<?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 Time Series Forecasting Detecting a Level Shift while Forecasting Accurately in Statistical Procedures</title>
    <link>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704429#M34067</link>
    <description>&lt;P&gt;I have attached the code I have used to generate the outputs in the pdf below.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I am struggling to find the right statements in my forecasts that would accurately forecast November 2020 and further months.Covid-19 has impacted the data and created a new level starting March 2020 to October 2020. This new level should be the starting point I know but I want to apply this level at the seasonality shown in 2017 through 2019.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;All of which routinely shows November and December taking a big dip from October. The drop from October to November in the years 2017 to 2019 is around 6,000. So I would expect to create a forecast that could predict November being between 22k and 25k. I would want to forecast for 12 months and show the trend gradually getting back to pre covid volumes.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thank you SAS community!&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;ods graphics on;&lt;BR /&gt;proc arima data=Nonship_Labor_Combined_Date plots(only)=(forecast(forecast forecastonly));&lt;BR /&gt;identify var=trans;&lt;BR /&gt;estimate p=1 q=1;&lt;BR /&gt;/*outlier type=(ls tc(6 12)) alpha=0.001 sigma=mse maxnum= 5 id=new_date; */&lt;BR /&gt;forecast lead=12 interval=month id=new_date nooutall out=results;&lt;BR /&gt;run;&lt;BR /&gt;quit;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;ods graphics on;&lt;BR /&gt;proc ucm data=shift_trans plots=all;&lt;BR /&gt;id new_date interval=month;&lt;BR /&gt;model trans = shiftMar2020;&lt;BR /&gt;level;&lt;BR /&gt;slope variance=0 noest;season length=12 type=trig print=smooth;&lt;BR /&gt;irregular;&lt;BR /&gt;estimate;&lt;BR /&gt;forecast lead=12 print= (decomp forecasts);&lt;BR /&gt;run;&lt;/P&gt;</description>
    <pubDate>Tue, 08 Dec 2020 15:57:36 GMT</pubDate>
    <dc:creator>Scottie_T</dc:creator>
    <dc:date>2020-12-08T15:57:36Z</dc:date>
    <item>
      <title>Time Series Forecasting Detecting a Level Shift while Forecasting Accurately</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704429#M34067</link>
      <description>&lt;P&gt;I have attached the code I have used to generate the outputs in the pdf below.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I am struggling to find the right statements in my forecasts that would accurately forecast November 2020 and further months.Covid-19 has impacted the data and created a new level starting March 2020 to October 2020. This new level should be the starting point I know but I want to apply this level at the seasonality shown in 2017 through 2019.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;All of which routinely shows November and December taking a big dip from October. The drop from October to November in the years 2017 to 2019 is around 6,000. So I would expect to create a forecast that could predict November being between 22k and 25k. I would want to forecast for 12 months and show the trend gradually getting back to pre covid volumes.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thank you SAS community!&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;ods graphics on;&lt;BR /&gt;proc arima data=Nonship_Labor_Combined_Date plots(only)=(forecast(forecast forecastonly));&lt;BR /&gt;identify var=trans;&lt;BR /&gt;estimate p=1 q=1;&lt;BR /&gt;/*outlier type=(ls tc(6 12)) alpha=0.001 sigma=mse maxnum= 5 id=new_date; */&lt;BR /&gt;forecast lead=12 interval=month id=new_date nooutall out=results;&lt;BR /&gt;run;&lt;BR /&gt;quit;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;ods graphics on;&lt;BR /&gt;proc ucm data=shift_trans plots=all;&lt;BR /&gt;id new_date interval=month;&lt;BR /&gt;model trans = shiftMar2020;&lt;BR /&gt;level;&lt;BR /&gt;slope variance=0 noest;season length=12 type=trig print=smooth;&lt;BR /&gt;irregular;&lt;BR /&gt;estimate;&lt;BR /&gt;forecast lead=12 print= (decomp forecasts);&lt;BR /&gt;run;&lt;/P&gt;</description>
      <pubDate>Tue, 08 Dec 2020 15:57:36 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704429#M34067</guid>
      <dc:creator>Scottie_T</dc:creator>
      <dc:date>2020-12-08T15:57:36Z</dc:date>
    </item>
    <item>
      <title>Re: Time Series Forecasting Detecting a Level Shift while Forecasting Accurately</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704718#M34098</link>
      <description>&lt;P&gt;I would recommend you re-post/move this to the Forecast and Econometrics community, where the question will be seen by people more familiar with the procedures in SAS/ETS.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;SteveDenham&lt;/P&gt;</description>
      <pubDate>Wed, 09 Dec 2020 14:04:15 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704718#M34098</guid>
      <dc:creator>SteveDenham</dc:creator>
      <dc:date>2020-12-09T14:04:15Z</dc:date>
    </item>
    <item>
      <title>Re: Time Series Forecasting Detecting a Level Shift while Forecasting Accurately</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704719#M34099</link>
      <description>Plz post it at Forecasting Forum . It is under SAS/ETS</description>
      <pubDate>Wed, 09 Dec 2020 14:04:40 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704719#M34099</guid>
      <dc:creator>Ksharp</dc:creator>
      <dc:date>2020-12-09T14:04:40Z</dc:date>
    </item>
    <item>
      <title>Re: Time Series Forecasting Detecting a Level Shift while Forecasting Accurately</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704721#M34100</link>
      <description>Plz post it at Forecasting Forum . It is under SAS/ETS</description>
      <pubDate>Wed, 09 Dec 2020 14:04:51 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/Time-Series-Forecasting-Detecting-a-Level-Shift-while/m-p/704721#M34100</guid>
      <dc:creator>Ksharp</dc:creator>
      <dc:date>2020-12-09T14:04:51Z</dc:date>
    </item>
  </channel>
</rss>

