<?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: Equation from Proc transreg in Statistical Procedures</title>
    <link>https://communities.sas.com/t5/Statistical-Procedures/Equation-from-Proc-transreg/m-p/232185#M12208</link>
    <description>&lt;P&gt;Try using BSPLINE and PSPLINE to get the coefficients. &amp;nbsp;BSPLINE should give what you need for the basis, realizing that SPLINE forms a linear combination of the columns of this basis. &amp;nbsp;PSPLINE is piecewise polynomial, and from your description, that may be what you are expecting, so it is probably worth exploring that option as well. &amp;nbsp;Good luck, and please share what you find back here.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Steve Denham&lt;/P&gt;</description>
    <pubDate>Thu, 29 Oct 2015 12:28:19 GMT</pubDate>
    <dc:creator>SteveDenham</dc:creator>
    <dc:date>2015-10-29T12:28:19Z</dc:date>
    <item>
      <title>Equation from Proc transreg</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/Equation-from-Proc-transreg/m-p/232123#M12205</link>
      <description>&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Proc Transreg looks like a powerful tool but it seems impossible to figure out the estimated model. From the SAS example I added "DETAIL" but only get knots and coefficients and can't relate them back to the transformation which I assume is cubic by default. The example below has 3 knots and therefore 4 cubic equations i.e. 12 coefficients plus an intercept but the output shows 7 coefficients. What am I doing wrong? is there a paper on this someplace? thanks for any help you can give. Joe.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;title2 'A Cubic Spline Fit with Knots at X=5, 10, 15';&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;proc transreg detail data=a;&lt;/P&gt;&lt;P&gt;model identity(y) = spline(x / knots=5 10 15);&lt;/P&gt;&lt;P&gt;output out=trans coefficients;&lt;/P&gt;&lt;P&gt;proc print;&lt;/P&gt;&lt;P&gt;run;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Wed, 28 Oct 2015 20:17:31 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/Equation-from-Proc-transreg/m-p/232123#M12205</guid>
      <dc:creator>jlynch</dc:creator>
      <dc:date>2015-10-28T20:17:31Z</dc:date>
    </item>
    <item>
      <title>Re: Equation from Proc transreg</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/Equation-from-Proc-transreg/m-p/232185#M12208</link>
      <description>&lt;P&gt;Try using BSPLINE and PSPLINE to get the coefficients. &amp;nbsp;BSPLINE should give what you need for the basis, realizing that SPLINE forms a linear combination of the columns of this basis. &amp;nbsp;PSPLINE is piecewise polynomial, and from your description, that may be what you are expecting, so it is probably worth exploring that option as well. &amp;nbsp;Good luck, and please share what you find back here.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Steve Denham&lt;/P&gt;</description>
      <pubDate>Thu, 29 Oct 2015 12:28:19 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/Equation-from-Proc-transreg/m-p/232185#M12208</guid>
      <dc:creator>SteveDenham</dc:creator>
      <dc:date>2015-10-29T12:28:19Z</dc:date>
    </item>
  </channel>
</rss>

