6

I'm not an expert in cluster analysis and thus not really familiar with all the "special" contributed packages out there. Hence, I just used base routines as the did here for a quick example I was asked to produce.

The problem is that my dataset has roughly 7800 observations and thus the leaf labels are by far too manifold and clutter up the whole x-axis. So, how can I suppress the plotting of the labels? I'm guessing there must be some "exotic" parameter accessible via par() that controls this?

Rappster
  • 12,762
  • 7
  • 71
  • 120
  • Would be nice to see why this Q got downvoted? BTW, Jonathan noticed something in a hclust plotting function: http://stackoverflow.com/questions/17232178/change-label-size-of-cluster-dendrogram-in-r-3-01 – Roman Luštrik Jun 21 '13 at 11:52
  • Sorry guys, totally missed that! Usually, I *do* read the help files ;-) – Rappster Jun 21 '13 at 17:01

1 Answers1

6

See the help for hclust

?hclust

There you will see an explanation of the labels argument: "If labels = FALSE no labels at all are plotted." Using the example from the help file,

hc <- hclust(dist(USArrests), "ave")
plot(hc, labels=FALSE)
Jean V. Adams
  • 4,634
  • 2
  • 29
  • 46