How to reshape 3D numpy array <10,3,2> to 4D array <10,1,3,2>?

Question

I have the following numpy arrays:

import numpy as np

np.ones((10, 3, 2))

and I need to reshape it to <10,1,3,2>.

How can I do so?

Answer 1

Like others mentioned, you can .reshape it. An alternative is to use np.newaxis or np.expand_dims like this:

arr = np.ones((10, 3, 2))

arr1 = arr[:, np.newaxis, ...]
print(arr1.shape)  # (10, 1, 3, 2)

arr2 = np.expand_dims(arr, 1)
print(arr2.shape)  # (10, 1, 3, 2)

# check if the two arrays are equal
print(np.array_equal(arr1, arr2))  # True

Answer 2

x = np.ones((10, 3, 2))
# in place
x.shape = (10,1,3,2)

# new view
x.reshape((10,1,3,2))

# Add new axis
x[:, np.newaxis, :, :]

Answer 3

How about this:

np.ones((10, 3, 2)).reshape([10,1,3,2])