[R] Rendering Dendrograms

Thomas Petzoldt thpe at simecol.de
Tue Nov 25 15:52:58 CET 2008

Hi Mark,

similar questions were at least two times during the last weeks, see





or search the archives yourself:


search for "cluster graph" or "plot.hclust"

Thomas P.

Mark Cook wrote:
> Hello all
> I've been using the hclust and as.dendrogram objects for hierarchical
> clustering. The problem I have is that my sample set is now so large (circa
> 500 points) that it isn't possible to view the leaf nodes.
> I'd like to be able to zoom in on specific areas of the graph by 
> selecting a
> region with the mouse.  I've deduced from trial and error that the xlim and
> ylim parameters don't work with these plots and the only method I've had 
> any
> success with is to cut the tree at a specific point and display the lower
> region.  However, the cut function returns a list of all subtrees below the
> specified height and I have no way of selecting a particular subtree (that
> is, cutting on the x-axis).
> The identify and rect.hclust functions seem to do this OK  and  in fact 
> these
> partially solve my problem but I want to use getGraphicsEvent() if possible
> since this also allows interaction with the keyboard (so that I can use the
> backspace key for example to return to the previous plot). I'm fairly 
> new to
> R so may be wrong but it appears that the plots are single threaded and can
> respond to either a set of graphics events or an identify loop but not 
> both)
> It has been suggested that I look at  Graphviz since this has more
> comprehensive mapping facilities. Is this likely to help? Is there for
> example a method for converting an R hclust object to an Graphviz ICLUST or
> some alternative method for describing an hclust in the Graphviz .DOT
> notation?
> Or can anybody suggest an alternative approach ??
> Regards
> Mark
> ______________________________________________
> 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.

More information about the R-help mailing list