Forecast Bias Percentage

Problem

Compute forecast bias as a percentage of total actual demand: 100 * sum(forecast - actual) / sum(actual).

Positive values mean forecasts ran above actuals. Return 0.0 when the actual values sum to zero.

Hint

The percentage sign makes bias comparable across series of different scale.

case 1

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

expected: 0

case 2

in: [[50,50],[60,50]]

expected: 10

case 3

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

expected: 0

case 4

in: [[10,20,30],[12,22,28]]

expected: 3.3333333333333335

Starter codePython

def forecast_bias_percentage(actual, forecast):
    # Your code here
    pass

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