I have a 3d matrix like this

arange(16).reshape((4,2,2))
array([[[ 0,  1],
        [ 2,  3]],

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

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

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

and would like to stack them in grid format, ending up with

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

Is there a way of doing without explicitly hstacking (and/or vstacking) them or adding an extra dimension and reshaping (not sure this would work)?

Thanks,


In [27]: x = np.arange(16).reshape((4,2,2))

In [28]: x.reshape(2,2,2,2).swapaxes(1,2).reshape(4,-1)
Out[28]: 
array([[ 0,  1,  4,  5],
       [ 2,  3,  6,  7],
       [ 8,  9, 12, 13],
       [10, 11, 14, 15]])

I've posted more general functions for reshaping/unshaping arrays into blocks, here.