Exercise

Holt-Winters with monthly data

In the video, you learned that the hw() function produces forecasts using the Holt-Winters method specific to whatever you set equal to the seasonal argument:

fc1 <- hw(aust, seasonal = "additive")
fc2 <- hw(aust, seasonal = "multiplicative")

Here, you will apply hw() to a10, the monthly sales of anti-diabetic drugs in Australia from 1991 to 2008. The data are available in your workspace.

Instructions

100 XP
  • Produce a time plot of the a10 data.
  • Produce forecasts for the next 3 years using hw() with multiplicative seasonality and save this to fc.
  • Do the residuals look like white noise? Check them using the appropriate function and set whitenoise to either TRUE or FALSE.
  • Plot a time plot of the forecasts.