You are on page 1of 10

Linear combinations

span basis linear


vectors independence

й
аш ї 25

ЇЇ і
j are basis

Span depends on what basis vectors we chose

Setof vectors is linearly independent it neither of vectors


can be expressed by linear combinations of other
vectors
The basis of a vector space is a set of
linearly independent vectors that span the full space

SPAN

span vi vn
Е Civile GEF

The
span is a set of all vectors that could be
reached with a linear combination of given vectors

Maybe a
point line 30space _or whole space

D в є span я vn itt Ax в has a solution


1
2 If span я vn Une say that he vn spans V
U vectorspace
3 A vectorspace U is finitedimensional if I w Visit
Span14 VIV
2
LINEAR INDEPENDENCE

List м ua in V is LI if the only choice


2 ant є F that makes 2,0 02,2 0 is
9 22 24 0

2 V0 0241 0 2 2 0
Set of vectors is linearly dependent if some of the
vectors could be expressed by linear combinations at
other vectors in the set Because some vector is in the
span of the others

Ae REFAA
Ах 0 х the only solution

Solution exists no contradictions по 10 01C с о

Unique solution REFhaspivot ineverycolumn As


3
Basis
A list of vectors Beer vn is called a basis for V it
win V admitsunique representation
any vector
adult приз Esri
for some diaper
піду
has a unique solution
v10 xpup re
Дтвday
2 calledcoordinates of u withrespectto ptosis
xp are

A list of vectors is a basis for V if it is LI andspans U

Linear dependence lemma


H lui vn isLDindandv.to
the
my 7jeL2
such that

а
Щє span V1 Vj 1
6 If theith form is removed from v1 rml the span of
the remaining list equals to span 4 4
Theorem Every span
riglistcontains a basis

or LDspanning list can be reducedto a basis

Algorithm
1 Remove Yeo from13 Blu it exists
2 If Ц is in span 4 remove v2 from р
j It у is And
in spanhe 4 remove
у
so on
See Sample test

Theorem Every LI listatvectors can be extended to a basis

Algorithm
Fix spanning list of vectors w vn
1 It w is in spanhe vn letBrlui rm
ifnot letpick um w
5 Ifwjisinspanlpj.nl etpj pj
ifmt
setpj lui un wp
Aftereach step pj is still LI accordingto LDlemma
Preposition for matrices
F field n positiveinteger

µ um list of vectors in F A.lv um nxm matrix with


columns 4 Un

list 4 um is LI itt A REFA haspivot in every column


list vi um spans V itt AeREFN haspivot inevery row
System u _um is a basis for F itt Aehas pivot in every
row andcolumn
Importantfor many theorems

Dimension
Dimension of vectorspace V is the length of any basis
of V Denoted dimN

И subspace of V dim 4 dim u

length of any LI list in V is atmost dim U

length of any spanning list is at least equalto dim10


Invertibili.ly
A is invertible it 7 Bst ABI In identitymatrix

Він її
Theorem

А is invertible it for
any
в Axl has a unique
solution

Corollary
A is invertible iff columns of A farm a basis for F

Find A
1 Farm
Alin augmented matrix
2 Through row operations make A reducednow echelonform
3 А shalldbecome In In shouldbecome searchedAt
А In

др Invertible
iff Are Aven ha e
unique solutions algorithm
solves everythingsimultaneously
Fundamental subspaces

1 Kernel herA ref Avo ef Allsolutions of Ан 0


2 Range ran A Anker IF F span ofcolumns ofA
Column space

3 Rank А dim ran A

AT Atranspose of A

A AT
rows become columns
columns became rows
or flip arounddiagonal

ran А herА ran At her Б fundamentalsubspaces of A

To compute fundamentalsubspaces and rank we need REF

let A єMm.nl F Ae echelonform

I Columns
withpivots basis of ran А ran А
deal
A dim herAll number of free variables dreduced row
echelon
Rank theorem

rank A ran AT
dimlherdtd.im ran1All
dimlherlAlltrankCA
n

1 41 _xp
д у у є F
У у у
spin Ran соĸира
1 dim span y xp rank
ХУ
2
Ує span x xp iff rank X rank 49 lxrx.gg
в є F is a solution of Ax biff rankА rankCAD

yp iff rank X rank 9


3 Span x span y rankXD
xp

Proposition

A is invertible rank An columns


off are LI
А І
II can be obtained from A through row operations
ranklldimlherlA rank.to n
Р
no tree variables

Changing basis

You might also like