On Dec 27, 2011, at 15:47 , Scott Raynaud wrote:
> I have the following code (which I did not write) that generates
> data based on a logistic model. I'm only getting a single record
> with y=1. It seems implausible that in 50k cases that have a
> single y=1. Does that ring alarm bells for anyone else?
>
Not really. As far as I can tell, "fixpart" is roughly -10.5 (= -1.5 -
.25*36), so binomprob is around 2.75e-5, which - nonlinearity notwithstanding -
suggests that the expected number of positives out of 50K is something like 1.4.
To do this more precisely, just compute and print sum(binomprob) in the code you
gave.
> beta<-c(-1.585600,-0.246900)
> betasize<-length(beta)
> meanpred<-c(0,35.900000)
> varpred<-c(0,1.000000)
> #loop code
> x<-matrix(1,length,betasize) #length set to 50k
> #loop code
> x[,2]<-rnorm(length,meanpred[2],sqrt(varpred[2])) #length set to 50k
> fixpart<-x%*%beta
> binomprob<-exp(fixpart)/(1+exp(fixpart))
> data$y<-rbinom(n1,1,binomprob)
> #more loop code
>
> ______________________________________________
> R-help at 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.
--
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com