Seasonal Naive with Drift

Problem

Forecast h steps ahead with a seasonal naive model plus drift: series[-season_length] + h * drift.

Return None if season_length is not positive, exceeds the series length, or h is negative.

Hint

The drift adds a constant amount for every extra step of horizon.

case 1

in: [[10,20,30,40],2,1,1]

expected: 31

case 2

in: [[10,20,30,40],2,1,3]

expected: 33

case 3

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

expected: 2

case 4

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

expected: null

Starter codePython

def seasonal_naive_drift(series, season_length, drift, h):
    # Your code here
    pass

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