Flatten 3-d numpy array

Question

How to flatten this:

b = np.array([
    [[1,2,3], [4,5,6], [7,8,9]],
    [[1,1,1],[2,2,2],[3,3,3]]
])

into:

c = np.array([
    [1,2,3,4,5,6,7,8,9],
    [1,1,1,2,2,2,3,3,3]
])

Niether of these work:

c = np.apply_along_axis(np.ndarray.flatten, 0, b)
c = np.apply_along_axis(np.ndarray.flatten, 0, b)

Just returns the same array.

It would be great to flatten this in place.

Miriam Farber · Accepted Answer · 2017-08-08 11:30:32Z

8

This will do the job:

c=b.reshape(len(b),-1)

Then c is

array([[1, 2, 3, 4, 5, 6, 7, 8, 9],
       [1, 1, 1, 2, 2, 2, 3, 3, 3]])

answered Aug 8, 2017 at 11:30

Miriam Farber

19.7k15 gold badges66 silver badges78 bronze badges

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

2 Comments

dokondr Over a year ago

What -1 does in reshape ?

Daniel F Over a year ago

It's a shortcut to automatically calculate the size of that dimension. Since b.size = 18 and len(b) = 2, the -1 is calculated to be 18 / 2 = 9

Ofer Sadan · Accepted Answer · 2017-08-08 11:30:52Z

2

You can completely flatten and then reshape:

c = b.flatten().reshape(b.shape[0],b.shape[1]*b.shape[2])

Output

array([[1, 2, 3, 4, 5, 6, 7, 8, 9],
       [1, 1, 1, 2, 2, 2, 3, 3, 3]])

answered Aug 8, 2017 at 11:30

Ofer Sadan

12k6 gold badges42 silver badges66 bronze badges

Comments

Tony · Accepted Answer · 2017-08-08 11:32:24Z

0

So you could always just use reshape:

b.reshape((2,9)) 
array([[1, 2, 3, 4, 5, 6, 7, 8, 9],
       [1, 1, 1, 2, 2, 2, 3, 3, 3]])

answered Aug 8, 2017 at 11:32

Tony

1,3002 gold badges14 silver badges38 bronze badges

Collectives™ on Stack Overflow

Flatten 3-d numpy array

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related