Scribd
Upload
89 views7 pages

Mid Solution

The document contains examples and explanations related to numerical analysis topics such as root finding, fixed point iteration, Newton's method, LU decomposition, spline interpolation, and matrix conditioning. Several questions are asked and answered regarding properties of functions, convergence of iterative methods, and characteristics of matrices and approximations.

Uploaded by

sibie
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
89 views7 pages

Mid Solution

The document contains examples and explanations related to numerical analysis topics such as root finding, fixed point iteration, Newton's method, LU decomposition, spline interpolation, and matrix conditioning. Several questions are asked and answered regarding properties of functions, convergence of iterative methods, and characteristics of matrices and approximations.

Uploaded by

sibie
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

1.

The quadratic f ( x ) = ( x 0.3)( x 0.5) obviously has zeros at 0.3 and 0.5.
(a) Why is the interval [0.1, 0.6] not a satisfactory starting interval for
bisection?
Ans :
f(0.1) = 0.08 and f(0.6) = 0.03 f(0.1)*f(0.6) > 0

(b) If you start with [0, 0.32], which root is reached with bisection?
Left

Right

Result

0
0.16

0.32
0.32

0.24
0.28

0.32
0.32

f(0.16) * f(0.32) < 0


f(0.24) * f(0.32) < 0
f(0.28) * f(0.32) < 0
f(0.30) * f(0.32) = 0

New

Iteration

0.16
0.24

1
2

0.28
0.30

3
4

Reach root 3.
(c) Using the interval in (b), how many iterations will be required to attain an
accuracy of 0.01 ?
Ans :
Version 1 :
Left

Right

Result

0
0.16
0.24

0.32
0.32
0.32

0.28

0.32

f(0.16) * f(0.32) < 0


f(0.24) * f(0.32) < 0
f(0.28) * f(0.32) < 0
f(0.30) * f(0.32) = 0

Version 2 :
0.32 0.0
2n

0.01 n = 5

New

Iteration

0.16
0.24
0.28

1
2
3

0.30

2.

The iteration scheme of Newtons method for root finding is


f ( xk )
x k +1 = x k
f ( x k )
(a) Please describe under what kind of conditions the algorithm does not
converge.
Ans :
1.
f ( x ) = 0
2.
f(x k + i ) = f(x k ) for some i > 0 (loop , see slide chap 1 page 19)
(b) Please show that Newtons method is also an example of the fixed point
method . (express Newtons method in the iteration scheme of the
fixed-point method)
Ans :
f ( xk )
x k +1 = g ( x k ) = x k
f ( x k )

3.

For the equation

f (x ) = x 2 x 2 = 0

Each of the following functions yields an equivalent fixed-point problem:


g1 ( x ) = x 2 2 ,
g 2 (x ) =

x+2 ,

g 3 ( x) = 1 +

2
,
x

x2 + 2
,
2x 1
Analyze the convergence properties of each of the corresponding fixed-point
g 4 ( x) =

iteration schemes for the root x = 2 by considering g i (2) .


Ans :
Converge if g i ( x) < 1 .

g1 ( x) = 2 x
g 2 ( x) =

1
2 x+2

g 3 ( x ) =

g 4 ( x) =

g1 (2) = 4 > 1

1
x2

2x 2 2x 4

(2 x 1)2

diverge.

g 2 (2) =

1
<1
4

converge.

g 3 ( 2) =

1
<1
4

converge.

g 4 (2) = 0 < 1

converge.

4.

Carry out one iteration of Newtons method applied to the system of nonlinear
equations with starting point x0 = [x0

y 0 ] = [1 2]
T

x + 2y 2
f ( x, y ) = 2

2
x + 4 y 4
Ans :
1 2
J ( x, y ) =

2 x 8 y
1

1 1 2
x1 = x0 + J ( x0 )( f ( x 0 )) = +

2 2 16
11
1 1 16 2 3 1 6
= +

= +
2 12 2 1 13 2 7
12
1

5.

3
13

5
6
=
17

12

A = LU
ATx = b
(LU) Tx = b
UTLTx = b
Q UT is a lower-triangular matrixand LT is a upper- triangular matrix
Let L1 = UTU1= LT
L1 U1x = b
First we get y = U1x from L1y = b by forward-substitution
then we get x from y = U1x by backward substitution.

6.

1 1 0
(a) det( 1 2 1 ) = 0
0 1 1
singular

1 0
1 0
1 1 0
1
1

1
1
(b) 1 2 1 (1) 1 1 (1)
0 1 1
(0) 1 1
(0) (1) 0
0 0
1

L = 1 1 0
0 1 1

1 1 0
U = 0 1 1
0 0 0

7.

(a) coefficient n*n matrix is diagonally dominant


n

aii > aij


j =1
j i

i = 1,2,,n

6 2 1
2
(b) Coefficient matrix is 2 7
1
2 5
8.

(a) From the condition we can find that

1
0

0 Qi Pi
1 Qi +1 Pi +1
=
0 Qi + 2 Pi + 2

3 Qi +3 Pi +3

0 0
0 0
3 0
0 3

replace it back

p(u) = u

= u

u2

2
3
1
0

u2

u1

u1

1
3
1
3

3
6
3
0

2
3
0
0
3
3
0
0

1
2
1
0

1 1
1 0
0 3

0 0

0
0
3
0

0
0
0
3

0 Qi
1 Q i +1
0 Q i + 2

3 Q i +3

1 Qi
0 Q i +1
0 Q i + 2

0 Qi+3

(b) No. Qi + 2 Qi +3 Qi + 4 need be collinear.

(c)

P0~3 = u 3

P3~ 6 = u 3

u2

u2

u1

1 3 3
3 6 3
1
3 3
0

0
0
1

1 10 10
0 50 15
= u3
0 75 60

0 90 100

u1

1 3 3
3 6 3
1
3 3
0

0
0
1

1 90
0 105
0 150

0 180

100
140
= u3

200

140

u2

u2

u1

u1

45
5
45 120

1
120 15

10
10

45 140
90
60
1
45
120

100
90

X = 5u 3 45u 2 + 120u + 10
P0~3 0~3
3
2
Y0~3 = 45u + 120u + 15u + 10
X = 45u 3 + 90u 2 + 45u + 90
P3~ 6 3~ 6
u [0 , 1]
3
2
Y
u
u
u
=

140
+
60
+
120
+
100
3~ 6

9.

X 3 X 5 a 0 + a1 x + a 2 x 2 + a3 x 3
10. X
+
=
3
5
1 + b1 x + b2 x 2 + b3 x 3
x 0 : a0 = 0
x1 : a1 = 1
x 2 : a 2 = b1
1
+ b2
x 3 : a3 =
3
1
x 4 : 0 = b3 b1
3
1 b
x5 : 0 = 2
5 3
b b
x6 : 0 = 1 3
5 3

11. (a) False.

3
5
b1 = b3 = 0
b2 =

a3 =

1 3 4
+ =
3 5 15

4 3
x
15
Pad approximation:
3
1+ x2
5
x+

0 0
2
0 1 0

2
det
1
0 0
2

0
...
...

0 ... ...

... 0

... 0

1
= ( ) n if n det 0

... 0
2

... 0
1
...

2 n*n

But it is nonsingular
(b) False.
(textbook p117)A small number means good conditioninga large one means
ill-conditioning(nearly singular).
(c) True.
Condition number = A A 1 .
cond(A) =

A A 1

A 1 A

= cond( A 1 )

(d) False.
(textbook p184) B-spline need not pass through any points of the set that are
used in its definition.
(e) True.
(textbook p170).cubic spline meets these conditions :

g i( xi +1 ) = g i +1( xi +1 ) C2 continuous.
(d) True

(e) False.

Almost linear parallel Small residual but big error of solution.

You might also like