Please use R studio $($R code$)$ to answer. If you load fpp$3$ library, all info for this question is available:

Consider Australian tobacco production from the $`$aus$\_$production$`$ tsibble included in the $`$fpp$3`$ package. The code below will help you isolate the time series for tobacco production from the dataset.

tobacco $<-$ aus$\_$production $|>$

select$(-$c$($Beer$,$Bricks,Cement,Electricity,Gas$))$

Create a training set that includes all data up to Q$4$ of $1998$ and use data from $1999$ on as your test set. The following code will help:

production$\_$train $<-$ tobacco $|>$

filter$($year$($Quarter$)<1999)$

Using the training data, fit the benchmark methods $($naive$,$ mean, seasonal naive, drift$)$ to the training set and forecast the periods covered by the test set. Which model performs best based on RMSE, MAE, MAPE and RMSSE?

a$.$

Mean

b$.$

Naive

c$.$

Seasonal naive

d$.$

Drift

For the "winning" forecasting model from the previous questions, calculate the p$-$value of the Ljung$-$Box test $($including $8$ lags$).$ Round your answer to $2$ decimals.