Limited Offer Get 25% off — use code BESTW25
No AI No Plagiarism On-Time Delivery Free Revisions
Claim Now
Prices start at $5 per page · plagiarism-free papers · free revisions for 14 days 24/7 support: support@bestwriters.com
Uncategorized

assignment in a Word document or a pdf

  • Posted
  • Authoradmin
  • Length961 words
  • Reading time5 min
Write My Assignment FREE

Submit this assignment in a Word document or a pdf

1. Problem 3.39 of the book (p. 162) asks students to sketch the ACF of a seasonal ARIMA (0,1)×(1,0)12 model with the non-seasonal MA = .5 and the seasonal AR . (They left out the differencing location, which means there is no differencing.)

Here’s how you can do it in R:

acfprob1=ARMAacf(ar = c(0,0,0,0,0,0,0,0,0,0,0,.8), ma=c(.5),lag.max=61)

plot(acfprob1, type=”h”)

The first command asks for 61 lags. The ar list gives coefficients that are 0 up to the 12th lag where the coefficient = .8. The ma list gives the information that the first lag coefficient = .5. In the plot command, the type = “h” causes projections from the value to the axis as we usually do in an ACF.

a. Use R to sketch the ACF in this situation. You may either type the plot command in R rather than copy and paste or you must paste as plain text. R doesn’t like Words’ quote marks. Give the ACF graph as the answer to this part.

b. To get the PACF of this model, modify the commands in the following way:

pacfprob1=ARMAacf(ar = c(0,0,0,0,0,0,0,0,0,0,0,.8), ma=c(.5),lag.max=61, pacf=TRUE)

plot(pacfprob1, type=”h”)

Use these commands to get the PACF of the model. Give the PACF as the answer to this problem.

c. Write a brief description of the ACF and PACF patterns for this model.

d. Now consider the ARIMA (1,0)×(0,1)12 model with non-seasonal AR=.8 and seasonal MA =.5. Graph the ACF of this model. (Reverse the seasonality options in the acfprob1 = command.) Give the graph as the answer to this part.

e. Refer to part d. Graph the PACF for the model. Give the graph as the answer to this part.

f. Write a brief description of the ACF and PACF patterns in parts d and e.

2. This is a more structured version of problem 3.43 in the book (p. 162) that asks students to fit a model of their choice to the US Live Birth Series. It gives n = 373 consecutive monthly measures of the number of live births in the United States. The data are in the dataset birth.dat.

Load the astsa library and read in the dataset:

library(astsa)

birth=scan(“birth.dat”)

Then, because we have linear trend and seasonality (monthly effects), apply both first and twelfth differences. Something like:

diff12 = diff(birth,12)

acf2(diff12)

diff1and12=diff(diff12,1)

a. Use acf2 to determine the ACF and PACF of the differenced data. Be sure that you supply the differenced series (diff1and12) as a parameter of the acf2 command. Give the ACF and PACF of the differenced data.

b. Consider the ACF and PACF values for the first two or three lags. Explain why they suggest either an MA(1) or AR(1) as a possible non-seasonal component.

c. Examine the ACF and PACF at lags 12, 24 and 36. What possibilities are suggested for a seasonal component of the model? Provide reasoning.

d. The ACF and PACF don’t actually give a clear pointer to the best model in this case, but here some possibilities (Note that this list is not all-inclusive):

ARIMA (0,1,1)×(0,1,1)12

ARIMA (0,1,2)×(0,1,1)12

ARIMA (1,1,1)×(0,1,1)12

The last model is most evident from the residuals of an ARIMA (0,0,0)×(0,1,1)12. Assuming that you loaded the astsa library, the sarima command will estimate the first model above.

sarima ( name of series, 0,1,1,0,1,1,12)

Making necessary adjustments to the sarima command, estimate the three models above. For each give the variance estimate and a brief description of the model diagnostics. (You don’t have to provide all of the graphs.)

e. Which of the three models in part d do you think is the best model? Explain.

3. This is a more structured version of problem 3.41 on page 162 of the book. The problem asks students to determine a model for a monthly series of number of unemployed persons in the United States. There are n = 372 data values, the monthly numbers for 1948-1978.

The data are in the file unemp.dat in the Datasets folder.

There is an upward trend because these are numbers of people rather than rates, so the number increases as the population size increases.

There also is seasonality. Unemployment numbers are affected by such things as seasonal agricultural production, holiday seasons, and so on.

a. Calculate a variable that is the result of applying a 12th difference and then a 1st difference. Something like the following should do the trick:

diff12=diff(x,12)

acf2(diff12)

diff1and12=diff(diff12,1)

Then determine the ACF and PACF of the differenced data series (diff1and12). Be sure to use the differenced series, not the original data for this step.

(i) Discuss why the ACF and PACF suggest that an AR(2) may be good for the non-seasonal component. (Look at what’s happening for the first few lags in the ACF and PACF.)

(ii) Discuss why the ACF and PACF suggest that a seasonal MA(1) may be good for the seasonal component. Look at what’s happening around lags 12, 24, and so on.

b. Fit an ARIMA (2,1, 0,)× (0,1,1)12 model to the original series. In the sarima command, specify the original series, not the differenced series because the differencing will be specified in the command itself. Discuss whether the diagnostics for the model suggest that the model is suitable. (Look at ACF of residuals, Ljung-Box-Pierce tests, and so on.)

c. Use the model fit in part b to forecast the unemployment values for the next 12 months following the end of the series. One possible command structure is

sarima.for (name of data series, 12, 2,1,0,0,1,1,12)

Give the 12 forecasts as the answer to this problem.

The post assignment in a Word document or a pdf appeared first on My Assignment Online.

Plagiarism Free Assignment Help

Expert Help With This Assignment — On Your Terms

✓ Native UK, USA & Australia writers ✓ Deadline from 3 hours ✓ 100% Plagiarism-Free — Turnitin included ✓ Unlimited free revisions ✓ Free to submit — compare quotes

admin

Academic writer and editor at BestWriters.

Get your paper off your plate tonight

Post your instructions for free. Prices start at $5 per page.

Place an order