<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: Do a Rolling ARIMA in Sample in New SAS User</title>
    <link>https://communities.sas.com/t5/New-SAS-User/Do-a-Rolling-ARIMA-in-Sample/m-p/554739#M9564</link>
    <description>&lt;P&gt;I think this would require a macro to loop through all of the dates and run PROC ARIMA on each group of 60 observations and do the predictions.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;As I have never seen an ARIMA(12,0,0) model, I wonder if somehow you can get PROC EXPAND to do the calculations without using a macro to do all the looping, but that would require some math to be performed to see if the model is even possible to compute in PROC EXPAND.&lt;/P&gt;</description>
    <pubDate>Mon, 29 Apr 2019 15:17:13 GMT</pubDate>
    <dc:creator>PaigeMiller</dc:creator>
    <dc:date>2019-04-29T15:17:13Z</dc:date>
    <item>
      <title>Do a Rolling ARIMA in Sample</title>
      <link>https://communities.sas.com/t5/New-SAS-User/Do-a-Rolling-ARIMA-in-Sample/m-p/554702#M9552</link>
      <description>&lt;P&gt;I am a very new user of SAS and am not comfortable with almost any aspect of it. I usually use Rstudio(and thus think in terms of that coding language), but have found myself in a situation where I have to use SAS. What I want to do is pretty straightforward, but I do not even know how to go about looking up how to do it(although I have spent awhile trying).&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I have a string of dates and 300 numbers. What I want to do is run an ARIMA(12,0,0) model(AR(12) model) in sas for every 60 observations, and then spit out the predicted variable in a new column lining up with the 61st row. In other words, say that column 2 was log differences in CPI, and column 1 was the date. I want the first 60 observations in the new column 3 to be blank, and then the 61st observation to be the ARIMA(12,0,0) projection using observations 49 to 60, where the coefficients of the model are estimated using observations 1 to 60.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;The 62nd observation is analogous, it is the projection using observations 50 to 61, coefficients estimated using data 2 to 61. I have found the proc arima command online, but I cannot find a way to forecast in sample, only to forecast out of sample(uses the data in the entire dataset and then projects into the future). This is such a standard process(I think) that there must be some simple source I can look up online to tell me how to do this. I was wondering if anyone can provide a link?&lt;/P&gt;</description>
      <pubDate>Mon, 29 Apr 2019 14:07:15 GMT</pubDate>
      <guid>https://communities.sas.com/t5/New-SAS-User/Do-a-Rolling-ARIMA-in-Sample/m-p/554702#M9552</guid>
      <dc:creator>stealth4933</dc:creator>
      <dc:date>2019-04-29T14:07:15Z</dc:date>
    </item>
    <item>
      <title>Re: Do a Rolling ARIMA in Sample</title>
      <link>https://communities.sas.com/t5/New-SAS-User/Do-a-Rolling-ARIMA-in-Sample/m-p/554739#M9564</link>
      <description>&lt;P&gt;I think this would require a macro to loop through all of the dates and run PROC ARIMA on each group of 60 observations and do the predictions.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;As I have never seen an ARIMA(12,0,0) model, I wonder if somehow you can get PROC EXPAND to do the calculations without using a macro to do all the looping, but that would require some math to be performed to see if the model is even possible to compute in PROC EXPAND.&lt;/P&gt;</description>
      <pubDate>Mon, 29 Apr 2019 15:17:13 GMT</pubDate>
      <guid>https://communities.sas.com/t5/New-SAS-User/Do-a-Rolling-ARIMA-in-Sample/m-p/554739#M9564</guid>
      <dc:creator>PaigeMiller</dc:creator>
      <dc:date>2019-04-29T15:17:13Z</dc:date>
    </item>
  </channel>
</rss>

