Information Technology Reference
In-Depth Information
The required number of dimensions in which to
calculate the
z
that gives the optimal PCA as
described in Section 8.8.
r
Controlling colouring of CLPs associated with ordinal
categories.
ord.col
NULL (the default) or integer-valued vector
specifying which sample points are to be drawn.
plot.samples
If set to TRUE, the ordering of an ordered
categorical variable is reversed. Default is
FALSE.
reverse
Specifies colour to use for plotting the sample points.
samples.col
Specifies plotting symbol for representing sample
points. Default is 15.
samples.pch
Specifies size of plotting character representing
sample points. Default is 1.
samples.size
An integer value controlling placement of label of a
sample point: 1
=
below, 2
=
to left, 3
=
above
and 4
=
to right of specified coordinates. Default
is 1.
samples.label.pos
Specifies colouring of labels of sample points.
samples.label.col
A numerical value controlling the offset from the
position specified by
samples.label.pos
.
Default is no offset.
samples.label.offset
A numerical value for controlling the size of the
plotting symbol representing sample points.
Defaults to 0.75.
samples.label.size
Character vector specifying the classes for which
α
-bags are to be drawn.
specify.bags
Factor controlling width of connecting lines for
ordered and nominal categorical variables. Defaults
to 2.
w.factor
Value
CATPCAbipl
constructs the categorical PCA biplot described in Section 8.8 as well as
a graphical representation of the optimal scores. In addition, a list is returned with the
following components:
The
H
matrix as defined in Section 8.8.
Hmat
The centred
H
matrix of Section 8.8.
Hmat.cent
A list with each component a matrix containing the details of a biplot
axis to be plotted.
z.axes
The
r
-dimensional Eckart - Young PCA solution of the catogorical
PCA criterion (8.16).
Yr
The optimal z-scores described in Section 8.8.
zlist
