Multiplicative Seasonal Indices

Problem

Estimate multiplicative seasonal indices using ratio-to-centered-moving-average. Build a centered moving average with window period (averaging two adjacent moving averages when period is even), then for each season position average series[t] / trend[t] over the positions where the trend is defined.

Return the period indices, or an empty list if the series is shorter than 2*period or some position has no observations.

Hint

For an even period the centered average gives half weight to the two end points.

Test cases

case 1

in: [[10,20,12,24,11,22],2]

expected: [0.6764705882352942,1.3212176283507495]

case 2

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

expected: [0.6666666666666666,1.3333333333333333]

case 3

in: [[5,10,15,20,25,30],2]

expected: [1,1]

case 4

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

expected: []

Starter codePython

def multiplicative_seasonal_indices(series, period):
    # Your code here
    pass

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