C2 Applied Linear Algebra
C2 Applied Linear Algebra
C2 Applied Linear Algebra
AM
PH
Duong T. PHAM
T.
g.
APPLIED LINEAR ALGEBRA
on
Du
1 Motivation
AM
2 Determinant
PH
3 Basic properties
T.
4 Laplace’s formula g.
on
5 Inverse matrix
Du
6 Cramer’s Rule
AM
If a 6= 0, Equ.(1) has a UNIQUE solution x = 0
PH
If a = 0, Equ.(1) has INFINITELY MANY solutions (all x ∈ R are
T.
solutions)
g.
⇒ number a decides the solvability of Equ.1
on
Du
Ex.2:
ax + by = 0 ax + by = 0
AM
(
ax + by = 0 cx + dy = 0 cx + dy = 0
⇐⇒ ⇐⇒
PH
cx + dy = 0 acx + bcy = 0
acx + bcy = 0
acx + ady = 0
(ad−bc)y = 0
T.
If ad−bc 6= 0, has a unique solution x = 0, y = 0
g.
on
If ad−bc = 0, has infinitely many solutions
Du
a b
ad−bc is called determinant of
c d
Definition
AM
A = a11 ; det A := a11
PH
a a12
A = 11 ; det A := a11 a22 − a12 a21
a21 a22
T.
a11 a12 a13
A = a21 a22 a23 ;
a31 a32 a33
g.
on
det A := a11 a22 a33 − a11 a23 a32 − a12 a21 a33
Du
a11 a12 ... a1n
AM
a21 a22 ... a2n
Given a square matrix A=
. . . . . .
... . . .
PH
an1 an2 ... ann
Determinant of A,
T.
a11 a12 ...
g. a1n
a a ... a2n X
det A = 21 22
on
:= s(f )a1f (1) a2f (2) . . . anf (n)
. . . . . . ... . . .
Du
AM
a11 = ?
X 1
s(f )a1f (1) = s a = a11
1 11
PH
f ∈S1
a11 a12 P
a21 a22 = ? f ∈S2 s(f )a1f (1) a2f (2)
T.
1 2 1 2
=s g. a a +s a a
1 2 11 22 2 1 12 21
on
= a11 a22 − a12 a21
Du
a11 a12 a13
X
a21 a22 a23 = s(f )a1f (1) a2f (2) a3f (3)
AM
a31 a32 a33 f ∈S3
PH
1 2 3 1 2 3
=s a a a +s a a a
1 2 3 11 22 33 1 3 2 11 23 32
T.
1 2 3 1 2 3
+s a12 a21 a33 + s a a a
2 g.
1 3 2 3 1 12 23 31
on
1 2 3 1 2 3
+s a13 a21 a32 + s a a a
2 1 13 22 31
Du
3 1 2 3
=
a11 a22 a33 − a11 a23 a32 − a12 a21 a33 + a12 a23 a31 + a13 a21 a32 − a13 a22 a31
Proposition.
A ∈ Mn (R); there holds det A = det At
AM
Proof. Assume that A = [aij ]. Then At = [aijt ], where aijt = aji
PH
P
det A = f ∈Sn s(f )a1f (1) . . . anf (n) (1)
det At = f ∈Sn s(f )a1f
t t
T.
P P
(1) . . . anf (n) = f ∈Sn s(f )af (1)1 . . . af (n)n (2)
Any f ∈ Sn ; f −1 ∈ Sn , g.
s(f −1 )af −1 (1)1 . . . af −1 (n)n = s(f )af −1 (1)1 . . . af −1 (n)n
on
= s(f )af −1 (1),f (f −1 (1)) . . . af −1 (n),f (f −1 (n))
Du
Proposition.
a11 ... a1n a11 . . . a1n a11 . . . a1n
AM
... ... ...
. . .
. . . . . . . . .
. . . . . .
λbi1 +µci1 . . . λbin +µcin = λ bi1
. . . bin + µ ci1
. . . cin
PH
... ... ...
. . .
. . . . . . . . .
. . . . . .
an1 ... ann an1 . . . ann an1 . . . ann
T.
Proof.
X g.
LHS = s(f )a1f (1) . . . (λbif (i) + µcif (i) ) . . . anf (n)
on
f ∈Sn
Du
=X X
λ s(f )a1f (1) . . . bif (i) . . . anf (n) + µ s(f )a1f (1) . . . cif (i) . . . anf (n)
f ∈Sn f ∈Sn
= RHS
Corollaries:
a11 . . . a 1n a11 . . . a1n a11 . . . a1n
... ... . . . . . . . . . . . . . . . . . . . . .
AM
bi1 +ci1 . . . bin +cin = bi1 . . . bin + ci1 . . . cin
... ... . . . . . . . . . . . . . . . . . . . . .
PH
an1 ... ann an1 . . . ann an1 . . . ann
T.
a11 . . . a1n a11 . . . a1n
... ... ... . . . . . . . . .
g.
λai1 . . . λain = λ ai1 . . . ain
on
... ... ... . . . . . . . . .
Du
Proposition.
. . . ... . . . . . . ... . . .
AM
ai1 ... ain aj1 ... ajn
. . . ... . . . = − . . . ... . . .
PH
aj1 ... ajn ai1 ... ain
. . . ... . . . . . . ... . . .
T.
Proof. Denote B := matrix in RHS
LHS =
X g.
s(f )a1f (1) . . . aif (i) . . . ajf (j) . . . anf (n)
on
f ∈Sn
Du
P
RHS = s(f )b1f (1) . . . bif (i) . . . bjf (j) . . . bnf (n)
P f ∈Sn
= s(f )a1f (1) . . . ajf (i) . . . aif (j) . . . anf (n)
Pf ∈Sn
= f ∈Sn s(f (i j))s(i j) a1f (1) . . . aif (j) . . . ajf (i) . . . anf (n)
P
=− f ∈Sn s(f )a1f (1) . . . anf (n) = −LHS
Proposition.
det A = 0 if A satisfies one of the following:
AM
has one zero row or one zero column
PH
has 2 equal rows or 2 equal columns
has 2 proportional rows or 2 proportional columns
T.
Proof.
a11 . . . a1n
g.
a11 . . . a1n
on
. . . . . . . . . . . . . . . . . .
Du
0 ... 0 = 0 0 ... 0 = 0
. . . . . . . . . . . . . . . . . .
an1 . . . ann an1 . . . ann
Other statements: Exercises
Corollary.
A ∈ Mn (R); Determinant of A does NOT change when multiplying one
AM
row with a number and adding it into another row, i.e.
PH
. . . . . . . . . . . . ... . . .
ai1 . . . ain ai1 + λaj1 . . . ain + λajn
T.
. . . . . . . . . = . . . ... . . .
aj1 . . . ajn aj1 ... ajn
g.
. . . . . . . . . . . . ... ...
on
Du
Corollary.
A ∈ Mn (R); Determinant of A does NOT change when adding into any
row with a linear combination of other rows.
Proposition.
A, B ∈ Mn (R); there holds det(AB) = det A det B
AM
1 + x1 y1 1 + x1 y2 . . . 1 + x1 yn
PH
1 + x2 y1 1 + x2 y2 . . . 1 + x2 yn
Ex. Find det A where A = ...
... ...
1 + xn y1 1 + xn y2 . . . 1 + xn yn
T.
1 x1 0 . . . 0 g. 1 1 ... 1
1 x2 0 . . . 0 y1 y2 . . . yn
on
A= 1 x3 0 . . . 0 0 0 ... 0
Du
. . . . . . . . . . . . . . . . . . . . .
1 xn 0 . . . 0 0 0 ... 0
(
0 when n > 2
det A =
(x2 − x1 )(y2 − y1 ) when n = 2
Definition.
A ∈ Mn (R);
AM
The minor Mi,j is the determinant of the (n − 1) × (n − 1)-matrix
that results from A by removing the ith row and the jth column
PH
The expression (−1)i+j Mi,j is known as cofactor
T.
−2 2 −3
Ex. A = −1 1 3
2 0 −1
g.
on
−1 3 −2 −3
Du
M1,2 =
= 1 − 6 = −5; M2,2 = = 8;
2 −1 2 −1
−2 −3
M3,2 =
= −3
−1 3
Theorem.
AM
a11 a12 . . . a1j ... a1n
a21 a22 . . . a2j ... a2n
PH
. . . . . . . . . ... ... . . .
A= ; Aij := (−1)i+j Mij cofactors of A
ai1 ai2 . . . aij ... ain
T.
. . . . . . . . . ... ... . . .
an1 an2 . . . anj g.
... ann
on
det A = ai1 Ai1 + ai2 Ai2 + . . . + ain Ain
Du
AM
−2 2 −3
1+2
−1 3 2+2
−2 −3
Ex. −1 1 3 = 2(−1)
2 −1 + 1(−1)
2 −1
PH
2 0 −1
3+2
−2 −3
+0(−1)
T.
−1 3
g.
= −2 × (−5) + 1 × 8
on
= 18
Du
Corollary.
AM
a11 a12 . . . a1n
PH
0 a22 . . . a2n
. . . = a11 a22 . . . ann
... . . .
T.
0 0 ... ann
a11
0 . . . 0 g.
on
a21 a22 . . . 0
. . . = a11 a22 . . . ann
... . . .
Du
an1 an2 . . . ann
AM
4 1 7 1 0 r4 →r4 −4r1 0 1 3 1 0
8 7 −1 2 3 r5 →r5 −8r1 0 7 − 9 2 3
PH
1 0 1 0 0 1 0 1 0 0
r2 ↔r4 0 1 3 1 0 r4 →r4 −2r2 0 1 3 1 0
T.
= − 0 0 −4 0 0 = − 0 0 −4 0 0
0 2 8 8 9 r5 →r5 −7r2 0 0
g. 2 6 9
0 7 −9 2 3 0 0 − 30 −5
on
3
Du
1 0 1 0 0 1 0 1 0 0
1
r3 → 4 r3 0 1
3 1 0 r4 →r4 +2r2 0
1 3 1 0
= − 4 0 0 −1
0 0 = − 4 0 0 −1 0 0
0 0
2 6 9 r5 →r5 −30r2 0
0 0 6 9
0 0 −30 −5 3 0 0 0 −5 3
Duong T. PHAM September 30, 2020 20 / 28
Triangularization of matrices
AM
1 0 1 0 0 1 0 1 0 0
PH
5
0
1 3 1 0 r5 →r5 + 6 r4 0
1 3 1 0
−4 0 0 −1 0 0 = − 4 0 0 −1 0 0 = 252
T.
0 0 0 6 9 0 0 0 6 9
21
0 0 0 −5 3 g. 0 0 0 0 2
on
Du
AM
Theorem.
Given A ∈ Mn (R); A is invertible if and only if det A 6= 0. Then
PH
A11 A21 . . . An1
A12 A22 . . . An2
T.
A−1 = det1 A
Here, Aij are cofactors of A
... ... ...
A1n g.
A2n . . . Ann
on
| {z }
:=P
Du
(AP)ii = ai1 Ai1 + ai2 Ai2 + . . . + ain Ain = det A ( Laplace’s formula )
AM
(
bk` = ak` ∀k 6= j, ` = 1, . . . , n
i 6= j; Define B ∈ Mn (R) whose
bj` = ai` ∀` = 1, . . . , n
PH
det B = 0 since rows ith and jth equal.
T.
Using Laplace’s formula
g.
0 = det B = bj1 Bj1 + . . . + bjn Bjn = ai1 Aj1 + . . . + ain Ajn = (AP)ij
on
Hence, AP = det A · In . Similarly, PA = det A · In .
Du
1
Therefore, A−1 = P
det A
AM
1+1
1 2 1+2
−1 2
A11 =(−1) 1 1 = − 1; A12 =(−1) 1 1 =3;
PH
1+3
−1 1 2+1
0 −1
A13 =(−1) 1 1 = − 2; A21 =(−1) 1 1 = − 1;
T.
2+2
1 −1 g. 2+3
1 0
A22 =(−1) 1 1 =2; A23 =(−1) 1 1 = − 1;
on
0 −1 1 −1
Du
3+1 3+2
A31 =(−1) 1 2 =1; A32 =(−1) −1 2 = −1;
−1 −1 1
1 0 = 1; ⇒ A−1 = 1 3
A33 =(−1)3+3 1 2 − 1
−1 1
−2 −1 1
Duong T. PHAM September 30, 2020 24 / 28
Find inverse of a matrix
AM
1 0 1 1 0 1 0 0 −→ 1 0 1 10 1 0 0
A=1 1 0 1 0
0 1 0 1 1 0 10 0 1 0
PH
1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1
1 31 1 1 1
1 1 1 1
1 1 1 1 3 3 3 1 1 1
T.
3 3 3 3
1 0 1 10 1 0 0 0 −1 0 0 − 1 2
− 1
− 1
→ 13 3
1 2
3 3
1 1 0 10 0 1 g.
0 0 0 −1 0 − 3 − 3 3 − 13
0 −1 − 13 − 13 − 31 2
on
1 1 1 0 0 0 0 1 0 0 3
Du
0 − 23 1 1 1
2 1 1 1
1 0 0 3 3 3 −3 3 3 3
0 −1 0 1 2 1 1 1 2 1 1
− 31
0 − − −
3 3 3 →I 3 3 3 3
4 1
0 0 −1 0 −
3 − 13 2
3 − 31 3
1
3 − 23 1
3
0 0 0 −1 − 13 − 13 − 31 2
3
1
3
1
3
1
3 − 2
3
AM
A . = . =⇒ . = A .
xn yn xn yn
PH
x + x + x = y1
0 1 1 1 x1 y1
2 3 4
1 0 1 1 x2 y2 x1 + x3 + x4 = y2
T.
Ex. 1 1 0 1 x3 = y3 ⇔ x + x + x
1 2 4 = y3
1 1 1 0 x4 g. y4
x + x + x
1 2 3 = y4
on
⇔
x1 = − 23 y1 + 31 y 2 + 31 y3 + 13 y4
Du
2 1 1 1
−3 3 3 3
x = 1 y − 2 y 2 + 1 y + 1 y 1
− 23 1 1
2 3 1 3 3 3 4 ⇒A−1 =
3
3 3 3
1 1
x 3 = 1
y1 + 1
y 2 − 2
y3 + 1
y 4
3 3 − 23 1
3
3 3 3 3 1 1 1
− 23
x = − 2 y + 1 y 2 + 1 y − 2 y
3 3 3
4 3 1 3 3 3 3 4
a11 a12 ... a1n x1 b1
a21 a22 ... a2n x2 b2
AM
... ...
= and det A 6= 0
... ... ...
PH
an1 an2 ... ann xn bn
| {z }
A
T.
The system has a unique solution
g.
xi =
det Ai
, i = 1, 2, . . . , n,
on
det A
Du
AM
We have det A = 4 6= 0 =⇒ The system has a unique solution
PH
1 −1 1 1 1 1
T.
1 1 −1 1 1 −1
3 1 1 4 1 3 1 4
x1 = = = 1;
g. x2 = = = 1;
det A 4 det A 4
on
Du
1 −1 1
1 1 1
1 1 3 4
x3 = = = 1.
det A 4