Information Technology Reference
In-Depth Information
Arguments
A two-component vector consisting of the coordinates of
the point to be projected. The coordinates must be in
terms of the scaffolding axes, e.g. the first two elements
in a row of the element
Z
returned by
Nonlinbipl
.
ToProject
The point of concurrency of the axes in the form of a
two-component vector consisting of the coordinates in
terms of the scaffolding axes.
g
A list with elements in the form of matrices: one matrix for
each trajectory. Each of these matrices has two columns
representing the two-dimensional coordinates for
constructing the trajectory.
Biplot.axis
A list with elements in the form of vectors: one vector for
each trajectory. Each of these vectors contains the
calibrations associated with a trajectory. Calibrations
correspond to the rows of the respective matrices in
Biplot.axis
.
Biplot.markers
Value
A list with the following two elements:
A
p
-vector, where
p
represents the number of biplot
trajectories. Each element in the vector is the prediction
for the (sample) for that particular variable.
Markers
The coordinates on the circle predictive biplot trajectory of
the corresponding prediction in
Markers
.
Coordinates
5.7 Examples
5.7.1 A PCA biplot as a nonlinear biplot
In Section 5.5 we showed under what circumstances a PCA biplot can be regarded as a
nonlinear biplot. Here, we use the
Ocotea
data of Table 3.9 to illustrate how to obtain a
PCA biplot with
Nonlinbipl
in comparison with the output of
PCAbipl
. These biplots
are given in Figure 5.25 and are constructed with the function calls
PCAbipl(X = Ocotea.data[,3:8], scaled.mat = TRUE,
colours = "blue", pch.samples = 15, pos = "Hor", offset =
c(-0.3, 0.1, 0.1, 0.1), offset.m = rep(-0.2, 6),
reflect = "x")
Nonlinbipl(X = Ocotea.data[,3:8], dist = "Pythagoras",
scaled.mat = TRUE, colours = "blue", pch.samples = 15)
