Prediction Interval Coverage

Problem

Compute the fraction of observations inside their prediction intervals [lower, upper] (inclusive).

Empty input returns 0.0.

Hint

Empirical coverage should roughly match the nominal confidence level.

case 1

in: [[[0,1],[2,3]],[0.5,2.5]]

expected: 1

case 2

in: [[[0,1],[2,3]],[0.5,4]]

expected: 0.5

case 3

in: [[[0,1]],[1.5]]

expected: 0

case 4

in: [[],[]]

expected: 0

Starter codePython

def interval_coverage(intervals, y_true):
    # Your code here
    pass

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