets jsut say you have two Images 
    old_image = np.reshape(old_image , (100,100,3,1))
    new_image = np.reshape(new_image , (100,100,3,1))
    directory = np.append( new_image , old_image , axis = 3)

answered Mar 7, 2018 at 10:46

Shaoor Ahmad Qureshi

898 bronze badges

Collectives™ on Stack Overflow

How can I save 3D array results to a 4D array in Python/numpy?

2 Answers 2

2 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related