Just want to preface that I'm relatively new to working with time series forecasts.
I'm pulling together a forecast I just want to understand a bit about the difference between an STL forecast and a TSLM forecast. I know that TSLM allows you to put in an external regressor, but I have only included the trend and seasonal in my model. Just want to know why the two forecasts look different/have such different prediction intervals.
library(forecast) library(tidyverse) autoplot(forecast(stl(time_series, s.window="periodic", robust=TRUE), h=123, method="ets")) +theme(legend.position="none")
autoplot(forecast(a_tslm, h=123)) +theme(legend.position="none")