Global Positioning System Reference
In-Depth Information
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
with a cofactor matrix for the residuals
P
−
1
AN
−
1
A
T
Q
v
=
−
(4.337)
Co
mpute the trace
Tr
I
AN
−
1
A
T
P
Tr
(
Q
v
P
)
=
−
Tr
N
−
1
A
T
PA
(4.338)
=
n
−
=
n
−
u
A more general expression is obtained by noting that the matrix
AN
−
1
A
T
P
is idem-
potent. The trace of an idempotent matrix equals the rank of that matrix. Thus,
Tr
AN
−
1
A
T
P
=
[15
R
A
T
PA
=
R(
A
)
=
r
≤
u
(4.339)
Lin
—
1.3
——
Nor
PgE
Thus, from Equations (4.338) and (4.339)
Tr
(
Q
v
P
)
=
Tr
(
PQ
v
)
=
n
−
R(
A
)
(4.340)
By denoting the diagonal element of the matrix
Q
v
P
by
r
i
, we can write
n
r
i
=
n
−
R(
A
)
(4.341)
i
=
1
[15
The sum of the diagonal elements of
Q
v
P
equals the degree of freedom. The element
r
i
is called the redundancy number for the observation
i
. It is the contribution of the
i
th observation to the degree of freedom. If the weight matrix
P
is diagonal, which is
usually the case when original observations are adjusted, then
=
r
i
q
i
p
i
(4.342)
where
q
i
is the diagonal element of the cofactor matrix
Q
v
, and
p
i
denotes the weight
of the
i
th observation. Equation (4.337) implies the inequality
1
p
i
0
≤
q
i
≤
(4.343)
Multiplying by
p
i
gives the bounds for the redundancy numbers,
0
≤
r
i
≤
1
(4.344)
Considering the general relation
Q
=
Q
b
−
Q
v
(4.345)
a