Lecture I

A linear Equation

+ azsezt tansen b- ①
ape , .
. . -_

Collection of points ,
& , .az .
-
.
,dn) c- Rn satisfying
① is called a

txampe ( Linnean) solution .

1. Use ,
-
5xz+2= se
,

te rearranged
3. se
,
-
5xz= -2

2. Nz=2( V6 -
xD + Nz

frae arranged

2sei-sez-sez-2.GE/ampIe- ( nonlinear)
1. 4N ,
-
Gnz=N , R2

and

2. Nz=2F -7
System of Linear equation

A finite set of linear equations

in
finitely many indeterminate .

Exampte :

2x -142=5 ①
-13g
-

set
y
-12=2 -

A solution
of of Linear
a
system
Equations .

A list ( Si , Sa ,
- - -

) of
.sn numbers

that makes each equation in

the values
the system true when

Si , sz ,
-
-
-

, Sn are
substituted for
Nc , Nz ,
- - -
, Nn respectively .

Examples ( Two equations two

in

variables ) .

Each equation determines a line in

2-
space .
Bo N, + xz=
10

-
N , + Nz =
0

n
Nz

Ni

Oneuniquesolution
ooo se , -
2xz= -3

2x ,
-
4nz= 8

I
0 solution
-
If Ai (i -1,2
-

) is the set of
solutions ith
of the equation ,

then the set of solutions of

the Almaz
system
is .

We assume for the time

being
that
the
coefficients of the eggs are

from IR .

A
system of
'

'm
equations in

'
'
n unknowns se
, , , Nz , - - -

Then can

be written as ,

91mi + aizrkz + tain Hn = b

,
-
- -

921 Ni t 92222 + - - -
1- Azn Nn =
bz
④ -
-
- - - - - -
-
-
- -
-
- -
- -
- - -

am , Ni + amzsez -1 - - -
+ 9mnnn=bn

where
aij c- IR ( lsism , lsjon)
called and
coefficients bi.bz ,
-
-

→ bm
ER ,

called constant team of the

system .

system (A)
This can be described

by the
following matrix notation ,

AX=B

where

( )
,= ÷""÷÷Éna+ai
A-
Mm,fnR)
an
area .
-
-

aan c-
-
- -

-
- - -
- -
- - -

am l
amz -
- -
9mn

)
b-
c- Mmxn ( IR) .

matrix
of constants
?⃝
and
se ,

Rn

-
matrix
of unknowns
.

/ B) C-
Mm×cn+ , (R) obtained
the column
by attaching
with A is called the
B

Augmented
matrix of the

system .

The solution set

of
A✗=B is a subset
of
Mna ( IR) element
every
as
of Mn× , (IR) defines an element

of IRN and vice versa .

④
Ni + net Nz =3

6
Zzz -13kg
=
se +
,

Nz + 223 =
I

B
Ni + Nz -1
Nz =3
Ni t 222 +
3az=6

se
, + sez -12mg =
4
 C

N + Katse
,
> =3
Ni +
2h2 -1 3xz=6

sea + Zsez =3

A -

no solution
(subtracting the 2nd
eq
from the
sum
of 1st & 3rd we get 0--2
)
B- one
solution
Ccnamme 's rule
) .

C -

infinitely many
sobs .

((7,3-2×-7) for any XEIR

is a
Sohr )

The collection (snsz )

of all -
- -
Sn

C- IRN

satisfying .
 n

2-
aixi-b.in
is called
hyperplane
.

Now if Ai 's ,
i= 1,2 .
-
-
m

is the soba set for each of

the above
eggs
in the
system
(5) then

MA ;
m

is the sole of the

i=i

system .

Ai 's
of these a
Each are

plane ( geometrically) .

So the solution of the system

is the intersection of planes .

Unique have
The planes
.

solution •
 point of intersection
a
unique
.

*e→qGzBBM

V0 Solution : Some
of the

equations are
contradictory ,

solution exists .

so no
 The planes intersect
Infinite Solutions ;

in a line .

13=0 Then the

:
system is

homogenous
13=10 Then the
system
: is

homogenous
non -

A
homogenous system always
has solution (
namely
a the

zero colomn ) .
 Basics of Matrices .

Matrices are
rectangular array
numbers
of .

Here the numbers that we will

consider are in IR on ¢ .

Ex :

-_( )
ay ,
9,2 -
- -
9in
A
°
" °" -
- -
"^

-
- -
9mn
am , amz

where aij
C- IR
( f)on

• Size
of a matrix :(No ofaons of
✗ No

columns
)
= man

( square [
' °
]
•

)
m=n
matrix
re

•
all , azz ,
- -
-

9mm : diagonal entries

of a
square matrix .

•
m=1 ( single now
) f. ] o

Row vector

•
n=1 ( single column
)
Column vector .
I :]

• Two matrices ,
A & B are

equal ,
ie ,

A=B where

A- =
Cai ;) ,
B=( bij)
if
if and
only
have size
They
• same

Hi ,j
•
aij =
bij , .

matrices
• Sum
of two .
 A=(aij) ,
B- ( bij )
of same size man is another

matrix
C =
( cij) where

cij-aij-bij.fi ,j .

Two matrices of different size

cannot be added .

scalanmultiph.ca/ionletcc-R(onc)
•

,
if AE Mm×n ( IR )

the CAE Mm×n ( IR) where

Yau
cA= (caij) can - - -
cain
=

' '
'
-

- -

- -
- -
-
- -
-
-

Cann
Cam ,
camzoui
 B C- Mm×n(IR)
Paopertiesofmataices : A ,

A
•• A -113=13-1

•• A + (
Btc ) =
☒ + B) + C

is
At Oman =
A ,
Oman a zero

••
matrix

• At C- A) =
0min

••c( At B) = CA + CB

DA
• Cc + d) A = CA +

• Ccd A) = (d) A
o• 1 A =
A

Mataixmultiplieah.in/t--(aij)mxn,B--(bij)n.xp
Then AB is defined
and
iff r
 matrix
1. AB=Ccij) is an
mxp
n

2. =
I aiebej
cij 1=1

by aiabzj + - -
-

= +
ai
.

ainbnj
+

Examine

[ ÷] I:÷÷:]
all 912

21
0h22 0h23

931 az ,

3×033×2
=

1 ::
231
:
:]
C-
32

3×2

Observation:

[ §;]
Consider
column vector .
that is ,

§ ] |?
all 912 • • • AIM

" a a"
" • • •

: :
•
③

9mi Amn
amz •
• r
bn

b)
1- ainbn

(
911 bit aizbzt .
. .

=
+ "

:b
+ b- + • •
•
"
a ' a-

:
amzbzt tamnbn
b,
. . .

+
am ,

Now we can define matrix

multiplication of two matrices

A and B ,

AB =
A[b , bz . . .

bp)
=
[Ab ,
Aba . . .

Abp] where
 B =
[b , ba -
- -

bp ] ,

( ;]
bi bi
=
Isis p
bzi ,

bni

Propfmatms
.

d # B)
1) @ A) B = A @ B) =

2) ACB c) =
CAB)c
3) + B) c = Ac + BC

4) CA + CB
CCA B)
=
+

µ )
00 - o

Inn
- -

••
=
1 0 - - -

: :
① - -
-
- l

Identity matrix .
12¥ :

•
Order of multiplication is
important .

In
general ,

AB # BA .

Eiample :
A-
[? :] .B=f:?)
AB=
[9-7]=113*[9-1]
A=OoaB=O
• AB=0 does not
imply
A-
-

f; :] .
13=1-1 :]
then AB=
[88]
• AC = BC does not
imply
A=B on C=O

A=f ? :] .

B=f :?)
c-
[
'

a :]
[ ? %]
then
Ac=Bc=

• One has to be careful

laws
using algebraic of matrices .

For example ,

( atb)? a2+2ab -15

This is not true for matrices

ie
(1++13)<=1 A2 + ZAB -1132 .
 matrix
•
Transpose of a .

A=(aij)m×n
Then
AT -_
Caji )n×m

[ First row
of At is first
column
of A
]

[
A
]
all 921
=
then
912 azz

At
[ ]
a
"
912
=

921 922

•
Properties
1) ( )T=A
At
a) (
AB )T= BTA '

A -113)T= A -413T
2) (
'

3) CAT
@ A) =
• An nxn matrix A is

invertible if there exists a

,

matrix B such that

AB=BA= -1-3×3

Matrices
Symmetric .

A
sq
matrix
,
A ,
is
symmetric
if fA-Af
-

Skew Matrices
symmetric
A = -
AT A square
,

matrix .

Diagonal Matrices

Square matrices which have

non -
zero entries
only along the
diagonal .

I
I
"

(C) ann

Upper triangular / Lower

triangular
Square matrices which have
entries
only
non -

zero on and

above the
(below)
diagonal .

①
i .g g#TMD
"
"" "
"
'

""

0 922 -
- -

92m

0 0 933 -
- -

93m
'

10
 ①
I :::÷÷
921

i
¢-1M

am ,
-
-
-
-
-
-
-
.
/ )

TnaceofaMata
A : matrix
square

for (A) = au tazz -

-
.
+ ann

= Sum of diagonal entries .

•
ta(A+B)=tn( A) + ta(B)

•
ta (A) = ta(AT)

•
ta ( AB) = ta( BA)

matrices
•
Idempotent are
square
matrices which satisfy .
AZA .