The plot shows that the series has seasonality, but the conclusion of adftest if there is no need to take the difference.

In this case, What is the best way to remove the seasonality in the model?

  • $\begingroup$ You need to supply more information about your data including the plot of the time series. $\endgroup$ – Michael R. Chernick Apr 18 '18 at 17:03

ADF test has little to do with seasonality. At best it can tell you whether there is a unit root. Differencing is one of the ways to treat unit roots and mean nonstationarity. It is not a go to tool for seasonality

Depending what kind of seasonality you have simple tools such as Holt_winters filter or Census Bureau's X-13 can be used

  • $\begingroup$ Thus how to deal with the seasonality... $\endgroup$ – JJJ.GGG Apr 18 '18 at 16:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.