EMA of Weights

Problem

Update an exponential moving average of the weights: ema_new = decay * ema + (1 - decay) * new_weights.

The signature is ema_update(ema_weights, new_weights, decay).

Hint

A decay near 1 keeps the old average while slowly tracking new weights.

case 1

in: [[1,2],[3,4],0.9]

expected: [1.2,2.2]

case 2

in: [[0],[1],0.5]

expected: [0.5]

case 3

in: [[1,1],[0,0],0.25]

expected: [0.25,0.25]

case 4

in: [[0,0],[1,1],1]

expected: [0,0]

Starter codePython

def ema_update(ema_weights, new_weights, decay):
    # Your code here
    pass

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