Information Technology Reference
In-Depth Information
r
0
−
2
g
RGF
0
5.5
q
0.1
SLF
u
h
v
5
c
2
j
m
3.5
p
3
4.5
2
2.5
i
f
4
1.5
n
b
1
0.2
a
0.5
4
6
0
d
k
w
3.5
s
e
8
SPR
0.3
3
t
PLF
Figure 2.20
The two-dimensional biplot of Figure 2.18, illustrating the use of the biplot
axes for prediction.
(
cx
,
cy
) of the centroid of the vertices of this polygon are calculated using the formula
1
n
n
n
1
n
(
cx
,
cy
)
=
x
i
,
y
i
.
(2.16)
i
=
1
i
=
1
The centroid of the polygon (quadrangle in Figure 2.23) is then drawn together with the
arrows pointing to the interpolated position of the new sample point.
As with prediction, we may shift the interpolation biplot axes to positions that are
more convenient. The process depends on the observation that for any set of markers
µ
1
,
µ
2
,
µ
3
,
...
µ
p
,
,
p
p
x
V
r
=
[
(
x
k
−
µ
k
)(
e
k
V
r
)
+
µ
k
(
e
k
V
r
)
]
=
[
(
x
k
−
µ
k
)(
e
k
V
r
)
]
+
p
µ
,
(2.17)
k
=
1
k
=
1
where
p
1
p
µ
=
1
µ
k
(
e
k
V
r
)
k
=
