Turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- Home
- /
- Analytics
- /
- Forecasting
- /
- How to get the critical values for acf, pacf?

Options

- RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page

🔒 This topic is **solved** and **locked**.
Need further help from the community? Please
sign in and ask a **new** question.

- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content

Posted 09-13-2018 11:15 AM
(3036 views)

I use

ods output Arima.Identify.AutoCorrGraph= ACF;

ods output Arima.Identify.PACFGraph = PACF;

to output the acf, pacf for arima. But they do not have critical values. Is there anyway to get it? Thanks.

1 ACCEPTED SOLUTION

Accepted Solutions

- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content

Hi @happydog,

The ODS OUTPUT data sets you created from PROC ARIMA include the approximate upper and lower 95% confidence limits about the autocorrelations and partial autocorrelations for each lag. These confidence limits allow you to determine the significance of the autocorrelation or partial autocorrelation at each lag using an approximate alpha level of 0.05.

P-values are more useful than critical values, since they can be used to determine significance at different alpha levels. Since you are looking at the ACF and PACF associated with the original (or differenced) series computed by the IDENTIFY statement in PROC ARIMA, another alternative to see these autocorrelation functions is PROC TIMESERIES. The CORR statement in PROC TIMESERIES allows you to explicitly indicate the statistics to write to the OUTCORR= data set. These statistics include the ACFPROB and PACFPROB, which are the p-values associated with the autocorrelations and partial autocorrelations.

Following, please find an example which compares the results of the OUTCOV= data set created by the IDENTIFY statement in PROC ARIMA, with the corresponding values written to the OUTCORR= data set by PROC TIMESERIES. In addition to the autocorrelations and partial autocorrelations, the OUTCORR= data set generated by PROC TIMESERIES also includes the ACFPROB and PACFPROB variables when these options are specified in the CORR statement:

```
data air;
set sashelp.air;
y=log(air);
run;
proc arima data=air plots=all;
identify var=y nlag=12 outcov=cov;
run;
quit;
proc timeseries data=air plots=corr outcorr=corr_pvals;
id date interval=month;
var y;
corr lag n acf pacf acfstd acfprob pacfprob / nlag=12;
run;
/* data set generated by PROC ARIMA OUTCOV= option */
proc print data=cov;
var lag n corr stderr partcorr;
run;
/* data set generated by PROC TIMESERIES OUTCORR= option */
/* which includes ACFPROB and PACFPROB variables */
proc print data=corr_pvals;
run;
```

For more details on PROC TIMESERIES, please see the following documentation link:

I hope this helps!

DW

3 REPLIES 3

- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content

Modif from a SAS Doc example:

```
proc arima data=seriesj plots(unpack)=all;
/*--- Cross-correlation of prewhitened series ---------------*/
identify var=y crosscorr=(x) nlag=12;
ods output SeriesACFPlot=SAP SeriesPACFPlot=SPP;
run;
```

PG

- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content

Using the PROC ARIMA procedure, we can have the confidence interval showing in the shaded area in the plots of ACF and PACF. The correspondence critical value is around 2 for a 95% CI.

- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content

Hi @happydog,

The ODS OUTPUT data sets you created from PROC ARIMA include the approximate upper and lower 95% confidence limits about the autocorrelations and partial autocorrelations for each lag. These confidence limits allow you to determine the significance of the autocorrelation or partial autocorrelation at each lag using an approximate alpha level of 0.05.

P-values are more useful than critical values, since they can be used to determine significance at different alpha levels. Since you are looking at the ACF and PACF associated with the original (or differenced) series computed by the IDENTIFY statement in PROC ARIMA, another alternative to see these autocorrelation functions is PROC TIMESERIES. The CORR statement in PROC TIMESERIES allows you to explicitly indicate the statistics to write to the OUTCORR= data set. These statistics include the ACFPROB and PACFPROB, which are the p-values associated with the autocorrelations and partial autocorrelations.

Following, please find an example which compares the results of the OUTCOV= data set created by the IDENTIFY statement in PROC ARIMA, with the corresponding values written to the OUTCORR= data set by PROC TIMESERIES. In addition to the autocorrelations and partial autocorrelations, the OUTCORR= data set generated by PROC TIMESERIES also includes the ACFPROB and PACFPROB variables when these options are specified in the CORR statement:

```
data air;
set sashelp.air;
y=log(air);
run;
proc arima data=air plots=all;
identify var=y nlag=12 outcov=cov;
run;
quit;
proc timeseries data=air plots=corr outcorr=corr_pvals;
id date interval=month;
var y;
corr lag n acf pacf acfstd acfprob pacfprob / nlag=12;
run;
/* data set generated by PROC ARIMA OUTCOV= option */
proc print data=cov;
var lag n corr stderr partcorr;
run;
/* data set generated by PROC TIMESERIES OUTCORR= option */
/* which includes ACFPROB and PACFPROB variables */
proc print data=corr_pvals;
run;
```

For more details on PROC TIMESERIES, please see the following documentation link:

I hope this helps!

DW

Are you ready for the spotlight? We're accepting content ideas for **SAS Innovate 2025** to be held May 6-9 in Orlando, FL. The call is **open **until September 25. Read more here about **why** you should contribute and **what is in it** for you!

Multiple Linear Regression in SAS

Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.

Find more tutorials on the SAS Users YouTube channel.