You are on page 1of 16

Linear Algebra and Differential Equations

Lecture 10
Linear transformations of R2

Jun 28, 2016

Linear Algebra and Differential EquationsLecture 10Linear transformations


Jun 28, 2016
of R2

1 / 17

Linear Transformations in R2
A transformation f from R2 R2 (i.e takes a vector as input and outputs
a vector) is linear if
f (au + bv) = a f (u) + b f (v)
One consequence of linaerity is that the zero vector is invariant i.e for any
linear function f , f (0) = 0.
Proof
f (0) = f (v v) = f (v) f (v) = 0
Another propert of linear functions is that if b1 and b2 are basis vectors
then we can find the action of f on any vector in R2 in terms of its
components with respect to the basis and the action of f on the basis
vectors.
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016
of R2

2 / 17

If


f (b1 ) =

b11
b21


f (b2 ) =

b12
b22

and if v = 1 b1 + 2 b2 then

f (v) = 1 f (b1 ) + 2 b2 = 1

b11
b21


+ 2

b12
b22

Thi can be written in the form



f (v) =

b11 b12
b21 b22



1
2

Given a basis B = {b1 , b2 } the action of a linear function on vectors


reduces to matrix multiplication. The matrix


b11 b12
[f ]B =
b21 b22
is the representation of f with respect to the basis B.
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016
of R2

3 / 17

An Example
Let E be the standard basis {e1 , e2 } for R2
 
 
1
0
e1 =
e2 =
0
1
and f be the linear transformation with


 
2
1
f (e1 ) =
f (e2 ) =
1
4
so the matrix representation of f in the standard basis E is


2 1
[f ]E =
1 4

Linear Algebra and Differential EquationsLecture 10Linear transformations


Jun 28, 2016
of R2

4 / 17

If we choose a second basis B = {b1 , b2 } with


 
 
3
0
b1 =
b2 =
1
2

f (b1 ) =

f (b2 ) =

2 1
1 4



2 1
1 4



3
1

0
2

7
1

2
8

=
=

and the matrix representation of f in the basis B is




7 2
[f ]B =
1 8

Linear Algebra and Differential EquationsLecture 10Linear transformations


Jun 28, 2016
of R2

5 / 17

Special Linear Transformations


Unless otherwise stated we will be representing linear functions with
respect to the standard basis. Since linear transformations must fix the
origin, translations are ruled out. Examples of linear transformations are
stretches, shears, rotations about the origin, and reflections in lines
through the origin.
To generate the matrix representations we need to look at the images of
the points (1,0) and (0,1) under the transformation, since these points
correspond to the basis vectors in the standard basis.
For a function that stretches by a factor of r in the x direction and a
factor of s in the y direction,the image of the point (1.0) is (r , 0) and the
image of (0,1) is (0, s),
 
 
r
0
f (e1 ) =
f (e2 ) =
0
s
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016
of R2

6 / 17

and the matrix representation is



[f ]E = F =

r 0
0 s

The image of the unit square with vertices (0,0),(1,0),(1,1),(0,1) is the


rectangle with vertices (0,0),(r , 0), (r , s), (0, s) and it is clear that one
effect of the transormation is to modify areas by a factor of |rs|.
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016
of R2

7 / 17

Shear in the x direction


In this deformation (1,0) remains fixed and (0,1) gets shifted x to the
right i.e. to the point (x,1) the matrix representation is


1 x
F=
0 1

The image of the unit square is the parellogram with vertices


(0,0),(1, 0), (1 + x, 1), (x, 1) and the transformation doesnt change the
area
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016
of R2

8 / 17

Shear in the y direction


In this deformation (0,1) remains fixed and (1,0) gets shifted y upwards
i.e. to the point (1.y ) the matrix representation is


1 0
F=
y 1

The image of the unit square is the parellogram with vertices


(0,0),(1, y ), (1, 1 + y ), (0, 1) and the transformation doesnt change the
area
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016
of R2

9 / 17

Effect of linear transformations on area


If f and g are two linear transformations of R2 then so is the composition
fog where v gets mapped to f (g (v)). If we consider a region R of the
plane then
Area of f (g (R))
Area of g (R)
Area of f (g (R))
=

Area of R
Area of g (R)
Area of R
So if f multiplies aras by f and g multiplies areas by g then fog will
multiply them by f g
A general linear transformation can be decomposed into a shear followed
by a stretch followed by a different shear. If a 6= 0

 



a b
1 0
r 0
1 x
A=
=
c d
y 1
0 s
0 1

Linear Algebra and Differential EquationsLecture 10Linear transformations


Jun 28, 2016of R2 10 / 17

Multiplying it out gives




a b
c d


=

r
ry

rx
s + rxy

thus

c
b
y=
a
a
The area magnification factor is given by
r =a

x=

s=d

bc
a

Area magnification factor for A = 1 |rs| 1 = |ad bc| = |det(A)|


So we have a geometric interpretation for the determinant as the
magnification factor for area.
Because of way the composition of functions affects area we can deduce
det(AB) = det(A)det(B)
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016of R2 11 / 17

det(AB)=det(A)det(B)
If




a11 a12
b11 b12
A=
B=
a21 a22
b21 b22


a11 b11 + a12 b21 a11 b12 + a12 b22
AB =
a21 b11 + a22 b21 a21 b12 + a22 b22

det(AB) = (a11 b11 + a12 b21 )(a21 b12 + a22 b22 )


(a11 b12 + a12 b22 )(a21 b11 + a22 b21 )
= a11 b11 a21 b12 + a11 b11 a12 b22 + a12 b21 a21 b12 + a12 b21 a22 b22
a11 b12 a21 b11 a11 b12 a22 b21 a12 b22 a21 b11 a12 b22 a22 b21
= a11 b11 a22 b22 + a12 b21 a21 b12 a11 b12 a22 b21 a12 b22 a21 b11
= a11 a22 (b11 b22 b12 b21 ) a12 a21 (b11 b22 b21 b12 )
= (a11 a22 a12 a21 )(b11 b22 b12 b21 )
= det(A)det(B)
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016of R2 12 / 17

Rotations
A rotation through a counter clockwise angle about the origin rotates e1
to (cos , sin ) and e2 to ( sin , cos ). Thus the matrix corresponding
to the rotation


cos sin
R =
sin cos

Since R+ = R R
 



cos sin
cos sin
cos( + ) sin( + )
=
sin cos
sin cos
sin( + ) cos( + )


cos cos sin sin sin cos cos sin
=
sin cos + cos sin cos cos sin sin
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016of R2 13 / 17

Reflections
Reflection in a line which makes an angle with the x axis sends (1, 0) to
(cos 2, sin 2) and (0, 1) to (sin 2, cos 2)

And thus has the matrix representation




cos 2
sin 2
P =
sin 2 cos 2
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016of R2 14 / 17

Badly behaved Linear transformations


Consider the linear transformation given by


1
5
M=
3 15
If M acts on a general vector
  



x
x + 5y
1
M
=
= (x + 5y )
y
3x 15y
3
No matter what vector we choose we end up with a vector proportional to
[1, 3]T . Now if we look at
   
5
0
M
=
1
0
Linear Algebra and Differential EquationsLecture 10Linear transformations
Jun 28, 2016of R2 15 / 17

Any vector proportional to [5, 1]T gets annihilated by M. These vectors


are called the null space or kernel of M.
So if we want to solve the system

Mx =
One solution is


M

3
1

2
6


2
6

But we can add any member of the null space to it, so the general solution
is given by
 
 
3
5
x=
+C
1
1

Linear Algebra and Differential EquationsLecture 10Linear transformations


Jun 28, 2016of R2 16 / 17

You might also like