Delayed Policy Update Counter

Problem

Return how many policy updates are now due under delayed policy updates: step // delay minus updates_done, floored at 0. A non-positive delay returns 0.

Hint

TD3 updates the actor and target networks less often than the critic.

Test cases

case 1

in: [10,2,3]

expected: 2

case 2

in: [10,2,5]

expected: 0

case 3

in: [1,2,0]

expected: 0

case 4

in: [8,0,0]

expected: 0

Starter codePython

def delayed_policy_updates(step, delay, updates_done):
    # Your code here
    pass

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