Reputation: 81
I have a numpy array of an images dataset of shape: (32, 32, 3, 7000) (height, width, channels, # of images)
but, I want the same dataset with shape: (7000, 32, 32, 3) (# of images, height, width, channels) without altering the images
what can I do? Thanks!
I used: np.moveaxis(X_train, -1, 0)
Upvotes: 0
Views: 58
Reputation: 568
Please use np.transpose function (https://numpy.org/doc/stable/reference/generated/numpy.transpose.html)
np.transpose(x, (3, 0, 1,2))
Upvotes: 1