Numpy Array summing with weights

Question

I have a two dimensional numpy array.

Each row is three elements long and is an integer 0-3. This represents a 6 bit integer, with each cell representing two bits, in order.

I'm trying to transform them into the full integer.

Eg

for i in range(len(myarray)):
  myarray[i] = myarray[i][0] * 16 + myarray[i][1] * 4 + myarray[i][2]

Eg I'm trying to sum each row but according to a certain weight vector of [16,4,1].

What is the most elegant way to do this? I'm thinking I have to do some sort of dot product followed by a sum, but I'm not 100% confident where to do the dot.

Answer 1

The dot product inclination is correct, and that includes the sum you need. So, to get the sum of the products of the elements of a target array and a set of weights:

>>> a = np.array([[0,1,2],[2,2,3]])
>>> a
array([[0, 1, 2],
       [2, 2, 3]])
>>> weights = np.array([16,4,2])
>>> np.dot(a,weights)
array([ 8, 46])