0

I am getting started using pmdarima and trying to resolve some confusion with m. I m appending to this related question and hoping you can clarify, I am having 5-minute intervals data. The seasonality is one day in that case, m = 24*12. When I try using auto_arima for 6 month of data with m=24x12, the process gets killed. How can I make this computationally feasible?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
AI bits
  • 1
  • 2
    See https://robjhyndman.com/hyndsight/longseasonality/ and [threads](https://stats.stackexchange.com/search?q=robjhyndman.com%2Fhyndsight%2Flongseasonality) that mention it. – Richard Hardy Feb 12 '21 at 11:56
  • @RichardHardy: do you want to post your comment(s) as an answer? [Better to have a short answer than no answer at all.](https://stats.meta.stackexchange.com/a/5326/1352) Anyone who has a better answer can post it. – Stephan Kolassa Feb 12 '21 at 15:21
  • @StephanKolassa, I would rather identify a duplicate and mark this thread as such. Just do not have the time for it now. – Richard Hardy Feb 12 '21 at 15:23
  • 1
    @RichardHardy: good point. I proposed a duplicate that is pretty close, at least in terms of the resolution, if not in terms of the original question. – Stephan Kolassa Feb 12 '21 at 15:27
  • @StephanKolassa, thank you. – Richard Hardy Feb 12 '21 at 16:18

0 Answers0