date	sales
2000Q1	6307
2000Q2	6053
2000Q3	6118
2000Q4	7433
2001Q1	5993
2001Q2	5960
2001Q3	5786
2001Q4	7016
2002Q1	5856
2002Q2	5795
2002Q3	6662
2002Q4	7170
2003Q1	6500
2003Q2	6377
2003Q3	7014
2003Q4	8549
2004Q1	7189
2004Q2	7471
2004Q3	7543
2004Q4	8666
2005Q1	7829
2005Q2	7715
2005Q3	7734
2005Q4	8854
2006Q1	7767
2006Q2	8474
2006Q3	8652
2006Q4	9581
2007Q1	7954
2007Q2	9045
2007Q3	8930
2007Q4	10452
2008Q1	8710
2008Q2	9236
2008Q3	9445
2008Q4	9599
2009Q1	8087
2009Q2	8596
2009Q3	9867
2009Q4	9739
2010Q1	8580
2010Q2	10002
2010Q3	9742
2010Q4	10716
2011Q1	9077
2011Q2	10675
2011Q3	10425
2011Q4	10779
2012Q1	9629
2012Q2	11088
2012Q3	10782
2012Q4	11341
2013Q1	10554
2013Q2	11578
2013Q3	11568
2013Q4	12309
2014Q1	11649
2014Q2	12466
2014Q3	12389
2014Q4	13391
2015Q1	12461
2015Q2	13101
2015Q3	13512
2015Q4	15244
2016Q1	12969
2016Q2	14277
2016Q3	13142
2016Q4	14784
2017Q1	13336
2017Q2	14238
2017Q3	12779
2017Q4	15351
2018Q1	14548
2018Q2	15228
2018Q3	14307
2018Q4	15303
2019Q1	14924
2019Q2	20262
2019Q3	19100
2019Q4	20858
2020Q1	18044
2020Q2	11779

Please answer and justify, showing all steps and results using RStudio.

1. Use 2000:1 to 2013:4 for training and 2014:1 to 2020:2 for testing. The random walk with drift MASE from the test data is:

A) 1.35

B) 2.57

C) 1.43

D) 4.52

2. The revenue forecast for 2021 Q1 from the random walk with drift model is:

A) 11982

B) 9301

C) 13600

D) 12111

3. Plot the ACF for revenue, using RStudio. The ACF shows:

A) seasonal

B) cycle

C) trend

D) noise

4. Perform time series cross-validation on the data for the same set of quarterly sales. Use a random walk with drift for the forecast function. Find the MAD at step 6.

A) 945

B) 808

C) 1045

D) 1151