Efficient product of 1D array and 3D array along one dimension - NumPy

Question

I have two numpy arrays:

• A 1D array called t of shape (70L,) with element called let s say ti
• A 3D array called I with shape (70L, 1024L, 1024L), with ea

Answer 1

Divakar gives the best (most efficient) answers. For completeness' sake, one other way of doing it is by using Numpy's broadcasting capabilities:

(t[:,np.newaxis,np.newaxis]*I).sum(axis=0)

By adding two axes to t, broadcasting becomes possible and one can use regular Numpy operations, which for some might be more readable.