R help - Oct 2009 - Use of the command 'replicate'

Dear R users,

I'd like to ask u whether you know how to sort out the following:

I'm trying to reproduce a dataset of clusters, and for that I need to build
up a cluster index (inside a function) using the command "replicate"
as
follows:


dataset<- function(
clusters=100,
cluster.size=50,
outcome.mean=list(mean=3, sd=1),
outcome.var=list(mean=0, sd=1),
treat=1,
unbal = FALSE){

if(clusters%%2!=0) stop("number of clusters, clusters, must be an even
number \n")

clust.size<-cluster.size
clusters<-clusters

clust.index <- rep(1:clusters, times=clust.size)               # cluster
index
treat.ind <- c(                                                            
# treatment index indicator
    rep(0, sum(clust.size[1:(clusters/2)])),
    rep(1, sum(clust.size[(clusters/2+1):clusters]))
    )

(...)

}

Doing this it gives me an error saying that the 'times' argument is
invalid
and that the cluster and treatment indicator indices are both invalid, and
therefore, I'm probably defining "clust.size" wrongly?!
Could you elucidate me about this??

Thanks a lot.

Manny Gomez
-- 
View this message in context:
http://www.nabble.com/Use-of-the-command-%27replicate%27-tp26001367p26001367.html
Sent from the R help mailing list archive at Nabble.com.

See if this gives you a hint:

  size <- 50
  num  <- 100
  sum(size[1:(num/2)])

  -Peter Ehlers

Manny Gomez wrote:
> Dear R users,
> 
> I'd like to ask u whether you know how to sort out the following:
> 
> I'm trying to reproduce a dataset of clusters, and for that I need to
build
> up a cluster index (inside a function) using the command
"replicate" as
> follows:
> 
> 
> dataset<- function(
> clusters=100,
> cluster.size=50,
> outcome.mean=list(mean=3, sd=1),
> outcome.var=list(mean=0, sd=1),
> treat=1,
> unbal = FALSE){
> 
> if(clusters%%2!=0) stop("number of clusters, clusters, must be an even
> number \n")
> 
> clust.size<-cluster.size
> clusters<-clusters
> 
> clust.index <- rep(1:clusters, times=clust.size)               # cluster
> index
> treat.ind <- c(
> # treatment index indicator
>     rep(0, sum(clust.size[1:(clusters/2)])),
>     rep(1, sum(clust.size[(clusters/2+1):clusters]))
>     )
> 
> (...)
> 
> }
> 
> Doing this it gives me an error saying that the 'times' argument is
invalid
> and that the cluster and treatment indicator indices are both invalid, and
> therefore, I'm probably defining "clust.size" wrongly?!
> Could you elucidate me about this??
> 
> Thanks a lot.
> 
> Manny Gomez