Question: Does nnetar_reg use the outcome variable as predictor?

[MxNl]

Hi,

I want to use the nnetar_reg() for timeseries modelling. After reading the documentation of forecast::nnetar(), I was a bit confused with how the definition of roles in a recipe translates to roles in the nnetar model. Because nnetar() uses lagged versions of y (the outcome) as predictor. But if I define the outcome in a recipe, I usually assume that it will not be used as predictor. If so, I could of course also just use a mlp with lagged predictors. Could you please help me with this? Thanks a lot!