Rollout Quality Metric

Problem

Compute a normalized rollout quality score: mean(returns) / optimal_return. An empty rollout or a zero optimal return gives 0.0.

Hint

Normalizing by the optimal return makes scores comparable across tasks.

case 1

in: [[5,10],10]

expected: 0.75

case 2

in: [[2],10]

expected: 0.2

case 3

in: [[1],0]

expected: 0

case 4

in: [[],10]

expected: 0

Starter codePython

def rollout_quality_metric(returns, optimal_return):
    # Your code here
    pass

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