You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When using auto ARIMA with no exogenous variables (r=0, xreg=NULL), in the case of differencing (d+D) equals 1 and a constant term being allowed, the auto ARIMA will find a model with exogenous variables. It will create helper variables from 1...N.
However, if you try to use auto_arima_predict with xreg=NULL it will inevitably fail, as the model will try to access exogenous variables, that are actually not present.
This should not be the case in my opinion, as this disallows using auto arima in a script. (Seg Faults appearing, even though everything is correct).
The text was updated successfully, but these errors were encountered:
When using auto ARIMA with no exogenous variables (r=0, xreg=NULL), in the case of differencing (d+D) equals 1 and a constant term being allowed, the auto ARIMA will find a model with exogenous variables. It will create helper variables from 1...N.
However, if you try to use auto_arima_predict with xreg=NULL it will inevitably fail, as the model will try to access exogenous variables, that are actually not present.
This should not be the case in my opinion, as this disallows using auto arima in a script. (Seg Faults appearing, even though everything is correct).
The text was updated successfully, but these errors were encountered: