Transcript A x
m
4.2 Linear Transformations from Rn to R
Functions from Rn to R
A function is a rule f that associates with each element in a set A one and only
one element in a set B.
If f associates the element b with the element a, then we write b = f(a) and say
that b is the image of a under f or that f(a) is the value of f at a.
The set A is called the domain of f and the set B is called the codomain of f.
The subset of B consisting of all possible values for f as a varies over A is
called the range of f.
Function from
Rn
to R m
If the domain of a function f is R n and the codomain is R m , then f is called a
m
map or transformation from R n to R . We say that the function f maps Rn
Into R m , and denoted by f : R n → R m .
If m = n the transformation f :
Rn → Rn is called an operator on Rn.
Suppose f1, f2, …, fm are real-valued functions of n real variables, say
w1 = f1(x1,x2,…,xn)
w2 = f2(x1,x2,…,xn)
…
wm = fm(x1,x2,…,xn)
These m equations assign a unique point (w1,w2,…,wm) in R m to each point
(x1,x2,…,xn) in R n and thus define a transformation from R n to R m . If we
m
n
denote this transformation by T: R → R then
T (x1,x2,…,xn) = (w1,w2,…,wm)
Example: The equations
w1 x1 x2
w2 3 x1 x2
w3 x12 x22
Defines a transformation T : R 2 R 3 .
With this transformation, the image of the point (x1, x2) is
T ( x1 , x2 ) ( x1 x2 ,3x1 x2 , x12 x22 )
Thus, for example, T(1, -2)=(-1, -6, -3)
Linear Transformations from
Rn
to R m
A linear transformation (or a linear operator if m = n) T:
defined by equations of the form
w1 a11 x1 a12 x2 ... a1n xn
w2 a21 x1 a22 x2 ... a2 n xn
...
wm am1 x1 am 2 x2 ... amn xn
or
R n→ R m
w1 a11 a12
w a
2 21 a22
wm am1 am 2
is
a1n x1
a2 n x2
amn xm
or
w = Ax
The matrix A = [aij] is called the standard matrix for the linear transformation T,
and T is called multiplication by A.
Example: If the linear transformation T : R 4 R 3 is defined by the equations
w1 2 x1 3x2 x3 5 x4
w2 4 x1 x2 2 x3 x4
w3 5 x1 x2 4 x3
Find the standard matrix for T, and calculate T (1, 3, 0, 2)
Solution: T can be expressed as
x1
w1 2 3 1 5
w 4 1 2 1 x2
2
x
w3 5 1 4 0 3
x4
So the standard matrix for T is
2 3 1 5
A 4 1 2 1
5 1 4 0
Furthermore, if ( x1 , x2 , x3 , x4 ) (1, 3,0, 2)
w1 2 x1 3x2 x3 5 x4 1
w2 4 x1 x2 2 x3 x4 3
w3 5 x1 x2 4 x3 8
Thus, T (1, 3, 0, 2) (1,3,8)
Or
1
w1 2 3 1 5 1
w 4 1 2 1 3 3
2
0
w3 5 1 4 0 8
2
Remarks
Notations:
If it is important to emphasize that A is the standard matrix for T. We denote the linear
transformation T: R n → R m by TA: R n → R m . Thus, TA(x) = Ax
We can also denote the standard matrix for T by the symbol [T], or T(x) = [T]x
Remark:
We have establish a correspondence between m×n matrices and linear
transformations from R n to R m :
To each matrix A there corresponds a linear transformation TA (multiplication by A), and
to each linear transformation T: R n → R m , there corresponds an m×n matrix [T] (the
standard matrix for T).
Examples
m
n
Zero Transformation from R to R
If 0 is the m×n zero matrix and 0 is the zero vector in Rn, then for every
vector x in Rn
T0(x) = 0x = 0
n
So multiplication by zero maps every vector in R into the zero vector in
.R m . We call T0 the zero transformation from Rn to R m .
Identity operator on Rn
If I is the n×n identity, then for every vector in Rn
TI(x) = Ix = x
So multiplication by I maps every vector in Rn into itself. We call TI the
identity operator on Rn .
Reflection Operators
2
In general, operators on R and R 3 that map each vector into its symmetric
image about some line or plane are called reflection operators.
Such operators are linear.
Projection Operators
In general, a projection operator (or more precisely an orthogonal
projection operator) on R2 or R 3 is any operator that maps each
vector into its orthogonal projection on a line or plane through the origin.
The projection operators are linear.
Rotation Operators
2
An operator that rotate each vector in R through a fixed angle θ is called a
rotation operator on R2 .
Operator
Rotation
through an
angle
Illustration
( w1 , w2 )
( x, y )
Equations
Standard
Matrix
w1 x cos y sin
cos
sin
w2 x sin y cos
sin
cos
Example: Use matrix multiplication to find the image of the vector (1, 1) when it is
rotated through an angle of 30 degree ( / 6 )
x
x
Solution: the image of the vector
y is
cos / 6 sin / 6 x 3 / 2
w
y
sin
/
6
cos
/
6
1/ 2
3
1
x
1/ 2 x 2
2
3 / 2 y 1
3
x
2
2
So
3/2
w
1/ 2
3 1
1/ 2 1 2
3 / 2 1 1 3
2
y
y
Dilation and Contraction Operators
2
If k is a nonnegative scalar, the operator on R or R 3 is called a contraction
with factor k if 0 ≤ k ≤ 1 and a dilation with factor k if k ≥ 1 .
Operator
Contraction
with factor k
on R 2
(0 ≤ k ≤ 1 )
Dilation with
factor k on R 2
(k ≥ 1 )
Illustrator
Equations
( x, y )
w1 kx
w2 ky
( kx, ky )
( kx, ky )
( x, y )
w1 kx
w2 ky
Standard
Matrix
k 0
0 k
k 0
0 k
Compositions of Linear Transformations
If TA : R n → R k and TB : R k → R m are linear transformations, then for
n
each x in R one can first compute TA(x), which is a vector in R k , and
then one can compute TB(TA(x)), which is a vector in R m .
Thus, the application of TA followed by TB produces a transformation from
to R m . This transformation is called the composition of TB with TA and is
denoted by TB TA . Thus (TB TA )(x) (TB (TA (x))
The composition TB TA
is linear since
(TB TA )(x) (TB (TA (x)) B( Ax) ( BA)x
The standard matrix for TB TA is BA. That is, TB TA TBA
Rn
Remark:
TB TA TBA captures an important idea: Multiplying matrices is equivalent to
composing the corresponding linear transformations in the right-to-left order of
the factors.
n
k
k
m
Alternatively, If T1 : R R and T2 : R R are linear transformations, then because
the standard matrix for the composition T2 T1 is the product of the standard matrices
of T2 and T1, we have
T2
T1 T2 T1
Example: Find the standard matrix for the linear operator T : R 2 R 2 that first reflects
A vector about the y-axis, then reflects the resulting vector about the x-axis.
Solution: The linear transformation T can be expressed as the composition
T T2 T1
Where T1 is the reflection about the y-axis, and T2 is the reflection about
The x-axis.
1 0
,
1
T1 0
Sine the standard matrix for T is
1
T2 0
T T2
0
1
T1 T2 T1
1 0 1 0 1 0
T
0 1 0 1 0 1
Which is called the reflection about the origin.
Note: the composition is NOT commutative.
Example: Let T1 : R 2 R 2 be the reflection operator about the line y=x, and let
T2 : R 2 R 2 be the orthogonal projection on the y-axis. Then
T1
0 1 0 0 0 1
T2 T1 T2
0 1 0 0
1
0
T2
0 0 0 1 0 0
T1 T2 T1
1 0 1 0
0
1
T2
T1 T1 T2
Thus, T2 T1 and T2 T1 have different effects on a vector x.
4.3 Properties of Linear Transformations form Rn to R m
One-to-One Linear transformations
Definition
n
m
A linear transformation T : R R
is said to be one-to-one if T maps
n
distinct vectors (points) in R into distinct vectors (points) in R m
Remark:
That is, for each vector w in the range of a one-to-one linear transformation
T, there is exactly one vector x such that T(x) = w.
Example:
2
2
The linear operator T: R R that rotates each vector through an angle
a one-to-one linear transformation.
is
In contrast, if T: R 2 R 2 is the orthogonal projection on the x-axis, then it’s
not a one-to-one linear transformation.
Theorem 4.3.1 (Equivalent Statements)
If A is an n×n matrix and TA : R R is multiplication by A, then the
following statements are equivalent.
n
•A is invertible
n
•The range of TA is R
•TA is one-to-one
n
Example
The rotation operator T : R 2 R 2 that rotates each vector through an angle
is one-to-one.
cos sin
T
The standard matrix for T is
sin cos
which is invertible since
det[T ]
cos
sin
sin
cos 2 sin 2 1 0
cos
Example
If T: R 2 R 2 is the orthogonal projection on the x-axis, then it’s
not one-to-one.
1 0
T
The standard matrix for T is
0 0
Which is not invertible since det[T] = 0
Inverse of a One-to-One Linear Operator
Suppose TA : R R is a one-to-one linear operator
⇒ The matrix A is invertible.
⇒ TA-1 : R n R n is itself a linear operator; it is called the inverse of TA.
⇒ T (T 1 (x)) AA1x Ix x
n
A
n
A
TA1 (TA (x)) A1 Ax Ix x
⇒
TA TA1 TAA1 TI
TA1 TA TA1 A TI
If w is the image of x under TA, then TA-1 maps w back into x, since
TA1 (w) TA1 (TA (x)) x
When a one-to-one linear operator on R n is written as T : R n R n , then
the inverse of the operator T is denoted by T 1 .
1
Thus, by the standard matrix, we have T T
1
Example
2
2
Show that the linear operator T : R R
w1= 2x1+ x2
w2 = 3x1+ 4x2
1
is one-to-one, and find T ( w1 , w2 )
defined by the equations
w1 2 1 x1
x
w
3
4
2
2
Solution: The matrix form of these equations is
2 1
3
4
So the standard matrix for T is [T ]
This matrix is invertible and the standard matrix for T 1 is
1
4
5
5
1
1
[T ] [T ]
3 2
5 5
Thus
1
1
4
4
w
w
1
2
5
w
5 w1 5
5
1 1
[T ]
w
3
2
w
3
2
2
2 w w
5 5
5 1 5 2
from which we conclude that
4
1
3
2
T 1 ( w1 , w2 ) ( w1 w2 , w1 w2 )
5
5
5
5
Linearity Properties
Theorem 4.3.2 (Properties of Linear Transformations)
A transformation T : R n R m is linear if and only if the following
relationships hold for all vectors u and v in Rn and every scalar c.
T(u + v) = T(u) + T(v)
T(cu) = cT(u)
Example: Determine whether T: R 2 R 2 is a linear operator if T(x, y)=(x, 3y).
Solution:
T (( x1 , y1 ) ( x2 , y2 )) T ( x1 x2 , y1 y2 ) ( x1 x2 ,3( y1 y2 ))
( x1 ,3 y1 ) ( x2 ,3 y2 ) T ( x1 , y1 ) T ( x2 , y2 )
T (c( x, y )) T (cx, cy ) (cx,3cy) c( x,3 y) cT ( x, y)
So T(x, y)= (x, 3y) is a linear operator.
We call the vectors e1, e2, …, en be the standard basis vectors for Rn if
1
0
0
1
e1 0 , e 2 0 ,
0
0
0
0
, e n 0
1
In particular, In R2 and R 3 the standard basis vectors are the vectors of length 1
Along the coordinate axes.
Theorem 4.3.3
If T : R n R m is a linear transformation, and e1, e2, …, en are the
standard basis vectors for Rn , then the standard matrix for T is
A = [T] = [T(e1) | T(e2) | … | T(en)]
Example:
2
2
Find the standard matrix for T: R R from the images of the standard
Basis vectors if T dilates a vector by a factor of 2, then reflects that vector about the
line y=x, and then projects that vector orthogonally onto x-axis.
Solution: Here
e1 (1, 0) (2, 0) (0, 2) (0, 0) 0e1 0e2
e2 (0,1) (0, 2) (2, 0) (2, 0) 2e1 0e2
Thus
0 2
[T ]
0 0
Eigenvalue and Eigenvector
Definition
n
n
If T: R R is a linear operator, then a scalar λ is called an eigenvalue of T
n
if there is a nonzero x in Rsuch
that
T(x) = λx.
Those nonzero vectors x that satisfy this equation are called the
eigenvectors of T corresponding to λ
Remarks:
If A is the standard matrix for T, then the equation becomes Ax = λx
•
•
The eigenvalues of T are precisely the eigenvalues of its standard matrix A
x is an eigenvector of T corresponding to λ if and only if x is an eigenvector of A
corresponding to λ
If λ is an eigenvalue of A and x is a corresponding eigenvector, then Ax = λx, so
multiplication by A maps x into a scalar multiple of itself
2
3
In R and R, this means that multiplication by A maps each eigenvector x into a
Vector x intro a vector that lies on the same line as x.
x
Ax x
x
0
Ax x
0
Example:
Let T: R 2 R 2 be the reflection about the y-axis. Find the eigenvalues and
corresponding eigenvectors of T. Check your calculations By calculating the
eigenvalues and corresponding eigenvectors from the standard matrix for T.
Solution: This transformation maps vectors on the x-axis to their negatives,
vectors on the y-axis into themselves, and maps no other vectors into
scalar multiples of themselves.
Thus the eigenvalues are λ = ±1 and the eigenvectors are vectors (x, y)
with either x = 0 or y = 0, but not both. To verify this, we observe that since
e1 → -e1 and e2 → e2, the standard matrix for the transformation is
1 0
0 1
. Hence the characteristic equation is
1 0 1 0
det(
)0
0 1 0 1
2
or 1 0 with solutions λ = ±1.
If (x, y) is an eigenvector corresponding to λ = 1,
Then 2 0 x 0
0 0 y 0
2 x 0
0 0
or x = 0, so the vector must lie on the y-axis.
If (x, y) is an eigenvector corresponding to λ = –1, then
0 0 x 0
0 2 y 0
0 0
2 y 0
or y = 0, so the vector must lie on the x-axis.
Theorem 4.3.4 (Equivalent Statements)
If A is an n×n matrix, and if TA : R n R n is multiplication by A, then the following
are equivalent.
• A is invertible
• Ax = 0 has only the trivial solution
• The reduced row-echelon form of A is In
• A is expressible as a product of elementary matrices
• Ax = b is consistent for every n×1 matrix b
• Ax = b has exactly one solution for every n×1 matrix b
• det(A) ≠ 0
• The range of TA is Rn
• TA is one-to-one