search

AutoViML / Auto_TS

Automatically build ARIMA, SARIMAX, VAR, FB Prophet and XGBoost Models on Time Series data sets with a Single Line of Code. Created by Ram Seshadri. Collaborators welcome.
Apache License 2.0
729 stars 114 forks source link

Update README file and init changes syntax a little. #21

Closed AutoViML closed 3 years ago

AutoViML commented 3 years ago

Please take a look and let me know if we can push this.

AutoViML commented 3 years ago

Nikhil: Ok I have modified it based on my understanding - please review and let me know. If not, you can change it too. Thx Ram

On Sun, Dec 20, 2020 at 11:02 AM Nikhil Gupta notifications@github.com wrote:

@ngupta23 commented on this pull request.

In auto_ts/init.py https://github.com/AutoViML/Auto_TS/pull/21#discussion_r546395704:

@@ -136,10 +131,10 @@ def init( and Scikit-Learn ML. It will automatically select the BEST model which gives best score specified. ##################################################################################################### """

  • lag_period = 5

Hi @AutoViML https://github.com/AutoViML

It is used to decide the cross validation steps (in fit), so it is needed. The idea is that you provide it once when you initialize the object and then you don't have to specify it again during fit or predict.

The only time you need to provide it during predict is when you need the forecast horizon to be different than the one used for training (which is rare). I just added that extra flexibility in there in case some one needed it, but by default the forecast_period is not needed to be specified in predict.

Could you please revert it back. If you are concerned that it needs a default value, you can add it to the init call and set it to 1 (instead of 5).

Thanks!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/AutoViML/Auto_TS/pull/21#discussion_r546395704, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMKBH6DVZRTM7KX6AMR5HTDSVYN2VANCNFSM4VB65ZBQ .