Hello
Why do we get NAs in residual and fitted values of forecast() function. Below is an example
fancy <- scan("http://robjhyndman.com/tsdldata/data/fancy.dat")
souvenir_series = ts(data =log(fancy),start = 1987, frequency = 12)
souvenir_series_forecasts = HoltWinters(souvenir_series)
souvenir_series_forecasts2 = forecast(souvenir_series_forecasts,h = 48)
souvenir_series_forecasts2$residuals
