Question

Consider the dataset below, it is a nonstationary time series dataset.

Monthly Unemployed Young Women between Ages 16 and 19 in the U.S. from January 1961 to August 2002 (in Thousands)

*note the data set is only a part of a whole so please I need the R codes so I can apply it to my whole dataset*

Year

Jan

Feb

Mar

Apr

May

June

July

Aug

Sept

Oct

Nov

Dec

1990	535	543	542	523	544	529	514	561	562	533	565	574
1991	659	578	587	608	579	586	620	618	586	620	594	661
1992	594	601	600	575	621	715	659	614	655	552	652	621
1993	629	621	614	584	661	601	560	571	576	591	589	567
1994	558	587	590	654	566	596	561	587	550	614	531	560
1995	575	565	581	630	600	594	658	584	622	585	610	625
1996	621	576	568	584	577	556	531	578	542	553	584	603
1997	604	646	572	551	618	551	638	546	582	516	545	554
1998	483	492	526	505	523	562	505	547	519	581	516	465
1999	567	551	528	546	502	522	498	517	587	530	515	472
2000	465	512	556	482	496	383	494	506	489	488	507	482
2001	499	473	527	515	465	513	549	523	531	542	540	584
2002	612	547	561	591	570	594	587	523

A. Build an appropriate model for the series.

Please use RStudio for the problem, also please include codes and packages/libraries used, for me to use to review.

Thank you!

Answer 1

R-codes for ARIMA time series model building

#Read data from Excel csv file
data=read.csv(file.choose(),header = T)

#convert data frame in to vector-matrix
data.vec <- as.vector(t(as.matrix(data[,2:13])))/1000

#convert vector data in to time sries dat
data.ts <- ts(data.vec, c(1990,1), end = c(2001,12), frequency = 12)

#Decompose time series
data.decompose<-decompose(data.ts)

#Plot components of time series
plot(data.decompose)

#Build ARIMA model

AutoArimaModel=auto.arima(data.ts)
AutoArimaModel

summary(AutoArimaModel)

# Forecast for next 12 months
forecast(AutoArimaModel,12)

plot(forecast(AutoArimaModel,12))