Sie sind auf Seite 1von 29

Linear transformations

Textbooks:
o Linear Algebra with Applications, Fifth Edition, by Otto Bretscher

Miguel Moscoso

Summary

What is a linear transformation?


Image and kernel of a linear transformation, and rank-nullity theorem.
The matrix associated to a linear transformation.
Linear transformations in geometry.
The inverse of a linear transformation.
Computation of the inverse matrix.
Isomorphisms

What is a linear transformation?


Definition: A linear transformation T : R R
the following two properties:
n

T u v T u T v u, v R n

T u T u u R n

is a function that satisfies

From this definition we can interpret a linear transformation as a linear vector


function from to . In other words,

is a linear transformation from to if there is an matrix


such that

T ( x ) Ax
for all in .

What is a linear transformation?


Example: Consider a transformation T : R R that rotates a vector
counterclockwise an angle :
Draw a sketch to illustrate that is a linear transformation
2

Image and kernel


Definition: If : , the set is is called the domain and the set the
codomain of the linear transformation.

Do not mix up the domain and the image of a linear transformation!

Definition: The image of a linear transformation : consists of all the


values it takes in its codomain. That is,
= { : = , for some X}.

Fact: The image of a linear transformation


=
is the span of the columns of , i.e., = ().

Image and kernel


Definition: The kernel of a linear transformation : is the set of all
zeros of the transformation. That is
k = { : = = 0}.

Remember that for a linear transformation : ,


() is a subset of the codomain of T, and
ker() is a subset of the domain of T.

Homework: Show that for every linear transformation : :


ker() is a subspace of
() is a subspace of

The rank-nullity theorem, again


Reminder: Given a linear transformation T,
its nullity is the dimension of the kernel, that is, = dim ker ,
its rank is the dimension of the image, that is, rank = dim im .
Theorem: (rank nullity theorem)
Given a linear transformation : with associated matrix

nul T rank T n

The matrix of a linear transformation


Fact: Given any linear transformation : observe that

a1,1 a1,n 0 a1,i


n
T (ei ) Aei

c
,
e

R
i

1 i

a
a

m
,
1
m
,
n
m
,
i

The matrix of a linear transformation


Fact: Given any linear transformation : observe that

a1,1 a1,n 0 a1,i


n
T (ei ) Aei

c
,
e

R
i

1 i

a
a

m
,
1
m
,
n
m
,
i

Theorem: Let : be a linear transformation.


Then, has an associated matrix given by:

A T e1 T en

The matrix of a linear transformation


Example: Consider a transformation : 2 2 that rotates a vector
counterclockwise an angle :
Draw a sketch to illustrate that is a linear transformation
Find the matrix of .

Linear transformations in geometry


Rotations: The matrix of a counterwise rotation through an angle is

Linear transformations in geometry


Rotations: The matrix of a counterwise rotation through an angle is

Rotation-Dilation: The matrix of a counterwise rotation through an angle such

that tan = , followed by a dilation by the factor = 2 + 2 is



Linear transformations in geometry


Definition: Let be a line in generated by the unit vector .
For every vector there is a unique vector on L such
that is perpendicular to , namely, = . This
vector is called the orthogonal projection of onto :

The transformation = ( ) from to is linear.

Linear transformations in geometry


Homework: Let be a line in , show that the reflection of in
is given by the vector
= 2 = 2 ,
where is a unit vector on .

Inverse of a linear transformation

Inverse of a linear transformation


Definition: A linear transformation : is invertible if the
linear system equation = = has a unique solution
for each . The inverse is denoted by 1 .

If : is invertible, then

T 1 :Y X

y x A1 y

Properties of the inverse transformation:

1
o T T x x x X

1
o T T y y y Y
o

1 1

Inverse of a linear transformation


If the linear transformation = = is invertible, its
inverse is
= 1 = 1 .

Computation of the inverse


Example: Find the inverse of

1
3

2
.
9

1
1 + 22
Solution: We consider the linear transformation
= and
3
+
9
2
1
2
solve for and .
1

Computation of the inverse


1 1
Example: Find the inverse of 2 3
3 8

1
2.
2

1
1 + 2 + 3
Solution: We consider the linear transformation 21 + 32 + 22 = 2
3
21 + 82 + 23
and solve for 1 , 2 and 3 .

Computation of the inverse

And therefore,

Computation of the inverse

We can write the previous calculation in matrix form, as

Computation of the inverse

Then, to compute the inverse of a matrix you can:


form the 2 matrix and
compute .
If = , then is invertible and
1 = .
If is of another form (the left half fails to
be _), then is not invertible.

Surjective, injective, bijective


Definitions:

Let T : R m R n be defined by y T x Ax . Then:


T is surjective dim im T rank T m
T is injective dimker T nul T 0
T is bijective T is injective and surjective. In this case we
say that T is an isomorphism and the spaces X and Y are
isomorphic.
Note: T is invertible T is bijective.

Surjective, injective, bijective

Surjective but
not injective

Injective but
not surjective

Neither injective
nor surjective

Bijective

Isomorphisms

Isomorphisms

Isomorphisms

Bases

Coordinates and matrices of with respect to different bases

Next week!!!

Das könnte Ihnen auch gefallen