I have learnt to create the LightGBM models with multiple features and target, but in ALL the examples I have seen thus far they only show how to do a "predict" on the x_test_df. What I am looking for is a way to use my current data that I have up until the last day, and then make a forecast n-steps into the future, like is done with ARIMA etc. Everything from what I can get is that it will only predict the dates that is in the df that is entered in the "predict".
Do I need to create an empty df with dates into the future, but then what about the inputs?
What am I missing, or am I just blind in missing this in the documentation?