How to flatten a numpy object array with different shaped arrays?

Question

I've got an array a

>> a = np.array([np.ones((4,5)), np.arange(6), np.arange(20).reshape((2,2,5))])
>> a
array([array([[1., 1., 1., 1., 1.],
       [1., 1., 1., 1., 1.],
       [1., 1., 1., 1., 1.],
       [1., 1., 1., 1., 1.]]),
       array([0, 1, 2, 3, 4, 5]),
       array([[[ 0,  1,  2,  3,  4],
        [ 5,  6,  7,  8,  9]],

       [[10, 11, 12, 13, 14],
        [15, 16, 17, 18, 19]]])], dtype=object)

and I want to flatten it to get something like array(1, 1, 1, ... 17, 18, 19). How do I do this in the most efficient way?

Divakar · Accepted Answer · 2019-05-01 06:17:06Z

1

Flatten and concatenate/stack-horizontally -

In [36]: np.concatenate([np.ravel(i) for i in a])
Out[36]: 
array([ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,
        1.,  1.,  1.,  1.,  1.,  1.,  1.,  0.,  1.,  2.,  3.,  4.,  5.,
        0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.,  9., 10., 11., 12.,
       13., 14., 15., 16., 17., 18., 19.])

In [42]: np.hstack([np.ravel(i) for i in a])
Out[42]: 
array([ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,
        1.,  1.,  1.,  1.,  1.,  1.,  1.,  0.,  1.,  2.,  3.,  4.,  5.,
        0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.,  9., 10., 11., 12.,
       13., 14., 15., 16., 17., 18., 19.])

Or with map -

np.concatenate(list(map(np.ravel,a)))
np.hstack(list(map(np.ravel,a)))

edited May 1, 2019 at 6:17

answered May 1, 2019 at 6:06

Divakar

222k19 gold badges273 silver badges374 bronze badges

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

2 Comments

Shree Singhi Over a year ago

Which method is faster?

Divakar Over a year ago

@ѕняєєѕιиgнι I would think np.concatenate([np.ravel(i) for i in a]) one. But, do test it out at your end.

Collectives™ on Stack Overflow

How to flatten a numpy object array with different shaped arrays?

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related