Reshape 4d numpy array to 2d array while preserving array locations

Question

I have a 4 dimensional numpy array of shape (N, N, Q, Q). So given a row and column index (i, j), mat[i,j] is a QxQ matrix. I want to reshape this array to shape (N*Q, N*Q) such that

array([[[[ 0,  1],
         [ 2,  3]],

        [[ 4,  5],
         [ 6,  7]]],


       [[[ 8,  9],
         [10, 11]],

        [[12, 13],
         [14, 15]]]])

goes to

array([[  0.,   1.,   4.,   5.],
       [  2.,   3.,   6.,   7.],
       [  8.,   9.,  12.,  13.],
       [ 10.,  11.,  14.,  15.]])

You can see that mat[0,0] goes to new_mat[0:2, 0:2]. Currently mat.reshape(N*Q, N*Q) takes mat[0,0] to new_mat[0:4, 0] (which is what I do not want). How can I use reshape or rollaxis or something similar to reshape this array? I eventually want to plot it with imshow, am currently stuck. I figure it's easy to do, I just haven't yet figured it out.

wflynny · Accepted Answer · 2013-10-29 20:19:51Z

9

Nevermind, I figured it out. np.swapaxes(1, 2) was the missing piece I needed.

The answer is just to do mat.swapaxes(1, 2).reshape(N*Q, N*Q).

Feel foolish for posting without attempting to figure it out myself for too long, but I'll leave it up so others can benefit from it.

answered Oct 29, 2013 at 20:19

wflynny

18.6k6 gold badges50 silver badges69 bronze badges

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

Collectives™ on Stack Overflow

Reshape 4d numpy array to 2d array while preserving array locations

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related