Date: Mon, 21 Jan 2002 09:27:10 -0800
Reply-To: Yvette <hollied@OLIN.WUSTL.EDU>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Yvette <hollied@OLIN.WUSTL.EDU>
Organization: http://groups.google.com/
Subject: Re: Dummy variables in regressions
Content-Type: text/plain; charset=ISO-8859-1
I tried proc glm but I still appear to have the same problem. For example:
CODE USED:
proc glm data = in.d_file_master;
class d1 d2 d3;
Model y = x1 x2 d1 d2 d3; (d are dummy variables)
run;
quit;
LOG REPORTS:
proc glm data = in.d_file_master;
473 class d1 d2 d3;
474 Model y = x1 x2 d1 d2 d3;
475 run;
476 quit;
NOTE: PROCEDURE GLM used:
real time 1.43 seconds
cpu time 0.11 seconds
OUPUT FILE SHOWS:
The GLM Procedure
Class Level Information
Class Levels Values
d1 1 1
d2 1 0
d3 1 0
Number of observations 10669
NOTE: Due to missing values, only 333 observations can be used in this analysis.
------
Dependent Variable: y
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 2 1.02314108 0.51157054 261.99 <.0001
Error 330 0.64437587 0.00195265
Corrected Total 332 1.66751696
R-Square Coeff Var Root MSE y Mean
0.613572 55.17548 0.044189 0.080088
Source DF Type I SS Mean Square F Value Pr > F
x1 1 0.66287916 0.66287916 339.48 <.0001
x2 1 0.36026193 0.36026193 184.50 <.0001
d1 0 0.00000000 . . .
d2 0 0.00000000 . . .
d3 0 0.00000000 . . .
Source DF Type III SS Mean Square F Value Pr > F
x1 1 0.96834274 0.96834274 495.91 <.0001
x2 1 0.36026193 0.36026193 184.50 <.0001
d1 0 0.00000000 . . .
d2 0 0.00000000 . . .
d3 0 0.00000000 . . .
Thoughts????? Thanks.
-----------------------------------
olaf.kruse@VST-GMBH.DE (Dr Olaf Kruse) wrote in message news:<416A07CFC6C1D311A796000083295C33FA32@VSTPC04>...
> Yvette wrote (in parts):
>
> Hello again,
>
> >I am trying to run regressions with and without dummary variables.
> >The regressions without the dummy variables is running smoothly.
> >However, the regressions with dummy variables is not giving me the
> >parameter estimates and the p-values. The information seems to be
> >consistent and fine for the independent variables (non-dummy) but the
> >output for the results show a '.' However, the log indicates no
> >errors or notes.
>
> >Proc reg data = in.file
> >Model Y = x1 x2 x3 d1 d2 d3 (d's represent the dummy variables)
> >run;
>
>
> Proc Reg doesn't support class-variables. Use Proc GLM instead:
>
> Proc glm data = in.file
> class d1 d2 d3
> Model Y = x1 x2 x3 d1 d2 d3 (d's represent the dummy variables)
> run;
>
> HTH (learned something on SAS-L recently ;-)) Olaf
