15
MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS SOLVING LINEAR SYSTEM BY MATRIX INVERSION CRAMER’S RULE LEAST SQUARE METHOD

MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

  • Upload
    others

  • View
    4

  • Download
    1

Embed Size (px)

Citation preview

Page 1: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS

SOLVING LINEAR SYSTEM BY MATRIX INVERSIONCRAMER’S RULELEAST SQUARE METHOD

Page 2: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 2

Solving Linear System by Matrix Inversion

If A is an invertible nxn matrix, then for each matrix nx1 b, the system of linear equations Ax=b has exactly one solution, namely x=A-1b

Example

Solve system of linear below

=

321

221321311

zyx

Page 3: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 3

Solving Linear System by Matrix Inversion

−−

−== −

321

110011342

1bAX

−−

−=−

110011342

1A

We had A-1 (see System of Linear Equations PPT page 1)

Solution of linear system is

−=

113

Page 4: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 4

Cramer’s Rule

( ) ( ) ( ))det(

det,...,)det(

det,)det(

det 22

11 A

AxAAx

AAx n

n ===

Cramer’s Rule

If Ax=b is a system of linear equations in n variable such that det(A)≠0, then system has one solution. When Aj is the matrix obtained by replacing the entries in the j th column of A by the entries of b, then the solution of system is

Page 5: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 5

Cramer’s Rule

4...212321543

===A

=

101

212321543

zyx

ExampleSolve system of linear equations below

Solution

See Determinant ppt page 9

Page 6: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 6

Cramer’s Rule

101

101

=== ...122143

3A

Solution (continued)

101

=== ...223153

2A

=== ...213254

1A 3

0

-1

41

41,0

40,1

44

321 −=−

===== xxx

Solution of system

Page 7: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 7

Least Square Fitting to DataLeast Squares Fit of a Straight Line

a Straight Line Model

0123456789

101112131415

0 1 2 3 4 5 6

X

Y

Suppose we want to fit a straight line y = a+bx to the experimentally determined point (x1,y1), (x2,y2),…, (xn,yn)

Y=a+bx a= ?b= ?

Page 8: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 8

Least Square Fitting to DataLeast Squares Fit of a Straight LineLet we have data (x1,y1), (x2,y2),…, (xn,yn)

If the data points are collinear, the line would pass through all n points and so the unknown coefficient a and b would satisfy

y1=a+bx1

y2=a+bx2

:

yn=a+bxn

=

nn y

yy

ba

x

xx

MMM2

1

2

1

1

11

M v = y

Page 9: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 9

Least Square Fitting to DataLeast Square Solution

v is given by formula v = (MTM)-1 MTy

Example

Find the least square straight line fit to the four points

(0,1), (1,3), (2,4) and (3,4)

Page 10: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 10

Least Square Fitting to Data

4431

14664

−32101111

2337

101

Solution

We have

MTM= (MTM)-1 =

v = (MTM)-1 MTy =

=

4431

y

=15.1

−2337

101

=

31211101

M

The least square straight line fit is y = 1.5 + x

Page 11: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 11

Least Square Fitting to DataFitting of a Quadratic Curve to DataLet we have data (x1,y1), (x2,y2),…, (xn,yn)

The technique described for fitting a straight line to data points generalizes easily to fitting a polynomial of any specified degree to data point.

y1=a0+a1x+a2x2

=

nnn y

yy

aaa

x

xx

x

xx

MMMM2

1

2

1

0

2

22

21

2

1

1

11

Let us attempt to fit a polynomial of fixed degree 2

We have

M v = y

Least Square Solution

v is given by formula

v = (MTM)-1 MTy

Page 12: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 12

Least Square Fitting to Data

Example

Find the polynomial of fixed degree 2 fit to the four points

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

1886862624

Solution

We have

MTM = (MTM)-1 =

=

4331

y

−−−−

5555935311

201

−=

4110

21111101

M

Page 13: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 13

Least Square Fitting to DataSolution

4431

−−−−

=411021101111

5555935311

201

v = (MTM)-1 MTy

−=201638

201

The polynomial of fixed degree 2 is

2

2016

2038 xxy +−=

Page 14: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 14

Exercises1. Let we have system of linear equations Ax=b

=

123

011132212

zyx

a. Find A-1

b. Solve that system using A-1

2. Let we have system of linear equations Ax=b

=

11

41

yx

ba

a. Find a and b such that system has one solution, write the solution

b. Find a and b such that system has infinitely many solution, write the solution

Page 15: MATRICES, DETERMINANT AND SYSTEM OF LINEAR EQUATIONS · August 7, 2008 Matrices, Det(A) and System of Linear 2 Solving Linear System by Matrix Inversion If A is an invertible nxn

August 7, 2008 Matrices, Det(A) and System of Linear 15

Exercises

3. Find the least square straight line fit to the five points (0,0), (1,2), (3,4), (4,6) and (5,7)

4. Find the quadratic polynomial that best fit to the five points (0,0), (1,2), (3,4), (4,6) and (5,7)