State-Value From Action Values

Problem

Compute the state value V(s) as the expectation of action values under a policy: V(s) = sum_a probs[a] * Q[a]. Empty inputs return 0.0.

Hint

This is just a dot product between the policy probabilities and the action values.

case 1

in: [[1,2,3],[0.2,0.3,0.5]]

expected: 2.3

case 2

in: [[5],[1]]

expected: 5

case 3

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

expected: 0.5

case 4

in: [[],[]]

expected: 0

Starter codePython

def state_value_from_q(Q, probs):
    # Your code here
    pass

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