Date: Mon, 31 Oct 2011 23:58:31 -0700
Reply-To: Daniel Nordlund <djnordlund@FRONTIER.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Daniel Nordlund <djnordlund@FRONTIER.COM>
Subject: Re: GML in SAS and R
In-Reply-To: <1320111383.29398.YahooMailNeo@web161403.mail.bf1.yahoo.com>
Content-Type: text/plain; charset="UTF-8"
> -----Original Message-----
> From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of oslo
> Sent: Monday, October 31, 2011 6:36 PM
> To: SAS-L@LISTSERV.UGA.EDU
> Subject: Re: GML in SAS and R
>
> Dear Farzeeza;
>
> Thanks a lot. Here is the SAS code.
>
> proc glm data= a;
> class t;
> model
> y=t/p ss3 solution;
> output out=pout1 RESIDUALS
> = ry;
> run;
> data b;
> set pout1;
> my= 7.273791363+ry; /*here 7.273791363 is means of y;
> run;
>
I am not sure if my original reply got lost in the luminiferous ether, so I have reposted it.
It would have been useful to see the exact R code that you used along with the SAS code. For your data, what do the values of t stand for? You are treating them as a class variable. Do you really want to do that? Did you create t as factor variable in R?
In any case, if I treat t as a class variable in SAS and treat t as factor variable in R, then the only difference between SAS and R is how the reference class is defined. In SAS, you could create a format to change the minimum value to the max value. With your sample data you could do
/*format minimum value to have largest formatted value*/
proc format;
value refmt
15 = '99'
;
run;
proc glm data= a order=formatted;
format t refmt.;
class t;
model y=t/p ss3 solution;
output out=pout1 RESIDUALS=ry;
quit;
Using this approach, I get the same results as in R with the syntax
summary(lm(y ~ factor(t))
Hope this is helpful,
Dan
Daniel Nordlund
Bothell, WA USA
>
> ________________________________
> From: Fareeza Khurshed <fkhurshed@GMAIL.COM>
> To: SAS-L@LISTSERV.UGA.EDU
> Sent: Monday, October 31, 2011 5:23 PM
> Subject: Re: GML in SAS and R
>
> Can you show the code you ran in SAS Oslo?
>
>
>
> On Mon, Oct 31, 2011 at 3:20 PM, oslo <hokut1@yahoo.com> wrote:
>
> > Peter;
> >
> > Thanks. ANOVA and p values are the same. But estimation for fixed
> effects
> > are not the same.
> >
> > Oslo
> >
> > From: Peter Flom <peterflomconsulting@MINDSPRING.COM>
> > To: SAS-L@LISTSERV.UGA.EDU
> > Sent: Monday, October 31, 2011 5:16 PM
> > Subject: Re: GML in SAS and R
> >
> > I ran both this in both SAS and R and got identical results, with no
> > parameter fixed at 0.
> >
> > In SAS
> >
> > proc glm data = a;
> > model y = t;
> > run;
> >
> > in R
> >
> > y <- c(8.45868914, 8.80001656, 11.19760853, 6.48274648, 4.13212718,
> > 2.56846860, 0.93439027, 10.42572772, 3.37216466, 7.97682921,
> > 3.90145850,
> > 1.42026869, 8.19922699)
> > t <- c(48, 28, 35, 15, 21, 50, 31, 46, 40, 26, 48, 32, 15)
> >
> > m1 <- lm(y ~ t)
> > summary(m1)
> >
> > Peter
> >
> > Peter Flom
> > Peter Flom Consulting
> > http://www.statisticalanalysisconsulting.com/
> > http://www.IAmLearningDisabled.com
> >
> > -----Original Message-----
> > From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of
> oslo
> > Sent: Monday, October 31, 2011 5:48 PM
> > To: SAS-L@LISTSERV.UGA.EDU
> > Subject: GML in SAS and R
> >
> > Dear SAS-users;
> >
> > I have run the same data in R and SAS but got different prediction
> results
> > after model fit. Suppose I want fit the following model for the data
> given
> > below (a part of data)
> >
> > y(ij)=mu + t(i) + e(ij)
> >
> > at the end I want to have
> >
> > y* =y(ij)-t(i)_hat
> > For this I run SAS and R but got differnt results.SAS and R are using
> > different constrain on fixed effect. In SAS the last effect, in R the
> first
> > fixed effect set to zero.
> >
> > In addition is it possible to set the first fixed effect to zero in SAS
> as
> > R
> > does?
> >
> > Regards,
> >
> > Oslo
> > data a;
> > input y t;
> > cards;
> > 8.45868914 48
> > 8.80001656 28
> > 11.19760853 35
> > 6.48274648 15
> > 4.13212718 21
> > 2.56846860 50
> > 0.93439027 31
> > 10.42572772 46
> > 3.37216466 40
> > 7.97682921 26
> > 3.90145850 48
> > 1.42026869 32
> > 8.19922699 15
> >
| 
