How To Add Numpy Arrays Of Different Shape

I am solving an Optimization problem, and after differentiation the equation becomes y + mu + b_i + c_j + (U.t,v) where shape of the variables are as follows: len(y) = 8992 mu = 2

Solution 1:

You can't add arrays of different dimensions elementwise, obviously, because it is not well defined.

What you can do though, is pad the smaller arrays with 0s to match the largest array, by:

def pad_with_zeros(b_00, b_11):
    b_max, b_min = (b_00, b_11) if b_00.shape[0] > b_11.shape[0] else (b_11, b_00)

    b_min_z = np.zeros_like(b_max)
    b_min_z[:b_min.shape[0],] = b_min
    return b_max, b_min_z

b_00 = np.array([2.2, 1.1, 4.4])
b_11 = np.array([1.2, 3.3])

b_max, b_min_z = pad_with_zeros(b_00, b_11)
b_max, b_min_z

and add them together when they are all of the same shape, the only question is if it is what you need?

Cheers.