In hierarchical cluster analysis dendrogram graphs are used to visualize how clusters are formed. I propose an alternative graph named clustergram to examine how cluster members are assigned to clusters as the number of clusters increases. This graph is useful in exploratory analysis for non-hierarchical clustering algorithms like k-means and for hierarchical cluster algorithms when the number of observations is large enough to make dendrograms impractical.

The clustergram is currently implemented in Stata and R.

Downloads:

• (pdf) Schonlau M. The clustergram: a graph for visualizing hierarchical and non-hierarchical cluster analyses. The Stata Journal, 2002; 2 (4):391-402.
  The paper introduces the clustergram and explains how to use the stata ado files.
• Schonlau M. Visualizing Hierarchical and Non-Hierarchical Cluster Analyses with Clustergrams. Computational Statistics: 2004; 19(1):95-111. (pdf)
  This paper points out that the y-axis in the clustergram may take different functions and gives further examples.
• Stata implementation clustergram ZIP File
  

• clustergram ado file
  
• clustergram Help File
  
• clustervar Ado file
  This supplementary ado file makes it easier to run various cluster algorithms. The stata paper describes how to run cluster analysis without using this supplementary ado file.
  
• clustervar help file
  Syntax Example:
  clustervar sepallen-petalwid, algorithm(singlelinkage) max(`max') distance(L2)
• Asbestos.dta (stata data set)
  Asbestos data set used in the paper

R implementation
Tal Galili has implemented the clustergram in R. He gives the code and talks about it in his blog: Clustergram blog as part of www.r-statistics.com

Return to Home Page
Remove navigation bar on the left