Combine arrays with specific shape using numpy

Question

I'm trying to mimic the combination of the aa and bb arrays shown below using zip() within a for block:

import numpy as np

aa = np.random.uniform(0., 1., (3, 566))
bb = np.random.uniform(0., 1., (3, 566))

cc = []
for a, b in list(zip(list(zip(*aa)), list(zip(*bb)))):
    cc.append(list(zip(*[a, b])))
cc = np.array(cc)

print(cc.shape)
(566, 3, 2)

I've tried vstack, hstack, column_stack, all of them combined with .reshape() to no avail. Obviously, not only the final shape should be equal, but the array itself too.

What is the proper numpy way to do this?

Another variation on stack: np.stack([aa,bb],1).transpose(2,0,1)`. This gets the (3,2) order right, and then transposes the 566 to the start. — hpaulj
– hpaulj, Commented Aug 30, 2018 at 19:39

akuiper · Accepted Answer · 2018-08-30 17:27:32Z

1

You can transpose aa and bb and then use numpy.dstack(stack arrays along third axis), i.e. np.dstack([aa.T, bb.T]):

np.dstack([aa.T, bb.T]).shape
# (566, 3, 2)

(np.dstack([aa.T, bb.T]) == cc).all()
# True

Or use np.stack(..., axis=-1):

(np.stack([aa.T, bb.T], axis=-1) == cc).all()
# True

answered Aug 30, 2018 at 17:27

akuiper

216k33 gold badges362 silver badges379 bronze badges

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

Collectives™ on Stack Overflow

Combine arrays with specific shape using numpy

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related