6
$\begingroup$

I'm looking for a Python alternative to R's ETS() from forecast().

It's my understanding that ETS() is one of the best performing forecasting program and I would like to use it. However I am extremely uncomfortable using R and I already have all my data cleaned and set up in Python.

Does anyone know any other packages similar to ETS?

$\endgroup$
3
  • $\begingroup$ It would be better to ask this over on StackOverflow. Technically this type of question is no longer allowed on CV, it is too programming focused. That said, last time I checked exponential smoothing/Forecasting are on the to-do list for statsmodels. $\endgroup$
    – Fraijo
    Jul 30, 2014 at 18:58
  • $\begingroup$ Sorry, I would have had I known. In fact I would have preferred to have asked on StackOverflow but I thought I would get told off and asked to post here haha! $\endgroup$ Jul 30, 2014 at 20:06
  • $\begingroup$ I think there is exponential smoothing in Pandas. It would probably be easiest to tag a SO question with Pandas (they are very active over there). $\endgroup$
    – Fraijo
    Jul 30, 2014 at 20:33

1 Answer 1

4
$\begingroup$

There is an open PR to add full ETS functionality to statsmodels here. I ran out of steam trying to code up all the heuristics for optimization that Hyndman suggests to make it less fragile. If someone wants to take up the torch here, I have some uncommitted code and thoughts on how to proceed.

$\endgroup$
1
  • 1
    $\begingroup$ This has been pull requested in. $\endgroup$
    – AZhao
    Oct 17, 2018 at 19:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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