Date:         Wed, 31 Jan 2007 16:23:44 -0500
Reply-To:     Wensui Liu <liuwensui@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Wensui Liu <liuwensui@GMAIL.COM>
Subject:      Re: Additive Regression Models using splines
Comments: To: aru Sutha <asutha@hotmail.com>
In-Reply-To:  <200701312045.l0VK9V2O012812@mailgw.cc.uga.edu>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

sutha,

take a look at proc gam, which should be available for SAS8.2 and above.

for univariate smoothing, proc gam uses spline and local reg. for
bivariate smoothing, it uses thin-plate smoother.

the default setting of df in proc gam is 4. but you can also use gcv
to do automatic df selection.

hth.

On 1/31/07, aru Sutha <asutha@hotmail.com> wrote:
> There is a function R, aregImpute, to do Additive Regression Models using
> spline.  Are there any proc or function in SAS to do a similar model?
> Proc MI does the multiple imputation but not using spline.
>


--
WenSui Liu
A lousy statistician who happens to know a little programming
(http://spaces.msn.com/statcompute/blog)