Linear independence Linear transformations

MATH 4A - Linear Algebra with Applications

Lecture 5: Linear independence and linear transformations

10 April 2019

Reading: §1.7-1.9 from Lay, 5th ed.

Recommended problems from §1.7: 1, 3, 5, 7, 9, 11, 13, 21-23, 27,
33, 35, 39
Recommended problems from §1.8: 1, 3, 7, 9, 11, 17, 21, 22, 29
Announcement: you can skip the sections 1.6 and 1.10 on
applications for now. We will double back on them in two weeks.
Linear independence Linear transformations

Lecture plan

1 Linear independence

2 Linear transformations
Linear independence Linear transformations

Motivation

Every linear system in n variables with more than one solution has
infinitely many solutions. Moreover the set of solutions is a
k-dimensional hyperplane in Rn of some dimension k. However, we
don’t yet know how to compute k (except in specific examples
maybe).

Linear independence is a key concept that eventually will allow us

to compute k.

I would argue that linear independence and linear dependence are

the single most important concepts in linear algebra. So pay
attention today!
Linear independence Linear transformations

Definition

A set of vectors {v1 , v2 , . . . , vp } in Rn is linearly independent if the

homogeneous vector equation

x1 v1 + x2 v2 + · · · + xp vp = 0

has only the trivial solution. The set {v1 , v2 , . . . , vp } is linearly

dependent if it is not linearly independent.
Linear independence Linear transformations

Equivalent definitions of linearly dependent

In other words, {v1 , v2 , . . . , vp } is linearly dependent if there exist

weights c1 , c2 , . . . , cp , not all zero, such that

c1 v1 + c2 v2 + · · · + cp vp = 0.

Another equivalent definition: {v1 , v2 , . . . , vp } is linearly

dependent if the homogeneous vector equation

x1 v1 + x2 v2 + · · · + xp vp = 0

has a nontrivial solution.

Linear independence Linear transformations

Linear dependence relations

If the set {v1 , v2 , . . . , vp } is linearly dependent, a choice of

c1 , c2 , . . . , cp , not all zero, such that

c1 v1 + c2 v2 + · · · + cp vp = 0

is called a linear dependence relation among v1 , v2 , . . . , vp .

Linear independence Linear transformations

Small sets: one vector, i.e. p = 1

Let v1 be some vector in Rn . When is the set {v1 } consisting of
only the one vector v1 linearly independent?

Well, if v1 6= 0, the single variable homogeneous linear system

x1 v1 = 0

has no nontrivial solutions. So {v1 } is a linearly independent set if

v1 6= 0.

On the other hand, if v1 = 0, we can plug any scalar in for x1 in

the equation
x1 v1 = 0
and the result is true. Thus the system has nontrivial solutions.
We conclude {v1 } is a linearly dependent set if v1 = 0. Every
choice of x1 6= 0 yields a linear dependence relation.
Linear independence Linear transformations

Small sets: two vectors, i.e. p = 2

Consider a set {v1 , v2 } consisting of two vectors in Rn . When is
this set linearly independent?

Well, we can always rewrite the vector equation

x1 v1 + x2 v2 = 0

as
x1 v1 = −x2 v2 .
Given any nontrivial solution, one of x1 or x2 must be nonzero.
Let’s assume x2 is nonzero. Then we can rewrite the equation as
x1
v2 = − v1 .
x2
Conclusion: {v1 , v2 } is linearly dependent if and only if one of the
vectors is a multiple of the other vector. {v1 , v2 } is linearly
independent if and only if neither vector is a multiple of the other.
Linear independence Linear transformations

In general: linear dependence can be recast in terms of a

homogeneous linear system

More precisely, a linear dependence releation among v1 , v2 , . . . , vp

is exactly the same thing as a nontrivial solution to the
homogeneous system:
Ax = 0
where A is the matrix


A = v1 v2 · · · vp

Thus, in general, we can use matrix methods (i.e. row operations,

reduced echelon forms, etc. ) to decide if a set of vectors is linearly
independent or not.
Linear independence Linear transformations

Linear independence of matrix columns

In other words:

The columns of a matrix A are linearly independent if and only if

the equation Ax = 0 has only the trivial solution.
Linear independence Linear transformations

iClicker 1

Consider the vectors

     
1 4 2
v1 = 0 v2 = −3 v3 = −3
0 0 0

Is the set {v1 , v2 , v3 } linearly dependent?

(a) Yes
(b) No
(c) Can’t nobody tell me nothin’
(d) You can’t tell me nothin’
(e) I’m gonna ride ’til I can’t no more
Linear independence Linear transformations

Characterizing linearly dependent sets in terms of spans

Theorem
A set {v1 , . . . , vp } of two or more vectors in Rn is linearly
dependent if and only if at least one of the vectors is a linear
combination of the others. In fact, if the vectors are linearly
dependent and v1 6= 0, then some vj (with j > 1) is a linear
combination of the preceding vectors v1 , . . . , vj−1 .

In other words, assuming v1 6= 0, then the set {v1 , . . . , vp } is

linearly dependent if and only if some vj (with 1 < j ≤ p) is in
Span{v1 , . . . , vj−1 }.
Linear independence Linear transformations

Example
Linear independence Linear transformations

Linear dependence “for free”

Sometimes it’s pretty easy to see that a set of vectors must be
linearly dependent.
Theorem
Let {v1 , . . . , vp } be a set of vectors in Rn with p > n. Then the
set must be linearly dependent.


Proof: Let A = v1 · · · vp . Then the equation Ax = 0 is a
system of n equations in p variables. Since p > n, there must be a
free variable.
Theorem
Let {v1 , . . . , vp } be a set of vectors in Rn . If at least one of the
vectors vi for some i = 1, . . . , p equals the 0 vector, then the set
{v1 , . . . , vp } is linearly dependent.

Proof: 0v1 + · · · + 0vi−1 + 1vi + 0vi+1 + · · · + 0vp = 0 is a linear

dependence.
Linear independence Linear transformations

iClicker 2

Are the vectors

       
1 432 1 532
 8   1  2 734
       
 234  −347e 3  3  1 
−4213π 4 4 1

linearly independent?
(a) Yes
(b) No
Linear independence Linear transformations

iClicker 3

Are the vectors      

2 0 1
3 0 1
5 0 1
linearly independent?
(a) Yes
(b) No
Linear independence Linear transformations

iClicker 4

Are the vectors    

32 64
4  8 
   
8  16 
13 −26
linearly independent?
(a) Yes
(b) No
Linear independence Linear transformations

Matrices as functions

So far, we’ve simply thought of matrices as convenient book

keeping tools. Namely, matrices provide a compact notation for
expressing linear systems.

However, we can also think of a matrix “dynamically:” if A is a

m × n matrix, it “converts” a vector x in Rn into a vector Ax in
Rm . Solving the equation Ax = b is equivalent to asking: which
vectors x get converted into b by A?
Linear independence Linear transformations

Generalities on functions

A transformation (or function or mapping) T from Rn to Rm is a

rule that converts each vector x in Rn to a vector T (x) in Rm .

If T is a transformation from Rn to Rm , we call Rn the domain of

T and Rm the codomain of T .

Given x ∈ Rn , we call the vector T (x) in Rm the image of x under

T.

The set of all vectors in Rm that are the image of some vector in
Rn is called the range of T .
Linear independence Linear transformations

A schematic picture of a transformation

range

T (x)
T
x

Rn
Rm
domain codomain
Linear independence Linear transformations

Matrix transformation example 1: projection

 
1 0 0
If A = 0 1 0, the transformation x 7→ Ax projects points in
0 0 0
R3 onto the x1 x2 -plane inside of R3 .
Linear independence Linear transformations

Matrix transformation example 1: demonstrating

projection with the room’s projector

The projector takes things (vectors) to their shadow (image) on

the projection screen (x1 x2 -plane). The line pointing from the
screen to the projector bulb is parallel to the x3 axis. (To be
precise, we should assume the projector is infinitely far away from
the screen. . . )
Linear independence Linear transformations

Matrix transformation example 2: here’s a sheep

Linear independence Linear transformations

Matrix transformation example 2: here’s a sheared sheep

Linear independence Linear transformations

Matrix transformation example 2: what just happened?

I applied the transformation

 
1 .6
0 1

from R2 to R2 . This is an example of a shear transformation.

So, we sheared the sheep. (Pun definitely intended.)

Linear independence Linear transformations

Matrix transformation example 3: dilation and contraction

Let r be a scalar and consider 2

 the matrix transformation from R
r 0
to R2 with matrix A = .
0 r

x2 Av1 x2

v1 Av3
v3
A
v4 x1 x1
v2 Av4
Av2

If r ≥ 1, we “dilate” the vectors (that is, we stretch them).

Linear independence Linear transformations

Matrix transformation example 3: dilation and contraction

Let r be a scalar and consider 2

 the matrix transformation from R
r 0
to R2 with matrix A = .
0 r

x2 x2

v1
v3 Av1 Av3
A
v4 x1 Av4 x1
v2 Av2

If r ≤ 1, we “contract” the vectors (that is, we shrink them).

Linear independence Linear transformations

Matrix transformations are extremely special because they

are linear

That is, if A is a m × n matrix, then A(u + v) = Au + Av and

A(cv) = c(Av) for all vectors u, v in Rn and all scalars c ∈ R.

This property is so useful, we abstract it out into its own

definition: a transformation T is linear if:
(i) T (u + v) = T (u) + T (v) for all u, v in the domain of T
(ii) T (cu) = cT (u) for all scalars c and all u in the domain of T

Of course, a matrix transformation is always a linear

transformation.
Linear independence Linear transformations

Extremely useful properties of linear transformations

If T is linear, then T (0) = 0.

If T is linear, then T (cu + dv) = cT (u) + dT (v) for all vectors

u, v in the domain of T and all scalars c, d.

More generally, if v1 , v2 , . . . , vp are in the domain of T and

c1 , c2 , . . . , cp are scalars, then

T (c1 v1 + c2 v2 + · · · cp vp ) = c1 T (v1 ) + c2 T (v2 ) + · · · + cp T (vp ).

In words: a linear transformation takes linear combinations to

linear combinations.
Linear independence Linear transformations

iClicker 5

Suppose Tis alinear transformation from 2 1

  R to R = R that
1 42
takes u = to 5 and v = to -3. Which of the
32 −4329
following vectors in R2 maps to 2?
(a) 2u + 3v
(b) 0 + 2
(c) 4u − 6v
(d) 2u − 3v
(e) 4u + 6v
Linear independence Linear transformations

Remarks

We don’t actually need to know what u and v are to answer this

question.

0 + 2 doesn’t make sense: 0 is in R2 but 2 is in R. We can’t add

them!
Linear independence Linear transformations

Beware!

Not every transformation from Rn to Rm is linear!

(Although, in linear algebra class, the main object of study is linear

transformations.)
Linear independence Linear transformations

iClicker 6

Only one of the following is always linear. Which is it?

(a) Any transformation that sends 0 to 0
(b) A transformation from R2 to R2 that sends (x1 , 0) to (−10, 0)
for all x1 and (0, x2 ) to (0, 9) for all x2
(c) A transformation from R10 to R = R1 that sends every vector
to (3).
(d) A transformation from R10 to R = R2 that sends every vector
to 0 = (0, 0).
(e) The transformation from R to R that sends x to 2x − 1.