2
$\begingroup$

I am using auto.arima for forecasting. I have more than one categorical variables having more than one level.

My questions are :

  1. Do I need to do dummy coding ?

  2. If I do dummy coding with my categorical variables, this will result into 20 variables. Is it good to have such number of variables in time series model.

  3. How can I do variable selections?

I would appreciate any kind of help.

$\endgroup$
3
$\begingroup$
  1. Yes, xreg must be a numeric vector or matrix, so you will need to code the dummy variables yourself.

  2. Provided you have enough data for each category to estimate all the coefficients, it should be ok.

  3. Minimize the AICc.

$\endgroup$
  • $\begingroup$ Thanks a lot for quick reply. So if I have a variable having entries "good","better","best","worst",then I would create 4 dummy variables or 3 dummy variables. Because in case of linear, it will be three. $\endgroup$ – Arushi Jul 17 '14 at 8:29

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.