Learn R
clusterfly (version 0.4)

ggobi.hierfly: Visualise hierarchical clustering with GGobi. Displays both data and dendrogram in original high-d space.

Description

This adds four new variables to the original data set:

Usage

"ggobi"(data, ...)

Arguments

data
hierfly object to visualise in GGobi
...
ignored

Details

  • ORDER, the order in which the clusters are joined
  • HEIGHT, the height of the branch, ie. the dissimilarity between the branches
  • LEVEL, the level of the branch
  • POINTS, the number of points in the branch

Make sure to select "attach edge set (edges)" in the in the edges menu on the plot window, when you create a new plot.

A tour over the original variables will show how the clusters agglomerate in space. Plotting order vs height, level or points will give various types of dendograms. A correlation tour with height/level/points on the y axis and the original variables on the x axis will show a mobile blowing in the wind.

See Also

cut.hierfly

Examples

Run this code
h <- hierfly(iris)
ggobi(h)
h <- hierfly(iris, method="single")

Run the code above in your browser using DataLab