本文为《Linear algebra and its applications》的读书笔记
目录
The Matrix of a Linear Transformation
- Let
V
V
V be an
n
n
n-dimensional vector space, let
W
W
W be an
m
m
m-dimensional vector space, and let
T
T
T be any linear transformation from
V
V
V to
W
W
W . To associate a matrix with
T
T
T , choose (ordered) bases
B
\mathcal B
B and
C
\mathcal C
C for
V
V
V and
W
W
W , respectively. Given any
x
\boldsymbol x
x in
V
V
V , the coordinate vector
[
x
]
B
[\boldsymbol x]_{\mathcal B}
[x]B is in
R
n
\mathbb R^n
Rn and the coordinate vector of its image,
[
T
(
x
)
]
C
[ T (\boldsymbol x)]_{\mathcal C}
[T(x)]C, is in
R
m
\mathbb R^m
Rm, as shown in Figure 1.
- Let
{
b
1
,
.
.
.
,
b
n
}
\{\boldsymbol b_1,..., \boldsymbol b_n\}
{b1,...,bn} be the basis
B
\mathcal B
B for
V
V
V. If
x
=
r
1
b
1
+
.
.
.
+
r
n
b
n
\boldsymbol x = r_1\boldsymbol b_1+...+ r_n\boldsymbol b_n
x=r1b1+...+rnbn, then
and
Since the coordinate mapping is linear, equation (1) leads toThe vector equation (2) can be written as a matrix equation, namely,where
The matrix M M M is a matrix representation of T T T, called the matrix for T \boldsymbol T T relative to the bases B \mathcal B B and C \mathcal C C. If B \mathcal B B and C \mathcal C C are bases for the same space V V V and if T T T is the identity transformation, then matrix M M M is just a change-of-coordinates matrix.
Linear Transformations from V V V into V V V
- In the common case where
W
W
W is the same as
V
V
V and the basis
C
\mathcal C
C is the same as
B
\mathcal B
B, the matrix
M
M
M in (4) is called the matrix for
T
T
T relative to
B
\mathcal B
B, or simply the
B
\mathcal B
B-matrix for
T
T
T, and is denoted by
[
T
]
B
[T]_{\mathcal B}
[T]B.
EXERCISES 7
Assume the mapping
T
:
P
2
→
P
2
T :\mathbb P^2 \rightarrow \mathbb P^2
T:P2→P2 defined by
is linear. Find the matrix representation of
T
T
T relative to the basis
B
=
{
1
,
t
,
t
2
}
\mathcal B = \{1, t, t^2\}
B={1,t,t2}.
SOLUTION
- A faster way is to realize that the information given provides the general form of
T
(
p
)
T(p)
T(p) as shown in the figure below:
The matrix that implements the multiplication along the bottom of the figure is easily filled
in by inspection:
Since [ T ] B [T]_{\mathcal B} [T]B can be represented by a matrix, [ T ] B [T]_{\mathcal B} [T]B is linear. Thus T T T is linear. ( [ x ] B [\boldsymbol x]_{\mathcal B} [x]B and [ x ] C − 1 [\boldsymbol x]_{\mathcal C}^{-1} [x]C−1 are all linear)
Linear Transformations on R n \mathbb R^n Rn
- In an applied problem involving R n \mathbb R^n Rn, a linear transformation T T T usually appears first as a matrix transformation, x ↦ A x \boldsymbol x \mapsto A\boldsymbol x x↦Ax. If A A A is diagonalizable, then there is a basis B \mathcal B B for R n \mathbb R^n Rn consisting of eigenvectors of A A A. Theorem 8 below shows that, in this case, the B \mathcal B B-matrix for T T T is diagonal. Diagonalizing A A A amounts to finding a diagonal matrix representation of x ↦ A x \boldsymbol x \mapsto A\boldsymbol x x↦Ax.
- The mappings x ↦ A x \boldsymbol x \mapsto A\boldsymbol x x↦Ax and u ↦ D u \boldsymbol u \mapsto D\boldsymbol u u↦Du describe the same linear transformation, relative to different bases.
Similarity of Matrix Representations
矩阵表示的相似性
- The proof of Theorem 8 did not use the information that
D
D
D was diagonal. Hence, if
A
A
A is similar to a matrix
C
C
C, with
A
=
P
C
P
−
1
A = PCP^{-1}
A=PCP−1, then
C
C
C is the
B
\mathcal B
B-matrix for the transformation
x
↦
A
x
\boldsymbol x \mapsto A\boldsymbol x
x↦Ax when the basis
B
\mathcal B
B is formed from the columns of
P
P
P.
- Conversely, if T : R n → R n T :\mathbb R^n \rightarrow \mathbb R^n T:Rn→Rn is defined by T ( x ) = A x T(\boldsymbol x)= A\boldsymbol x T(x)=Ax, and if B \mathcal B B is any basis for R n \mathbb R^n Rn, then the B \mathcal B B-matrix for T T T is similar to A A A. In fact, the calculations in the proof of Theorem 8 show that if P P P is the matrix whose columns come from the vectors in B \mathcal B B, then [ T ] B = P − 1 A P [T]_{\mathcal B} = P^{-1}AP [T]B=P−1AP.Thus, the set of all matrices similar to a matrix A A A coincides with the set of all matrix representations of the transformation x ↦ A x \boldsymbol x \mapsto A\boldsymbol x x↦Ax.
Tip
- The t r a c e trace trace of a square matrix A A A is the sum of the diagonal entries in A A A and is denoted by t r A trA trA. It can be verified that t r ( F G ) = t r ( G F ) tr(FG)= tr(GF) tr(FG)=tr(GF) for any two n × n n \times n n×n matrices F F F and G G G. It can be shown that if A A A and B B B are similar, then t r A = t r B trA = trB trA=trB. If B B B is diagonal, then the diagonal entries of B B B must be the eigenvalues of A A A. So t r A = t r B = { s u m o f t h e e i g e n v a l u e s o f A } tr A = tr B = \{sum\ of\ the\ eigenvalues\ of\ A\} trA=trB={sum of the eigenvalues of A}.
- In general, it can be shown that for any square matrix A A A, the trace of A A A is the sum of the eigenvalues of A A A, counted according to multiplicities. You can use this fact to provide a quick check on your eigenvalue calculations. The sum of the eigenvalues must match the sum of the diagonal entries in A A A (even if A A A is not diagonalizable).