Upscaling a numpy array and evenly distributing values

Question

I have a 2 dimensional numpy array representing spatial data. I need to increase its resolution. I also need to evenly distribute values across the space. For example, a value of

5

would become:

1.25 1.25

1.25 1.25

I've looked at imresize but I don't think the interpolation options will work for this. Maybe there's another way? I'd like to avoid iterating rows and columns if I can. Any help would be greatly appreciated! Thank you!

Answer 1

Simply divide by the number of elements in the block defined by its height and width and then replicate/ expand. To replicate, we can use np.repeat or np.lib.stride_tricks.as_strided.

With np.repeat -

def upscale_repeat(a, h, w):
    return (a/float(h*w)).repeat(h, axis=0).repeat(h, axis=1)

With np.lib.stride_tricks.as_strided using tile_array -

def upscale_strided(a, h, w):
    return tile_array(a/float(h*w), h, w)

Sample run -

In [140]: a
Out[140]: 
array([[ 7,  6,  9],
       [ 6,  6, 10]])

In [141]: upscale_repeat(a, 2, 2)
Out[141]: 
array([[ 1.75,  1.75,  1.5 ,  1.5 ,  2.25,  2.25],
       [ 1.75,  1.75,  1.5 ,  1.5 ,  2.25,  2.25],
       [ 1.5 ,  1.5 ,  1.5 ,  1.5 ,  2.5 ,  2.5 ],
       [ 1.5 ,  1.5 ,  1.5 ,  1.5 ,  2.5 ,  2.5 ]])

In [142]: upscale_repeat(a, 2, 3)
Out[142]: 
array([[ 1.17,  1.17,  1.17,  1.  ,  1.  ,  1.  ,  1.5 ,  1.5 ,  1.5 ],
       [ 1.17,  1.17,  1.17,  1.  ,  1.  ,  1.  ,  1.5 ,  1.5 ,  1.5 ],
       [ 1.  ,  1.  ,  1.  ,  1.  ,  1.  ,  1.  ,  1.67,  1.67,  1.67],
       [ 1.  ,  1.  ,  1.  ,  1.  ,  1.  ,  1.  ,  1.67,  1.67,  1.67]])