Mean Absolute Percentage Error with Zero Handling

Problem

Compute the mean absolute percentage error in percent. Terms with an actual value of 0 are skipped and the mean is taken over the remaining samples; if all actuals are 0, return 0.0.

Empty input returns 0.0.

Hint

Divide the absolute error by the actual value and report a percentage.

Test cases

case 1

in: [[100,200],[110,190]]

expected: 7.500000000000001

case 2

in: [[10],[12]]

expected: 20

case 3

in: [[0,10],[0,5]]

expected: 50

case 4

in: [[],[]]

expected: 0

Starter codePython

def mean_absolute_percentage_error(y_true, y_pred):
    # Your code here
    pass

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