[R] Problem with clusplot

Sarah Goslee sarah.goslee at gmail.com
Tue Dec 6 14:33:41 CET 2011

Well, if I had to guess (and I do, since we have no idea what your data
look like, and calling your data matrix is a very bad idea): you have more
variables than units, so clusplot() can't use princomp() to create a reduced-
dimension plot.

You could roll your own using a different ordination method, or you could
remove some variables from your data, or you could think about whether
the data you have are appropriate for what you're trying to do with them.


On Tue, Dec 6, 2011 at 12:58 AM, elisacarli21 <elisacarli21 at gmail.com> wrote:
> Dear all
> I'm trying to run a cluster analysis with R
> Here are the commands:
> mydata <- na.omit(matrix) # listwise deletion of missing
> mydata <- scale(matrix) # standardize variables
>  fit <- kmeans(mydata, 8) # 8 cluster solution
>  # get cluster means
> aggregate(mydata,by=list(fit$cluster),FUN=mean)
>  # append cluster assignment
>  mydata <- data.frame(mydata, fit$cluster)
>  library(cluster)
> clusplot(mydata, fit$cluster, color=TRUE, shade=TRUE,  labels=2, lines=0)
> I get the following error
> Error in princomp.default(x, scores = TRUE, cor = ncol(x) != 2) :
>  'princomp' can only be used with more units than variables
> How can I fix it?
> Thanks

Sarah Goslee

More information about the R-help mailing list