RND Predictor Error

Problem

Compute the relative RND predictor error ||predicted - target|| / (||target|| + epsilon). An all-zero target with a nonzero prediction yields a large finite value thanks to epsilon.

Hint

Dividing by the target norm makes the error scale-free.

Test cases

case 1

in: [[1,1],[1,1]]

expected: 0

case 2

in: [[0,0],[3,4]]

expected: 0.9999999980000001

case 3

in: [[1,2],[2,2]]

expected: 0.35355338934327374

case 4

in: [[2,0],[0,1]]

expected: 2.2360679551391103

Starter codePython

import math
def rnd_predictor_error(predicted, target, epsilon=1e-8):
    # Your code here
    pass

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