Naive Forecast MAE

Problem

Evaluate the one-step naive forecast on the last h observations of a series, where the forecast for each point is simply the previous observation. Return the mean absolute error over those h steps.

Return 0.0 when h is not positive or when h is larger than the number of available one-step errors, n - 1.

Hint

Each error is just the absolute change between consecutive observations.

Test cases

case 1

in: [[1,2,4,7],3]

expected: 2

case 2

in: [[5,5,5],2]

expected: 0

case 3

in: [[1,2],1]

expected: 1

case 4

in: [[1,2,3],5]

expected: 0

Starter codePython

def naive_holdout_mae(series, h):
    # Your code here
    pass

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