You are on page 1of 29

I .

1 & 1.2
1.1 .
Linear Equations

at t 22=1 ( each variable is


power of 1 )
y
( each variable is
multiplied by a constant)

General form :

A t b
,
N ,
t 92 Uz t .
. . -

an Nn =

D All variables are raised power I .

b are constants
2) Az
i
a
-
-
-

, ,



at t 2 = I not a linear equation

a u
, ta
z Nz -
t
-
-
-

t an un a there b -
- o
)
,

is a linear homogenous equation



5 u
they &82 -
I a o

linear but not homogenous

It
II} } system of linear equations

④ a
-
.

I , y
-
-

I
There is
exactly I solution

n
ty
-
I
23kt 2
2g
=

& substitution methods


get using elimination
we o - o

Guess a solution

solution
ez
is
a- -

ta , y a

'
" '
' '

NZ I y
z O
,

' ' "

A- 0
1
,
y =

we have a
many
solutions


n*yzt
Ly
2kt z 3

& substitution methods


get using elimination
we 021

wrong

If
aety = I then 2xt2y must be 2
,

No solutions .
Trick
-

Solve

htt2gal
-

2
-

2y
=
a
-

don't need write "


"
to
we my , a

row I → I 2 I

row 2 → I -2 2

add row I to row 2

I 2 I

2 O 3

from 2nd row

2h =3
toy

N -
-

3-
2
Coefficientmatrix

⑨ 4
2K t 3 z a

y
-

4 t 2
y z -
a t

3 2 =
1
N t
y t

i. I
ii.


matrix

moon
,

efficient
C-
@
:
matrix
r

row
} ( Rs)

Augmented

write the augmented matrix

N
,
t 2C
z
t
Nz z I
4
n
,
t 223 a

5 rewrite
N t Nz =
uz
-

N , t
Nz t
uz
z 5

:lit

.
.
:,
write the system .

need 3 war

n 2
, y ,

3 x t 4 t 52 I
y
-
=

n t 3
y
2

22 4 2
Zy
t t 2 =

write the system

I 2 2 I

-
↳ because of
the answer

we
got
var this triangle of zeroes

W n , y 2
, ,

W t 2x t
Ly t 2 a 5 -


3 n t
y
-
2 z a I -

-

y t 2 z
a
-
I -

4 z z 4 -40
↳ 2=1

Sub t
2 I
y
-
-
=

-
3
y

Sub z in ②
,
y
3 2 I
3 t =
-

2=0

Sub a
, y iz
in ①
W to t 6 t I a
5

W = -
2

" ' '

Back substitution

Solve
3 1
2g
at t a a

+72
Sy
I
3 a t =

Hit
:
2 n t
by t 2 2 22

any . mat

I 2

I
3

we need


:L
:L
÷
"
"

we use Elementary row operations


.
Nt
2g t z =
A-10
3kt
Sy t 72 = I -

I 3) a ① ②
-

¥
Hit

first
step
Put zero here

An

multiply
row I with -3 and add it to row 2 Elementary I row
operation

C- 3) R *Rz → Rz
,

{3*1}×1 ! )
Sttgxz
7+2733×1 ) It x

to i: t 'd
a I 2 3

§
step 2

Put zero here


add - 2 times now 1 to rows

( •
2) R ,
t Rz → Rz

I "
I 2 3

±: .

§
step 3

Put zero here

Rz2. t Rz →
Rg

-1¥
I :
I 2 3

rewrite the
equations '

32 I
a t
Ly t -
-

2 2
y
-

z
-
z
-

-8 z
-
-
4 T Back Substitution

22 -
Iz

y 2
( I) 2 ⇒
y -
I → yal
- -
- -
- a

CD I I)
n t 2 t 3 a
I

§
-

2L 2
Twomorerowoperationscl
::,
we want 1 here

with
& we can multiply an entire now a non -

zero number ,

So here multiply row I with Iz

:÷L

we
*
⇒:
want a non -
zero number here

* we can switch 2 rows here switch row 1 and row 3


any
.

R
, et Rs

71
:)
3 4

:

solve
t 2
y
set a 1

42 3
22 t
2g t a

3kt 3 t 62 =3
y

:L
:L
I I I

I
2

:
we want

f.II
:
. .

. : :

( -
2) R , t Rz →
Rz

:B
c:
( 3) -
R
,
t
Rz → B

l :"
t t 2 2 at
a y
I
y =

0 = 0

: .
N =
23

Can put any value for z -

So 2 is a FREE VARIABLE

Let t
za

i. a =
-2 t

convention

-
{ L -
it
,
I
,
t ) :

tianya.az#lnmber}
( set of solutions

solve

11¥
:
C- 3) R , tRz → Rz

i.it!
I l 2 3

I.

( -
2) R t
Rg → R3
,

:
°
%
:* , O I

t *22 3
se
y
.
-

-4
ya
o -
I

NO solutions .

Inconsistent .

[ all zero
log
( non - zero

2yt3z
I
at z -

72 5
3kt
by t z

c ::L :L
we need improve
the method .

RowEchelonf.org
*
Leading ones and leading zeros .

! ⑥

Leading zeros

000
element
It is a
1st
eaodifngruownets
.

of rows non zero


-

,
.

÷:
deciding I 2 o o o

zero ← O O O o O

:÷÷
:!;
¥
of manson .

row
one
2
row

.su?...c:.::::
: ④ he
of row )

No zeroes
* leading in rows
A matrix is in Row Echelon form if
The first number of is
① non
-
zero each row I
[ Does not apply to zero -
rows ]

If at the bottom the


② there are zero
-

rows
, they are of

matrix .

③ Number of leading zeros of a row is


larger than
number of leading zeros of row above it .

[ Does not apply to zero rows ]

4
0J
2

O O I Echelon form

8)
O O O

" '

{ o on not echelon LR , has

Kg
more
leading
O I O Zeros than

,
§§! not echelon ( not increasing zeros )
echelon L not
[ )
not )
to
q
increasing zero

↳! ) not echelon I Rz starts with non -

zero)

I o 3 4

( )
Echelon

!
Gaussiandiminationi
the matrix
) Write augmented
echelon form
2) Use row
operations to
get an

3) Assign values to free variables ( if


any )
4) Back substitute

④ Some

3ns I
a
,
+ znz -
the -
-

*723
6
Z
Nz t
2123 Ky
- -
=
,

-2n ,
-

422 t -

My = I

leading one
C
pivot)

µ
I 2 - 3 I

II.
: I 'd

R R 2K R
,
t
Rz →
z , ,
t
z

Rz

o.in?i'oli7
pivot

(

O '
( now is in an echelon form)
o

o o I 3
pivot

choosingfneevan.at
Look for all
pivot
There are in columns I 2 and 3
pivots , ,

: . we
say n
, , nz ,
n
,
are
leading variables
Remaining variables are free variables
Ny is free

323
t
a
, Inez -

try =/

it 7
Kz Zz
-

=3
Mz t
My
t
my
-

Backsubs
-

3 -
f
Ng
z

4 #t
Kz
=

2
n
,
=
-

ft t

Set of solutions

{ ( 2- Gt
,
4ft ,
3 -
t
,
t ) i t ER }

Solve
2x
a
, ten , t
,
t
my 23

2
U
, taz t 3mg z

2N ,
t
Zzz t
523 t
2h42 3


pivot

1%
:

µ
- R , tRz → Rz
-

2R ,
t Rs → Rz
,

pivot

III
'

I o -3
) not echelon

° o
-
Rz t
Rz → Rs

( I 2 I 3

( 0 o I -
I / -

I
] EchelonForm
O O O I - 2

No pivots in column 2

o :
Nz free

t 3
u t
Nz My
-
,

+2kg
ng
-
my e
-
I
-
2
Uy
=

Let t
uz
=

3
-

Ng
-

t 6
n
,
t -
t L -

2) a 3

U a 11 -
t
,

Solutions

{ ( It - t
,
t
,
-3
,
- 2) it ER }

:* : :L

- I
these columns
pivots
in
no

Nz are free variables


, ay

Let nz a S

ne a
t
,

Solutions

{ I 4 -
t
,
t
,
- s
-
t
,
s ,
3) i s
,
t EIR}
Reduced Row Echelon Form ( RR E )
-

A matrix is in RRE if

② It is in an echelon form

⑦ It a column contains a pivot ( leading one ) ,


everything else
in that column must be zero .

)
I I 2 3

\ to
ft
Oo
Echelon but not RRE

-1
I 3

I
o o

me
:

:/!
I o o o -
I

I
name
Gauss -
Jordan reduction method
-

D Write augmented matrix

2) Get the MRE form .

Rest is same

Solve
N
,
t 3 Nz t
uz t my
= 3

2n 2K t
224 L
-

tmz
a
,

3N t
Mz
t
2mg my = - I
-

:L
:L
l 3 I I 3

I -

in
:
ZR tRzsRz Rs
-

, -3K t Rs →
, ,

2)
I 3 I I 3

| o -8 -1 o )
O -
8 -

I -4 -
to

I Rz
-

need zero here

go
to
:*:
:÷ .
'
8 Rz t Rz → Rs ,
-

3RztR ,
→ R
,

:÷:t±t
" o I I ¥
I
:
If Rs
-

:¥÷t¥t
i:
-
t
Rz
R ,
→ R ,

.it:1
I It %
I:
o o

: :
RRE

N
} %
¥
-
t
,

tf ¥
-

Nz Nz
z

3
14 =

a
,
is free

Let as a t
¥ ft
-

Uz
-
=

n
42 It
-
=
,

{( I -

It ,
-

I -

It ,
t
,
3 ) :
ter }


'
at
2y a

2n 3
t
try
a

If 4=4
,
inconsistent 2nt4y - 3 not 2
?

:*
*, A

For what does the f.


system have D
many solutions ?

:p
:::)L
to ( close ) echelon
Try get or an form
2h R2 Rz R

tRz R3
-
t →
, , ,

I
I 2 I o

I !:
:L
I 2 l O

:p
For a
solutions we need a free variable
many , .

If
p 2=0 then free We have solutions
-

,
a
,
is . a
many .

Ans .

p =L

Ii

① find all a
,
b such that
system have a
many
solutions
"

inconsistent
" "
" " "
is

solution
" " "
" "

has

"
a
unique

::e* ,
too
!:* !
① A -
520
,
b - 4=0

② a -5=0 ,
b -
4 40

③ a -5
to and b can take any value

:}
at
Ly
°
t 2 2

an

2-273
If:I
zeros

homogeneous
system

Find B so that system is consistent

this solution
N 20
,
yeo ,
2=0 is a
?
Yes will make it inconsistent
, no
B
Homogeneous systems always is consistent .

There has to be a solution .

Yess
it is solution ( trivial solution ]
No such B

⑨ arty z I atb =
2

4 2b =3
n
2y a
- -
- -

X
both have same coefficients

Trick

list
:

c.
H
's
.
ran .
- r
.

tr
-

I 1¥
III.
! y
I atb 2
arty
- -
-

I be Is
- -

y
=

Definition
If B is obtained by performing row
operations
on A
,
we say A and B are Row Equivalent

You might also like