Information Technology Reference
In-Depth Information
Gender
Hair
George
Manual
Jeremy
England
Male
Grey
Dark
Harriet
Ivor
School
Alisdair
Postgrad
Education
University
Clerical
Professional
Wales
Brown
Scotland
Female
Fair
Work
Myfanwy
Jane
Region
Figure 8.11
Categorical PCA. Scales on axes marked with solid black circles are
labelled in red according to category. Each axis is in different colours, indicating that
the scales are not continuous. Predictions are shown for
Jane
.
8.9 Functions for constructing MCA related biplots
8.9.1 Function
cabipl
In Section 8.2 we remarked that MCA may be regarded as a CA of the categorical data
matrix
G
as if it were a two-way contingency table. Although we also expressed our
reservations about this practice, it does open up the use of
cabipl
for constructing MCA
biplots. Therefore, all the facilities of
cabipl
as discussed in Section 7.6.1 are available
when the indicator matrix is used as input. However, some notational precautions must
be kept in mind. These are summarized in Table 8.4.
8.9.2 Function
MCAbipl
This is our main function for constructing the two-dimensional MCA biplots based on
the Burt matrix and the EMC discussed in Sections 8.3 and 8.4, respectively.
