Thanks for this help. I'll read the documentation and see if I can work through the problem. I actually look forward to Mr. Harrell's reply. I have been taught that stepwise is not a good approach to use and should be avoided at all costs, so he may give me a cogent argument. The problem I posted came from my supervisor to me or I wouldn't be trying to work through the stepwise procedure. But on the brightside, it gives me an opportunity to learn something in R.
Zeda. ________________________________ From: Uwe Ligges <lig...@statistik.tu-dortmund.de> Cc: r-help@r-project.org Sent: Fri, April 15, 2011 10:28:10 AM Subject: Re: [R] Stepwise Regression with no Origin On 12.04.2011 21:52, Zd Gibbs wrote: > Sorry, the first version was incomplete. I'm trying again. > > I am running a regression equation and i want to enter in 12 IV then stepwise > enter 8 variables and not have an origin. > > DV is "shfl". > I want to enter in the following 12 independent dummy variables > ajan > bfeb > cmar > dapr > emay > fjun > > And then I want to enter in a stepwise fashion > slag6 > slag7 > slag8 > slag9 > slag10 > slag11 > slag12 > > And finally, I want there to be no origin. > > I've done simple regression, but nothing quite like this. See ?step which allows to specify minimal, maximal models as well as forward stepwise selection. BTW: Someone called Frank Harrell will shortly let you know about possible problems in model interpretation when using stepwise variable selection. Best, Uwe Ligges > I would appreciate your help. > > Thanks > > Zeda > [[alternative HTML version deleted]] > > > > > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. [[alternative HTML version deleted]]
______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
