R help - Nov 2011 - help with formula for clogit

I would like to know if clogit function can be used as below
clogit(group~., data=dataframe)

When I try to use in above format it takes a long time, I would appreciate
some pointers to get multiple combinations tested.

set.seed(100)
 d=data.frame(x=rnorm(20)+5,
 x1=rnorm(20)+5,
 x2=rnorm(20)+5,
 x3=rnorm(20)+5,
 x4=rnorm(20)+5,
 x5=rnorm(20)+5,
 x6=rnorm(20)+5,
 x7=rnorm(20)+5,
 x8=rnorm(20)+5,
group=rep(c(1,2),10), Age=rnorm(20)+35,strata=c(rep(1,10), rep(2,10))) 

nam=names(d)[1:9]
results <- c("Protein", "OR", "p-val")
pc3=combinations(n=length(nam),r=2)

for (len in 1:dim(pc3)[2])
 {
  prs=pc3[len,]
  mols=nam[prs]
  d2=d[,c(mols,'group','Age','strata')]
  log.reg<-clogit(group~.,data=d2)
  a = summary(log.reg)$conf.int
  z= summary(log.reg)$coefficients[1,4]  #ncol in coefficients is 3 * number
of parameters
  pval  = 2*pnorm(-abs(z))
  res2 = c(paste('IL8+',molecule,sep=''), paste
(round(a[1,1],2), " ("  ,
round(a[1,3],2), " - " , round(a[1,4],2), ")" ,
sep=""), pval)
  results  = rbind (results ,res2 )
}

Thanks
Sharad

--
View this message in context:
http://r.789695.n4.nabble.com/help-with-formula-for-clogit-tp3998967p3998967.html
Sent from the R help mailing list archive at Nabble.com.

On Tue, Nov 8, 2011 at 4:19 AM, 1Rnwb <sbpurohit at gmail.com>
wrote:
> I would like to know if clogit function can be used as below
> clogit(group~., data=dataframe)
>
Not usefully.  That syntax does not specify a strata() term, which is
why the computation is very slow and probably not what you intended.

   -thomas

-- 
Thomas Lumley
Professor of Biostatistics
University of Auckland

clogit needs to spell out formula plus strata, you can try modify the code
...
d2=d[,c(mols,'group','Age','strata')]
fo<-as.formula(paste(paste('group~','Age',sep=''),'strata(strata)',sep='+'))
log.reg<-clogit(fo,data=d2)
...

Weidong

On Mon, Nov 7, 2011 at 10:19 AM, 1Rnwb <sbpurohit at gmail.com>
wrote:
> I would like to know if clogit function can be used as below
> clogit(group~., data=dataframe)
>
> When I try to use in above format it takes a long time, I would appreciate
> some pointers to get multiple combinations tested.
>
> set.seed(100)
> ?d=data.frame(x=rnorm(20)+5,
> ?x1=rnorm(20)+5,
> ?x2=rnorm(20)+5,
> ?x3=rnorm(20)+5,
> ?x4=rnorm(20)+5,
> ?x5=rnorm(20)+5,
> ?x6=rnorm(20)+5,
> ?x7=rnorm(20)+5,
> ?x8=rnorm(20)+5,
> group=rep(c(1,2),10), Age=rnorm(20)+35,strata=c(rep(1,10), rep(2,10)))
>
> nam=names(d)[1:9]
> results <- c("Protein", "OR", "p-val")
> pc3=combinations(n=length(nam),r=2)
>
> for (len in 1:dim(pc3)[2])
> ?{
> ?prs=pc3[len,]
> ?mols=nam[prs]
> ?d2=d[,c(mols,'group','Age','strata')]
> ?log.reg<-clogit(group~.,data=d2)
> ?a = summary(log.reg)$conf.int
> ?z= summary(log.reg)$coefficients[1,4] ?#ncol in coefficients is 3 * number
> of parameters
> ?pval ?= 2*pnorm(-abs(z))
> ?res2 = c(paste('IL8+',molecule,sep=''), paste
(round(a[1,1],2), " (" ?,
> round(a[1,3],2), " - " , round(a[1,4],2), ")" ,
sep=""), pval)
> ?results ?= rbind (results ,res2 )
> }
>
> Thanks
> Sharad
>
> --
> View this message in context:
http://r.789695.n4.nabble.com/help-with-formula-for-clogit-tp3998967p3998967.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.
>