Example of Decomposition

A marketing analyst wants to predict sales of a golf driver. The analyst collects previous sales data to predict the sales of the product for the next 3 months.

  1. Open the sample data, GolfDriverSales0.MTW.
  2. Choose Stat > Time Series > Decomposition.
  3. In Variable, enter Sales.
  4. In Seasonal length, enter 12.
  5. Under Model Type, choose Additive.
  6. Select Generate forecasts. In Number of forecasts, enter 3.
  7. Click OK.

Interpret the results

The time series decomposition plot shows that the model underpredicts the data at the end of the series. This indicates that Decompostion does not adequately model the trend or the seasonal pattern. The analyst should try Winters' method to determine whether it provides a better fit to the data.

Time Series Decomposition for Sales

Method Model type Additive Model Data Sales Length 48 NMissing 0
Fitted Trend Equation Yt = 173.06 + 2.111×t
Seasonal Indices Period Index 1 -42.8472 2 -32.2639 3 -25.4306 4 -18.5972 5 -1.3056 6 47.3194 7 84.1111 8 30.5278 9 23.2361 10 4.1111 11 -22.8472 12 -46.0139
Accuracy Measures MAPE 7.265 MAD 16.621 MSD 518.119
Forecasts Period Forecast 49 233.672 50 246.367 51 255.312 52 264.256 53 283.659 54 334.396

Time Series Decomposition Plot for Sales

Decomposition - Component Analysis for Sales

Decomposition - Seasonal Analysis for Sales

By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy