Advantage Normalization

Problem

Normalize a batch of advantages by its maximum absolute deviation.

Subtract the mean and divide by the largest |a - mean|. If every centered advantage is zero return zeros; an empty batch returns [].

Hint

Centering plus max-abs scaling bounds every normalized advantage in [-1, 1].

case 1

in: [[1,2,3]]

expected: [-1,0,1]

case 2

in: [[0,0,0]]

expected: [0,0,0]

case 3

in: [[]]

expected: []

case 4

in: [[-2,2]]

expected: [-1,1]

Starter codePython

def normalize_advantages_max(advantages):
    # Your code here
    pass

4 test cases · run your code against them in the editor