library(pcaMethods)  #this is purely an example to show ease of missing value imputation
library(psych) #has the data
data(bfi)  #load the data

#bfi=read.table("http://www.unt.edu/rss/class/mike/data/bfi.txt", header=T)  #alternatively one can download it

pcamod <- pca(bfi, nPcs = 5, method = "bpca")
#summary(pcamod)  #if you want to look at the pca output
bfi2=data.frame(pcamod@completeObs)  #the now completed dataset
bfi3=t(bfi2)  #transpose the data so that we can cluster variables

library(cluster)
data(bfi3)
agn1 <- agnes(bfi3, method = "ward")  #agnes = agglomerative nesting
agn1  #AC is a quantity varying between 0 and 1. AC close to 1 indicates that a very clear structuring has been found.
      #AC close to 0 indicates that the algorithm has not found a natural structure. In other words, the data consists of only one big cluster.
plot(agn1)  #I personally don't care for the banner plot other than it shows clean cluster breaks.  This is interesting to compare to a factor analysis.  They serve different purposes but there are similarities in the results.  For example, FA would show poor loading for A1, good loading for Ns.

#membership
cutree(agn1,2)  #2 for 2 cluster solution
classmemb=data.frame(FFA=rownames(bfi3),class=cutree(agn1,2))  #creates a dataset with just variables (five factor attribute) and cluster membership
classmemb

#Factor analysis comparison. Remove comment marks to runs
#FAmod <- factanal(~A1+A2+A3+A4+A5+C1+C2+C3+C4+C5+E1+E2+E3+E4+E5+N1+N2+N3+N4+N5+O1+O2+O3+O4+O5,factors=5, rotation="promax", scores="none", data=bfi2)
#print(FAmod, digits=2, cutoff=.3)
#FAmod$factors