Hi All. I am currently working on a forecasting model using Time Gen 1, and I have a question regarding the use of exogenous variables for historical data. For example, external sales are correlated with internal purchases. I would like to leverage this historical correlation in my model to improve forecast accuracy. Could you please advise on the best way to incorporate these historical exogenous variables into the Time Gen 1 model? thanks.

Hello! Take a look at this tutorial to learn how to work with exogenous variables and TimeGPT.

Thanks, Marco. That documentation only specifies cases with known future variables. Is there a way to include a correlation from the historical dates?

In the documentation, the variables are also historical features. However, to use them, you need to provide their future values over the forecast horizon. This is also covered in the tutorial.

thanks.