Discounted Return List

Problem

Compute the discounted return-to-go G_t for every timestep.

G_t = rewards[t] + gamma * G_{t+1}, with the value after the last step equal to 0. Return a list of the same length as rewards; an empty episode returns [].

Hint

Sweep backwards and carry one accumulated return.

Test cases

case 1

in: [[1,2,3],0.5]

expected: [2.75,3.5,3]

case 2

in: [[0,0,1],0.9]

expected: [0.81,0.9,1]

case 3

in: [[],0.5]

expected: []

case 4

in: [[4],1]

expected: [4]

Starter codePython

def discounted_returns_list(rewards, gamma):
    # Your code here
    pass

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