I really haven't seen the answer to this question here. I have a dataset with the seasonally adjusted brazilian GDP and I'm trying to model it with an ARIMA(1,1,1) model.
However, my residuals mean are not equal to zero. Can you please help me?
dadospib$pib = ts(dadospib$PIB_pm, start = c(1996,1), end = c(2017,4), frequency = 4) #it's a quarterly data
pib %>% arima(order = c(1,1,1)) %>% residuals %>% mean
[1] 0.3065052