CODEWITHSUNDEEP
pythonarraysnumpy3dreshape
mit10000
mit10000

Reputation: 23

numpy 3D array reshape/flattern to 1D array based on row order

I have a 3D NumPy array

x = np.array([[[0, 0, 0, 0],
               [1, 1, 1, 1],
               [2, 2, 2, 2]],

              [[0, 0, 0, 0],
               [1, 1, 1, 1],
               [2, 2, 2, 2]]])

If I want to flatten the 3D array to 1D array by taking row n°0 of x[0], row n°0 of x[1], row n°1 of x[0], row n°1 of x[1], row n°2 of x[0], row n°2 of x[1], and get the following layout:

[0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2]

How can I achieve this? Tried to reshape, flatten, none of them worked.

Upvotes: 1

Views: 600

Answers (1)

Ivan
Ivan

Reputation: 40678

You can stack then flatten the array:

>>> np.stack(x, axis=1).flatten()
array([0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2])

Upvotes: 1

Related Questions