Average Reward Optimality

Problem

Compute the average reward (gain) of a policy from its stationary distribution and per-state rewards: sum of d[s] * rewards[s].

Hint

Under the average-reward criterion the gain is the stationary-weighted mean reward.

case 1

in: [[0.5,0.5],[2,4]]

expected: 3

case 2

in: [[1,0],[1,10]]

expected: 1

case 3

in: [[0.25,0.25,0.5],[1,2,3]]

expected: 2.25

Starter codePython

def average_reward(d, rewards):
    # Your code here
    pass

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