Recitation 2.

based on material by Carles Domingo-Enrich

Center for Data Science

September 15, 2023

Concept Review: Guassian Elimination
Elimentary Row Operations
I Swapping two rows
I Multiplying a row by a nonzero number
I Adding a multiple of one row to another row
Gaussian Elimination - From Lecture 2
2 3 2 3
1 1 0 1 1
4
A= 2 0 1 5
1 2R n⇥m and y = 15 2 Rn
4
1 5 2 0 4
Cinean AxY if x =
(π) a ,
*2 X3 X4 ER
leading coefficieat
, ,

!x, ;

{
⼀
⼀
-
Xzt X4 = R
1 3

恐
t x3 X4 1
) CRz
- =
r

ttpg
)
esxrtzx-
-

X4 = 1 3
R
X2 t

「
x ,
-

3-
( R2) 2 CRD
X 4 =- 1
-

→ + zX2+ X 3

↑
5 ) +4X 2 +2 X 3+
X
4
= CR 3 ) + [R,

X, x
4= 1 3

δ
-

X 2 f R
o t 2x =+ X 3
) 3X 4 yR -
=

X4 = l
R
oto + 0 + 7x 4 7 =
Gaussian Elimination - From Lecture 2
x, 1 号

「
-

{
xz + 1 =
=
* -

> 1
号
—

1 xn +xr = x2 = 1 -

2x 4
= XΨ = 1

{(f ) H
∞| rtk =
s =
) tx }

{({ ) +v
| vaspanl
☆ a
Concept Review: Subspaces
Definition (Subspace)
A non-empty subset S of a vector space V is called a subspace if it
is closed under addition and scalar multiplication:
1. Closure under Addition: x + y 2 S for all x, y 2 S.
2. Closure under Scalar Multiplication: ↵x 2 S, for all x 2 S and
↵ 2 R.
I A subspace is also a vector space!
I A subspace always contains the zero vector.
Concept Review: Span & Linear Dependence

- Let x1 , ..., xk 2 V .
def
Span(x1 , ..., xk ) = {↵1 x1 + ... + ↵k xk |↵1 , ..., ↵k 2 R}.
Concept Review: Span & Linear Dependence

- Let x1 , ..., xk 2 V .
def
Span(x1 , ..., xk ) = {↵1 x1 + ... + ↵k xk |↵1 , ..., ↵k 2 R}.

- The family (x1 , ..., xk ) are linearly dependent if there exists

↵1 , ..., ↵k 2 R that are not all zero s.t.

↵ 1 x1 + · · · + ↵ k xk = 0
⼀
Concept Review: Span & Linear Dependence

- Let x1 , ..., xk 2 V .
def
Span(x1 , ..., xk ) = {↵1 x1 + ... + ↵k xk |↵1 , ..., ↵k 2 R}.

- The family (x1 , ..., xk ) are linearly dependent if there exists

↵1 , ..., ↵k 2 R that are not all zero s.t.

↵ 1 x1 + · · · + ↵ k xk = 0

Ex. Consider the two vectors v = (1, 1) and w = ( 1, 2), is

S = {↵v + w |↵,
…
0} a subspace of R2 ? F ) (i +r (点 )
(i ) → - 5) (i ) =
(f ) a+ 2
B ≥O
Concept Review: Basis & Dimension

- A family (x1 , . . . , xn ) of vectors of V is a basis of V if

1. x1 , . . . , xn are linearly independent,
2. Span(x1 , . . . , xn ) = V .
Concept Review: Basis & Dimension

- A family (x1 , . . . , xn ) of vectors of V is a basis of V if

1. x1 , . . . , xn are linearly independent, ∠
⼀

2. Span(x1 , . . . , xn ) = V . ←

- If V admits a basis (v1 , . . . , vn ), then every basis of V also has n

vectors. We say that V has dimension n and write

dim(V ) = n.
⼀
Concept Review: Basis & Dimension

- A family (x1 , . . . , xn ) of vectors of V is a basis of V if

1. x1 , . . . , xn are linearly independent,
2. Span(x1 , . . . , xn ) = V .
- If V admits a basis (v1 , . . . , vn ), then every basis of V also has n
vectors. We say that V has dimension n and write

dim(V ) = n.

0
Ex. Let v1 , v2 , v3 , v4 2 R3 . Let C1 = {v1 , v2 } and C2 = {v3 , v4 }.
⼀ ⼀

If C1 and C2 are both linearly independent, what are the possible

values for dim(Span(v1 , v2 , v3 , v4 ))? (No formal proof necessary)
2, 3 . 4
 1 Linear Transforiuath

!
.
.

2 .
Matrix
Repesentcathm ,

3 ornel (
.

Imef ,

Recitation 2.2

based on material by Carles Domingo-Enrich

Center for Data Science

September 15, 2023

Concept Review: Linear Tranformations
⼀

←
Definition
A function L : Rm ! Rn is linear if 0 6
⼀号 I For all v 2 Rm and all ↵ 2 R we have L(↵v ) = ↵L(v ).
⼀
⼀

→ I For all v , w 2 R
m we have L(v + w ) = L(v ) + L(w ).
←
⼀

-p
Remark: < coto ) ∞
∞ I L(0) = L(0 + 0) = 2L(0) =) L(0) = 0
D
∞
I L(⌃ki=1 ↵i vi ) = ⌃ki=1 ↵i L(vi ) ↑
2L ( )
0 =
< co 了
…

G spant
≈ =≥ C 0o) 20 ,
d
question 1: Linear Transformations fcx) fa α ) =α
{{ fcaey
x

is fcxltfoy)
) =

Which of the following functions are linear? If the function

∞

µ
,

linear, what is the kernel?

∞

v
1. f1 : R2 ! R2 such that f1 (a, b) = (2a, a + b)

!
{ 2. f2 : R2 ! R3 such that f2 (a, b) = (a + b, 2a + 2b, 0)
⼀…

→ 3. f2 : R2 ! R3 such that f3 (a, b) = (2a, a + b, 1) 身

p
4. f2 : R2 ! R such that f4 (a, b) = a2 + b 2
→ 5. f2 : R ! R such that f5 (x) = 5x + 3 x
∞
Kercfis : fi 0
lab)
。
Ψ
=>
{皆
a (er (f ) =
(8 }
)
{ atb
2a
= 0 ,

a =0
= 0 .
.

frca b ) =0

= {
ercfbp
,
,

G+ b = 0 s a= b
- s )

=>
{ zaezb = 0
{
O = O
=
g
span (
Concept Review: Geometry of Linear Tranformations
Homothety

t
.
H : Rn ! Rn
x 7! x
.
i. =

]
Rotation

卡
R ✓ : R2 ! R2
x 7! x rotated by angle✓

]
cos0
Sino

!
-

Ño -

no coso

→
question 2: Linear Transformations
Find a 2x2 matrix A (not identity matrix) such that A4 = I
「
-
p

1 icencescinny
←

←
d
x∅
←

只
"

乡∅

(∞
ei _
(。) ⼀

[ ∵)
0
巩 a
4
= 1
,
.

.
Concept Review: Matrix Product
Definition
Let L 2 Rnxm and M 2 Rmxk . We define the matrix product LM as
the nxk matrix of the linear transformation L M. Whose
coefficients are given by the formula:

(LM)i,j = ⌃m
p=1 Li,p Mp,j for all 1  i  n, i  j  k
K

n
t
:
t
…

…
间
question 3: Matrix Product
Calculate the product AB and verify the following properties for
matrix multiplication:
I Non-commutativity: AB 6= BA
…
I Associativity: A(BC ) = (AB)C

( (.
. ) 0
Concept Review: Matrix Inverse
Definition
A square matrix M 2 Rnxn is invertible if there exists M
⼀
1 2 Rnxn
…

such that
MM 1 = M 1 M = Idn

o is unique
such that M 1

M N . = N M
- = I
f
N = M
Question 5: Matrix Inverse
Find the inverse of the matrix below and verify. What do you
observe? θ 10 1I 江1
" p #
1 3
A= p2 2
3 1

l
2 2

1∴
p

⼩
AB⼆ I
⼼Δ

l ↑
…
ε …
,

…
*

:
:! ( 1 0)

皇誉 )
,

主
β“
(
1

* Cheer ac Hurme: AB
= BA I
=
 )
.
Concept Review: Kernels and Images
m coam →
Kernel
The Kernel Ker (L) (or nullspace) of a linear transformation !
L : Rm ! Rn is defined as the set of all vectors v 2 Rm such that
L(v ) = 0. i.e. Inge .

def
Ker (L) = {v 2 Rm |L(v ) = 0}

Image
The image Im(L) (or column space) of a linear transformation
L : Rm ! Rn is defined as the set of all vectors u 2 Rn such that
there exists c 2 R m such that V (v ) = u. Im(L) is also the span of
⼀ θ

the columns of the matrix representation of L.

⼀
“

K
m

( C ( el Lie ) .
Lcem ) )
= L(x txmn ]
xeit at txmen Lx ← …

x= ,
. …

0 β Δ
Question 6: Kernels and Images xi) tx
= 2L (e) -
←
…
txmL(
)
Let T : M22 7!θ
R2 be defined by =
spa ( L (G ) Ez) ,
…

CC3),
 
a b a b
T =
c d c td

Then T is a linear transformation. Find a basis for Ker (T ) and ⼀

Im(T )
⼀

eamiitaim
Givie
-

…a
-
,

r ,
l
x

Span
↑ ↑
笻] ken() : bass ;

.( ' 0 d )
( ) } NOT
品 UNIQUE ANS

“
[ IuCr ] basis
, { ( ' ) 10
,
:
,

}ALSO NOTUNIQUEANS
6
Concept Review: Solving Linear Systems