資訊科學數學 14 : determinants & inverses

67
1 資資資資資資 資資資資資資 14 14 : : Determinants & Inverses Determinants & Inverses 陳陳陳陳陳陳陳 陳陳陳陳陳陳陳 (Kuang-Chi Chen) (Kuang-Chi Chen) [email protected] [email protected]

Upload: miracle

Post on 11-Jan-2016

137 views

Category:

Documents


0 download

DESCRIPTION

資訊科學數學 14 : Determinants & Inverses. 陳光琦助理教授 (Kuang-Chi Chen) [email protected]. Linear Equations and Matrices Determinants. 3.1 Determinants. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: 資訊科學數學 14  : Determinants & Inverses

11

資訊科學數學資訊科學數學 14 14 ::

Determinants & InversesDeterminants & Inverses

陳光琦助理教授 陳光琦助理教授 (Kuang-Chi Chen)(Kuang-Chi Chen)[email protected]@mail.tcu.edu.tw

Page 2: 資訊科學數學 14  : Determinants & Inverses

22

Linear Equations and MatricesLinear Equations and Matrices

DeterminantsDeterminants

Page 3: 資訊科學數學 14  : Determinants & Inverses

33

3.1 Determinants3.1 Determinants• With each With each nnnn matrix matrix AA it is possible to associate a it is possible to associate a

scalar scalar det(det(AA)), called the , called the determinantdeterminant of the matrix, w of the matrix, whose value will tell us whether the matrix is hose value will tell us whether the matrix is singular singular or notor not..

• Case 1: 1Case 1: 11 matrices 1 matrices

- If - If AA = ( = (aa), then ), then AA will have a multiplicative inverse will have a multiplicative inverse iff iff aa≠0 ≠0 ..

- - AA is nonsingular iff det( is nonsingular iff det(AA))≠≠0 .0 .

Page 4: 資訊科學數學 14  : Determinants & Inverses

44

222 Matrices2 Matrices

• Case 2: 2Case 2: 22 matrices 2 matrices

- Let - Let AA = . = .

- - AA will be will be nonsingularnonsingular iff iff det(det(AA) = ) = aa1111aa2222 – – aa1212aa2121≠ ≠ 00 . .

11 12

21 22

a a

a a

Page 5: 資訊科學數學 14  : Determinants & Inverses

55

333 Matrices3 Matrices• Case 3: 3Case 3: 33 matrices 3 matrices

- Let - Let AA = . = .

- - AA will be will be nonsingularnonsingular iff iff

det(det(AA) = ) = aa1111aa2222aa3333 + + aa1212aa3131aa23 23 + + aa1313aa2121aa3232 – – aa1111aa3232aa23 23 – – aa1212

aa2121aa33 33 – – aa1313aa3131aa22 22 ≠ ≠ 00 . .

11 12 13

21 22 23

31 32 33

a a a

a a a

a a a

Page 6: 資訊科學數學 14  : Determinants & Inverses

66

Example 4 & 5Example 4 & 5

• Example 4Example 4

If If AA = [ = [aa1111] is a 1] is a 11 matrix, then det(1 matrix, then det(AA) = ) = aa1111 . .

• Example 5Example 5

IfIf

⇒⇒ det(det(AA) = ) = aa1111aa2222 – – aa1212aa2121

⇒⇒ det(det(AA) = (2)(5) – (-3)(4)) = (2)(5) – (-3)(4) = 22 = 22

11 12

21 22

a aA

a a

54

32A

Page 7: 資訊科學數學 14  : Determinants & Inverses

77

Example 6 & 7Example 6 & 7• Example 6Example 6

IfIf

⇒⇒ det(det(AA) =) = aa1111aa2222aa3333 + + aa1212aa3131aa23 23 + + aa1313aa2121aa3232

– – aa1111aa3232aa23 23 – – aa1212aa2121aa33 33 – – aa1313aa3131aa22 22

• Example 7Example 7

IfIf ⇒ ⇒ det(det(AA) = (1)(1)(2) + (3)(2)(1)) = (1)(1)(2) + (3)(2)(1) + (2)(3)(3)+ (2)(3)(3)

– – (3)(1)(3)(3)(1)(3) – (1)(1)(3)– (1)(1)(3) – (2)(2)(2) = 6– (2)(2)(2) = 6

333231

232221

131211

aaa

aaa

aaa

A

213

312

321

A

Page 8: 資訊科學數學 14  : Determinants & Inverses

88

Properties of DeterminantsProperties of Determinants

• Theorem 3.1Theorem 3.1

The determinants of a matrix and its The determinants of a matrix and its transposetranspose are are eqequalual, i.e., det(, i.e., det(AA) = det() = det(AATT).).

Page 9: 資訊科學數學 14  : Determinants & Inverses

99

Example 8Example 8• Example 8Example 8

IfIf

⇒ ⇒ det(det(AATT) = (1)(1)(2) + (3)(1)(2)) = (1)(1)(2) + (3)(1)(2) + (2)(3)(3)+ (2)(3)(3)

– – (3)(1)(3)(3)(1)(3) – (1)(1)(3)– (1)(1)(3) – (2)(2)(2)– (2)(2)(2)

= 6 = det(= 6 = det(AA))

233

112

321TA

213

312

321

A

Page 10: 資訊科學數學 14  : Determinants & Inverses

1010

Theorem 3.2 & 3.3Theorem 3.2 & 3.3

• Theorem 3.2Theorem 3.2

If matrix If matrix BB results from matrix results from matrix AA by by interchanginginterchanging t two rows (or two columns) of wo rows (or two columns) of AA, then, then

det(det(BB) = -det() = -det(AA).).

• Theorem 3.3Theorem 3.3

If If two rows (or columns) of two rows (or columns) of AA are equal are equal, then, then

det(det(AA) = 0.) = 0.

Page 11: 資訊科學數學 14  : Determinants & Inverses

1111

Example 9 & 10Example 9 & 10• Example 9Example 9

IfIf

• Example 10Example 10

IfIf

712

23 and 7

23

12

0

321

701

321

Page 12: 資訊科學數學 14  : Determinants & Inverses

1212

Theorem 3.4Theorem 3.4

• Theorem 3.4Theorem 3.4

If a row (or column) of If a row (or column) of AA consists entirely of consists entirely of zeroszeros, t, then det(hen det(AA) = 0.) = 0.

• Example 11Example 11

0

000

654

321

Page 13: 資訊科學數學 14  : Determinants & Inverses

1313

Theorem 3.5Theorem 3.5• Theorem 3.5Theorem 3.5

If If BB is obtained from is obtained from AA by by multiplyingmultiplying a row (colum a row (column) of n) of AA by a real number by a real number cc, then, then

det(det(BB) = ) = cc det( det(AA)) . .

• Example 12Example 12

1814641

1132

121

312

121

62

Page 14: 資訊科學數學 14  : Determinants & Inverses

1414

Example 13Example 13

• Example 13Example 13

0032

141

151

121

32

341

351

321

2

682

351

321

Page 15: 資訊科學數學 14  : Determinants & Inverses

1515

Theorem 3.6Theorem 3.6• Theorem 3.6Theorem 3.6

If If BB = [ = [bbijij] is obtained from ] is obtained from AA = [ = [aaijij] by adding to eac] by adding to eac

h element of the h element of the rrthth row (column) of row (column) of AA a constant a constant cc ti times the corresponding element of the mes the corresponding element of the ssthth row (colum row (column) n) rr≠≠ss of of AA, then det(, then det(BB) = det() = det(AA) .) .

• Example 14Example 14

101

312

905

101

312

321

Page 16: 資訊科學數學 14  : Determinants & Inverses

1616

Theorem 3.7Theorem 3.7• Theorem 3.7Theorem 3.7

If a matrix If a matrix AA = [ = [aaijij] is ] is upper (lower) triangularupper (lower) triangular, then, , then,

then det(then det(AA) = ) = aa11 11 aa22 22 … … aannnn . .

• Corollary 1.3Corollary 1.3

The determinant of a The determinant of a diagonal matrixdiagonal matrix is the product o is the product of the entries on its main diagonal.f the entries on its main diagonal.

Page 17: 資訊科學數學 14  : Determinants & Inverses

1717

Example 15Example 15

• Example 15Example 15

2 3 4 3 0 0 5 0 0

0 4 5 , 2 5 0 , 0 4 0

0 0 3 6 8 4 0 0 6

A B C

120)det( ,60)det( ,24)det( CBA

Page 18: 資訊科學數學 14  : Determinants & Inverses

1818

Elementary OperationsElementary Operations• Elementary row and elementary column operationsElementary row and elementary column operations

I - Interchange rows (columns) I - Interchange rows (columns) ii and and jj : :

rrii ⇔ ⇔ rrjj ( (ccii ⇔ ⇔ ccjj ) )

II - Replace row (column) II - Replace row (column) ii by a nonzero value by a nonzero value kk times times row (column) row (column) ii : :

krkrii ⇔ ⇔ rrii ( (kckcii ⇔ ⇔ ccii ) )

III - Replace row (column) III - Replace row (column) jj by a nonzero value by a nonzero value kk times times row (column) row (column) i+ i+ row (column) row (column) jj : :

krkrii + + rrjj ⇔ ⇔ rrjj ( (kckcii + + ccjj ⇔ ⇔ ccjj ) )

Page 19: 資訊科學數學 14  : Determinants & Inverses

1919

… … then …then …

det( ) det( ),

det( ) det( )

det ( ) det( ),

i j

i i

i j j

r r

kr r

kr r r

A A i j

A k A

A A i j

Page 20: 資訊科學數學 14  : Determinants & Inverses

2020

Example 16Example 16• E.g. 16E.g. 16

642

523

234

A

13 32

1 3

4 3 2

det( ) 2det( ) 2det ( 3 2 5 )

1 2 3

4 3 2 1 2 3

2det( 3 2 5 ) ( 1) 2det( 3 2 5 )

1 2 3 4 3 2

r r

r r

A A

Page 21: 資訊科學數學 14  : Determinants & Inverses

2121

Example 16Example 16 (cont’d)(cont’d)

1 2 2

1 3 3

52 3 38

-3r4

304

1 2 3 1 2 3

2det( 3 2 5 ) 2det( 0 -8 4 )

4 3 2 0 -5 10

1 2 3 1 2 3

2det( 0 -8 4 ) 2det( 0 -8 4 )

0 -5 10 0 0

r rr r r

r r r

det( ) 2(1)( 8)( 30 / 4) 120A

Page 22: 資訊科學數學 14  : Determinants & Inverses

2222

Theorem 3.8Theorem 3.8• Theorem 3.8Theorem 3.8

The determinant of a product of two matrices is the prThe determinant of a product of two matrices is the product of their determinants oduct of their determinants det(det(ABAB) = det() = det(AA)det()det(BB)) . .

• Example 17Example 17

43

21A

21

12B

2A 5B

510

34AB BAAB 10

Page 23: 資訊科學數學 14  : Determinants & Inverses

2323

Example 17 Example 17 (cont’d)(cont’d)

• RemarkRemark

ABAB≠≠BABA

||BABA| = || = |BB| || |AA|= -10 = ||= -10 = |ABAB||

107

01BA

Page 24: 資訊科學數學 14  : Determinants & Inverses

2424

Corollary 3.2Corollary 3.2

• Corollary 3.2Corollary 3.2

If If AA is is nonsingularnonsingular, then , then det(det(AA) ) ≠≠ 0 0,,

thus thus det(det(AA-1-1) = 1/det() = 1/det(AA))..

If If AA is singular, then det( is singular, then det(AA) = 0) = 0

( 1 = |( 1 = |II| = || = |AAAA-1-1| = || = |AA| || |AA-1-1| )| )

Page 25: 資訊科學數學 14  : Determinants & Inverses

2525

Example 18Example 18• Example 18Example 18

43

21A 2)det( A

1 2 1

3/ 2 1/ 2A

)det(

1

2

1)det( 1

AA

Page 26: 資訊科學數學 14  : Determinants & Inverses

2626

Cofactor Expression and Cofactor Expression and ApplicationsApplications

Page 27: 資訊科學數學 14  : Determinants & Inverses

2727

3.2 Cofactor Expression and 3.2 Cofactor Expression and ApplicationsApplications

Cofactor expression and applicationsCofactor expression and applications

• Definition – Minor and cofactorDefinition – Minor and cofactor

Let Let AA = [ = [aaijij] be an ] be an nnnn matrix. Let matrix. Let MMijij be the ( be the (nn-1)-1)

((nn-1) -1) submatrixsubmatrix of of AA obtained by deleting the obtained by deleting the iithth row row and and jjthth column of column of AA. The . The determinant det(determinant det(MMijij)) is calle is calle

d the d the minorminor of of aaijij. The . The cofactor cofactor AAijij of of aaijij is defined as is defined as

)( det)1( ijji

ij MA

Page 28: 資訊科學數學 14  : Determinants & Inverses

2828

Example 1Example 1• E.g. 1E.g. 1

LetLet

217

654

213

A

12

23

31

4 6det( ) 8 42 34

7 2

3 1det( ) 3 7 10

7 1

1 2det( ) 6 10 16

5 6

M

M

M

1 212 12

2 323 23

3 131 31

( 1) det( ) ( 1)( 34) 34

( 1) det( ) ( 1)(10) 10

( 1) det( ) (1)( 16) 16

A M

A M

A M

Page 29: 資訊科學數學 14  : Determinants & Inverses

2929

Theorem 3.9Theorem 3.9

• Theorem 3.9Theorem 3.9

Let Let AA = [ = [aaijij] be an ] be an nnnn matrix. Then matrix. Then

for each 1for each 1≤ ≤ ii ≤ ≤ nn,,

det(det(AA) = ) = aaii11AAii11 + + aaii22AAii22 + … + + … + aaininAAinin , and , and

for each 1for each 1≤ ≤ jj ≤ ≤ nn,,

det(det(AA) = ) = aa11jjAA11jj + + aa22jjAA22jj + … + + … + aanjnjAAnjnj . .

Page 30: 資訊科學數學 14  : Determinants & Inverses

3030

Example 2Example 2To evaluate the determinantTo evaluate the determinant

3202

3003

3124

4321

3 1 3 2

3 3 3 4

1 2 3 42 3 4 1 3 4

4 2 1 3( 1) (3) 2 1 3 ( 1) (0) 4 1 3

3 0 0 30 2 3 2 2 3

2 0 2 3

1 2 4 1 2 3

( 1) (0) -4 2 3 ( 1) ( 3) 4 2 1

2 0 3 2 0 2

( 1)(3)(20) 0 0 ( 1)( 3)( 4) 48

Page 31: 資訊科學數學 14  : Determinants & Inverses

3131

Example 3Example 3Consider the determinant of the matrixConsider the determinant of the matrix

4 1 4

1 2 1

3 1 4

4

1 2 3 4 1 2 3 5

4 2 1 3 4 2 1 1

3 0 0 3 3 0 0 0

2 0 2 3 2 0 2 5

2 3 5 0 4 6

( 1) (3) 2 1 1 ( 1) (3) 2 1 1

0 2 5 0 2 5

( 1) (3)( 2)( 8) 48

c c c

r r r

Page 32: 資訊科學數學 14  : Determinants & Inverses

3232

Theorem 3.10Theorem 3.10• Theorem 3.10Theorem 3.10

If If AA = [ = [aaijij] be an ] be an nnnn matrix, then matrix, then

aaii11AAkk11 + + aaii22AAkk22 + … + + … + aaininAAknkn = 0, for = 0, for ii≠≠kk , ,

aa11jjAA11kk + + aa22jjAA22kk + … + + … + aanjnjAAnknk = 0, for = 0, for jj≠≠kk . .

Page 33: 資訊科學數學 14  : Determinants & Inverses

3333

Example 4Example 4• E.g. 4E.g. 4

033142191

032145194

354

211

1424

311

1925

321

254

132

321

231322122111

233322322131

3223

2222

1221

AaAaAa

AaAaAa

A

A

A

A

Page 34: 資訊科學數學 14  : Determinants & Inverses

3434

AdjointAdjoint• Definition – AdjointDefinition – Adjoint

Let Let AA = [ = [aaijij] be an ] be an nnnn matrix. The matrix. The nnnn matrix matrix adjadj AA,,

called the called the adjoint of adjoint of AA, is the matrix whose , is the matrix whose jj, , iithth ele element is the ment is the cofactor cofactor AAijij of of aaijij . Thus . Thus

nnnn

n

n

AAA

AAA

AAA

adjA

21

22212

12111

Page 35: 資訊科學數學 14  : Determinants & Inverses

3535

RemarkRemark

• RemarkRemark

The adjoint of The adjoint of AA is formed by taking the is formed by taking the transptransposeose of the matrix of of the matrix of cofactorscofactors AAijij of the elemen of the elemen

ts of ts of AA..

Page 36: 資訊科學數學 14  : Determinants & Inverses

3636

Example 5Example 5• Example 5Example 5

Compute adj Compute adj AA

301

265

123

A

Page 37: 資訊科學數學 14  : Determinants & Inverses

3737

SolutionSolution

1 1

11

1 2

12

1 3

13

2 1

21

2 2

22

2 3

23

6 21 18

0 3

5 21 17

1 3

5 61 6

1 0

2 11 6

0 3

3 11 10

1 3

3 21 2

1 0

A

A

A

A

A

A

2865

231

125

131

1026

121

3333

2332

1331

A

A

A

2826

11017

10618

adj Then, A

Page 38: 資訊科學數學 14  : Determinants & Inverses

3838

Theorem 3.11Theorem 3.11

• Theorem 3.11Theorem 3.11

If If AA = [ = [aaijij] be an ] be an nnnn matrix, then matrix, then

AA(adj (adj AA) = (adj ) = (adj AA))AA = det( = det(AA) ) IInn . .

Page 39: 資訊科學數學 14  : Determinants & Inverses

3939

Example 6Example 6• E.g. 6 E.g. 6

Consider the matrixConsider the matrix

3 2 1 18 6 10 94 0 0 1 0 0

5 6 2 17 10 1 0 94 0 94 0 1 0

1 0 3 6 2 28 0 0 94 0 0 1

and

18 6 10 3 2 1 1 0 0

17 10 1 5 6 2 94 0 1 0

6 2 28 1 0 3 0 0 1

301

265

123

A

Page 40: 資訊科學數學 14  : Determinants & Inverses

4040

Corollary 3.3Corollary 3.3• Corollary 3.3Corollary 3.3

If If AA = [ = [aaijij] be an ] be an nnnn matrix and det( matrix and det(AA))≠≠0, then0, then

A

A

A

A

A

A

A

A

A

A

A

AA

A

A

A

A

A

adjAA

A

detdetdet

detdetdet

detdetdet

det

1 1

Page 41: 資訊科學數學 14  : Determinants & Inverses

4141

Example 7Example 7• Example 7Example 7

Consider the matrixConsider the matrix

Then det(Then det(AA) = -94, and) = -94, and

3 2 1

5 6 2

1 0 3

A

18 6 1094 94 94

1 17 10 194 94 94

6 128294 94 94

1

det( )A adjA

A

Page 42: 資訊科學數學 14  : Determinants & Inverses

4242

Theorem 3.12Theorem 3.12

• Theorem 3.12Theorem 3.12

A matrix A matrix AA = [ = [aaijij] is ] is nonsingularnonsingular iff iff det(det(AA) ) ≠≠ 0 0..

• Corollary 3.4Corollary 3.4

For an For an nnnn matrix matrix AA, the homogeneous system , the homogeneous system AAx x = 0 has a = 0 has a nontrival solutionnontrival solution iff iff det(det(AA) = 0) = 0..

Page 43: 資訊科學數學 14  : Determinants & Inverses

4343

Example 8Example 8• Example 8Example 8

Let Let AA be a 4x4 matrix with det( be a 4x4 matrix with det(AA) = -2) = -2

(a) describe the set of all solutions to the homoge(a) describe the set of all solutions to the homogeneous system neous system AAxx = 0. = 0.

(b) If (b) If AA is transformed to reduced row echelon form is transformed to reduced row echelon form BB, , what is what is BB??

(c) Given an expression for a solution to the linear syst(c) Given an expression for a solution to the linear system em AAxx = = bb, where , where bb = [ = [bb11 , , bb22 , , bb33 , , bb44 ] ]TT . .

(d) Can the linear system (d) Can the linear system AAxx = = bb have more than one sol have more than one solution? Explain.ution? Explain.

(e) Does (e) Does AA-1-1 exist? exist?

Page 44: 資訊科學數學 14  : Determinants & Inverses

4444

Solutions of Example 8Solutions of Example 8• SolutionsSolutions

(a) Since det((a) Since det(AA))≠≠0, 0, AxAx = 0 has = 0 has only the trivial solutiononly the trivial solution..

(b) Since det((b) Since det(AA))≠≠0, 0, AA is a nonsingular matrix, so is a nonsingular matrix, so BB = = IInn

(c) A solution to the given system is given by (c) A solution to the given system is given by xx = = AA-1-1bb

(d) No. The solution is unique.(d) No. The solution is unique.

(e) Yes. (e) Yes.

Page 45: 資訊科學數學 14  : Determinants & Inverses

4545

Nonsingular EquivalenceNonsingular Equivalence• List of nonsingular equivalenceList of nonsingular equivalence

The following statements are equivalent.The following statements are equivalent.

1.1. AA is is nonsingularnonsingular..

2.2. xx = 0 = 0 is the is the only solutiononly solution to to AxAx = 0. = 0.

3.3. AA is is row equivalencerow equivalence to to IInn . .

4. The linear system 4. The linear system AAxx = = bb has a has a unique solutionunique solution for for every every nn1 matrix 1 matrix bb..

5. 5. det(det(AA))≠≠00 . .

Page 46: 資訊科學數學 14  : Determinants & Inverses

4646

DeterminantsDeterminants

• Linearly independentLinearly independent

• NonsingularNonsingular

• Trivial solution Trivial solution xx = 0 to = 0 to AxAx = 0 = 0

• det(det(AA) ) ≠≠ 0 0

Page 47: 資訊科學數學 14  : Determinants & Inverses

4747

DeterminantsDeterminants

• Linearly dependentLinearly dependent

• SingularSingular

• Nontrivial solution to Nontrivial solution to AAxx = 0 = 0

• det(det(AA) = 0) = 0

Page 48: 資訊科學數學 14  : Determinants & Inverses

4848

Cramer’s RuleCramer’s RuleTheorem 3.13 (Cramer’s Rule)Theorem 3.13 (Cramer’s Rule)

Let Let aa1111xx11 + + aa1212xx22 + … + + … + aa11nnxxnn = = bb11

aa2121xx11 + + aa2222xx22 + … + + … + aa22nnxxnn = = bb22

……

aann11xx11 + + aann22xx22 + … + + … + aannnnxxnn = = bbnn

Then,Then,

xx11 = det( = det(AA11)/det()/det(AA)) , , xx22 = det( = det(AA22)/det()/det(AA)) , … , , … ,

xxnn = det( = det(AAnn)/det()/det(AA)) . .

Page 49: 資訊科學數學 14  : Determinants & Inverses

4949

Cramer’s RuleCramer’s Rule

Cramer’s Rule for solving the linear system Cramer’s Rule for solving the linear system AAxx = = bb, wh, where ere AA is is nnnn, is as follows:, is as follows:

Step 1. Compute det(Step 1. Compute det(AA). If det(). If det(AA) = 0, Cramer’s rule is ) = 0, Cramer’s rule is not applicable. Use not applicable. Use Gauss-Jordan ReductionGauss-Jordan Reduction..

Step 2. If det(Step 2. If det(AA))≠≠0, for each 0, for each ii,,

xxii = det( = det(AAii)/det()/det(AA)) , ,

where where AAii is the matrix obtained from is the matrix obtained from AA by replacing t by replacing t

he he iithth column of column of AA by by bb. .

Page 50: 資訊科學數學 14  : Determinants & Inverses

5050

Example 9Example 9

• Consider the following linear system:Consider the following linear system:

-2-2xx11 + 3 + 3xx22 – – xx33 = 1 = 1

xx11 + 2 + 2xx22 – – xx33 = 4 = 4

-2-2xx11 – 2 – 2xx22 + + xx33 = -3 = -3

ThenThen 2 3 1

1 2 1 2

2 1 1

A

Page 51: 資訊科學數學 14  : Determinants & Inverses

5151

Example 9Example 9 (cont’d) (cont’d)

2

2 1 1

1 4 1

2 3 1 63

2x

A

1

1 3 1

4 2 1

3 1 1 42

2x

A

3

2 3 1

1 2 4

2 1 3 84

2x

A

Hence,Hence,

Page 52: 資訊科學數學 14  : Determinants & Inverses

5252

Polynomial Interpolation Polynomial Interpolation RevisitedRevisited

• Polynomial Interpolation RevisitedPolynomial Interpolation Revisited

To find a quadratic polynomial that interpolates the To find a quadratic polynomial that interpolates the following points:following points:

((xx11, , yy11), (), (xx22, , yy22), (), (xx33, , yy33),),

where where xx11≠≠xx22 , , xx11≠ ≠ xx33 , , xx22≠ ≠ xx33 . .

Page 53: 資訊科學數學 14  : Determinants & Inverses

5353

… … more …more …

The polynomial has the form:The polynomial has the form:

yy = = aa22xx22 + + aa11xx + + aa00 . .

The corresponding linear systemThe corresponding linear system

yy11 = = aa22xx1122 + + aa11xx11 + + aa00 , ,

yy22 = = aa22xx2222 + + aa11xx22 + + aa00 , ,

yy33 = = aa22xx3322 + + aa11xx33 + + aa00 . .

Page 54: 資訊科學數學 14  : Determinants & Inverses

5454

… … more …more …

The coefficient matrixThe coefficient matrix

The Vandermount determinantThe Vandermount determinant

((xx11 – – xx22 )( )( xx11 – – xx33 )( )( xx22 – – xx33 ) )

1

1

1

3

2

3

2

2

2

1

2

1

xx

xx

xx

Page 55: 資訊科學數學 14  : Determinants & Inverses

5555

Page 56: 資訊科學數學 14  : Determinants & Inverses

5656

Linear Equations and MatricesLinear Equations and Matrices

LU-FactorizationLU-Factorization

Page 57: 資訊科學數學 14  : Determinants & Inverses

5757

LU-FactorizationLU-Factorization

• 1.8 LU-Factorization1.8 LU-Factorization

If a square matrix can be reduced to upper If a square matrix can be reduced to upper triangular form using only 3 row operations, triangular form using only 3 row operations, then it is possible to represent the then it is possible to represent the reduction reduction processprocess in terms of a matrix factorization. in terms of a matrix factorization.

Page 58: 資訊科學數學 14  : Determinants & Inverses

5858

Type I OperationType I Operation

An elementary matrix of type I is a matrix obtained by An elementary matrix of type I is a matrix obtained by interchanginginterchanging two rows of identity matrix two rows of identity matrix II..

ExampleExample1

1 1

1

1

0 1 0 0 1 0

1 0 0 , 1 0 0

0 0 1 0 0 1

Interchange the first row and

the second row of

Interchange the first column

and the second column of

E E

E A

A

AE

A

Page 59: 資訊科學數學 14  : Determinants & Inverses

5959

Type II OperationType II Operation

An elementary matrix of type II is a matrix obtained An elementary matrix of type II is a matrix obtained by by multiplyingmultiplying a row of a row of II byby a nonzero constant.a nonzero constant.

ExampleExample

12 2

2

2

1 0 0 1 0 0

0 1 0 , 0 1 0

0 0 3 0 0 1/ 3

Multiply the third row of by 3

Multiply the third column of by 3

E E

E A A

AE A

Page 60: 資訊科學數學 14  : Determinants & Inverses

6060

Type III OperationType III Operation

An elementary matrix of type III is a matrix obtained An elementary matrix of type III is a matrix obtained from from II by adding a multiple of one row to another row. by adding a multiple of one row to another row.

ExampleExample1

3 3

3

3

1 0 3 1 0 3

0 1 0 , 0 1 0

0 0 1 0 0 1

Add three times the third row to

the first row of

Add three times the third column

to the first column of

E E

E A

A

AE

A

Page 61: 資訊科學數學 14  : Determinants & Inverses

6161

In GeneralIn GeneralIn general, the elementary matrix by adding In general, the elementary matrix by adding mm times times of row of row ii to row to row jj

1000

10

10

1

,

1000

10

10

1

133

m

E

m

E

Row j

Column iaji

Page 62: 資訊科學數學 14  : Determinants & Inverses

6262

TheoremTheorem• TheoremTheorem

If If AA and and BB are are nonsingular squarenonsingular square matrices, then matrices, then ABAB i is also s also nonsingularnonsingular..

i.e. (i.e. (ABAB))-1-1 = = BB-1-1AA-1-1 . .

In general, if In general, if EE11, , EE22, …, , …, EEkk are all are all nonsingularnonsingular, then t, then t

he product he product EE11EE22 … … EEkk is also is also nonsingularnonsingular and and

((EE11EE22 … … EEkk ) )-1-1 = = EEkk-1-1 … … EE22

-1 -1 EE11-1-1 . .

Page 63: 資訊科學數學 14  : Determinants & Inverses

6363

Row EquivalenceRow Equivalence

• A matrix A matrix BB is is row equivalent torow equivalent to AA if there exists a fin if there exists a finite sequence ite sequence EE11EE22 … … EEkk of elementary matrices such of elementary matrices such

that that BB = = EEkk … … EE22 EE11AA . .

Page 64: 資訊科學數學 14  : Determinants & Inverses

6464

LU-FactorizationLU-Factorization

• LU-FactorizationLU-Factorization

If a square matrix can be reduced to upper If a square matrix can be reduced to upper triangular form using only 3 row operations, triangular form using only 3 row operations, then it is possible to represent the then it is possible to represent the reduction reduction processprocess in terms of a matrix factorization. in terms of a matrix factorization.

Page 65: 資訊科學數學 14  : Determinants & Inverses

6565

ExampleExample• LetLet

914

251

242

A

1

1 21

2 4 2 2 4 2

1 5 2 0 3 1

4 1 9 4 1 9

1 0 0

1/ 2 1 0 , 1/ 2

0 0 1

E A

E a

Page 66: 資訊科學數學 14  : Determinants & Inverses

6666

(cont’d)(cont’d)

2

2 31

2 4 2 2 4 2

0 3 1 0 3 1

4 1 9 0 9 5

1 0 0

0 1 0 , 2

2 0 1

E A

E a

3

3 32

2 4 2 2 4 2

0 3 1 0 3 1

0 9 5 0 0 8

1 0 0

0 1 0 , 3

0 3 1

E A

E a

Page 67: 資訊科學數學 14  : Determinants & Inverses

6767

(cont’d)(cont’d)

3 2 1

1 1 1 13 2 1 1 2 3

1 1 11 2 3

2 4 2

0 3 1

0 0 8

( )

1 0 01 0 0 1 0 0

11 0 0 1 0 0 1 0

22 0 1 0 3 10 0 1

E E E A U

A E E E U E E E U LU

L E E E