Graphics Reference
In-Depth Information
Figure
.
.
Projection of the German election data onto the first two principal components. he arrows
give the projected directions of unit vectors corresponding to the five original variables
Software
11.1.2
All of the figures presented in this chapterhave been created with R (RDevelopment
Core Team,
; Murrell,
) using the extension packages
cluster
,
color-
space
,
ellipse
,
flexclust
,
flexmix
,
fpc
,
gplots
,
grid
,
kohonen
,
lattice
,
mvtnorm
,and
vcd
.All ofthose are available fromthe comprehensive R
archivenetworkathttp://cran.R-project.org.hereferencesassociatedwiththepack-
ages are usually given in the text when actually used. he two data sets are contained
in package
flexclust
, while the R code used to generate all figures is available
from the author's homepage.
Hierarchical Cluster Analysis
11.2
Hierarchical clustermethods are probably the “most intuitive” approach to grouping
data, because they approach the problem in a similar way to how a human would
