Use the time series auscafe from the fpp2 package to answer the following questions. (a) Is this ts variance stable? How do you know? If not, what can you do to make it variance stable? (b) Is this time series stationary? How do you know? If not, how can you make it stationary?

(c) Use ggtedisplay( to plot your variance stable, stationary time series. Based on this output, what ARIMA model do you believe is a good starting point for estimating the data? How do you know? (Remem- ber to look for seasonal patterns in additional to regular patterns) (d) Look at nearby models to see if you can find a better model. What is the best model you found? How do you know it's the best model? (e) Are the residuals of your best model white noise? How do you know? (Note it is possible that you don't find a model that has white noise residuals. If that's the case move forward forecasting your best fitting model.) (f) Now estimate the time series using an ETS model. What ETS model did R choose? (g) Does the ARIMA model or the ETS model fit this time series better? How do you know? (h) Forecast your best model 1 year ahead.