For my project, I'm choosing my initial process based on stationarity, seasonality, and constant variance.

I have two processes right now: one with log transformed data and differenced once at lag 1 and the other with Box-Cox transformed data and also differenced once at lag 1.

The acf and pacf plots for both are almost identical, and my problem right now is determining which process has better constant variance. The log process has a lower variance than the Box-Cox process does, but is there a test in R that can test which model has better constant variance?

  • 1
    $\begingroup$ Plot the residuals against time? $\endgroup$ – kjetil b halvorsen Nov 21 '17 at 9:29

Your Answer

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

Browse other questions tagged or ask your own question.