Lecture Note : Numerical Analysis (13) Newton-Cotes Integration Formula

1. Introduction to the Numerical Integration and Newton-Cotes Integration

 Numerical Integration
b b
S   f ( x )dx  I n   f n ( x ) dx where f n (x) is the approximating function
a a

 Exact Integration of Polynomials and Newton-Cotes Integration

𝑥 𝑙𝑛 𝑏 − 𝑙𝑛 𝑎 if 𝑘 = −1
𝑥 𝑑𝑥 = = 𝑏 𝑎
𝑘+1 − otherwise
𝑘+1 𝑘+1
For f ( x)  a0  a1 x  a2 x 2  a3 x 3   am 1 x m 1  am x m  a x j
j

I 
x2

x1
f ( x)dx  
x1
x2
 a x dx   a 
j
j
x2
j x
1
x j dx  
aj
j 1
x 2
j 1
 x1
j 1


The Newton-Cotes integration method approximates f (x) with interpolating polynomials to

b
numerically calculate the integration of I  a
f ( x) dx .

1
2. Divided-Difference Interpolation Formula revisited and Integration

 nth order polynomial interpolation function can be defined using the divided-difference formula as

f n ( x)  b0  b1 ( x  x0 )  b2 ( x  x0 )( x  x1 )    bn 1 ( x  x0 )( x  x1 )( x  xn 1 )
 bn ( x  x0 )( x  x1 )( x  xn 1 )( x  xn )

b0  f [ x0 ]
b1  f [ x1 , x0 ]
b2  f [ x2 , x1 , x0 ]

bn  2  f [ xn  2 , x2 , x1 , x0 ]
bn 1  f [ xn 1 , xn  2 , x2 , x1 , x0 ]
bn  f [ xn , xn 1 , xn  2 , x2 , x1 , x0 ]

In general, if we define the following divided difference formula,

0th order: f [ xi ]  f ( xi )
f [ xi ]  f [ x j ]
1st order: f [ xi , x j ] 
( xi  x j )

f [ xi , x j ]  f [ x j , xk ]
2nd order: f [ xi , x j , xk ] 
( xi  xk )
…….
f [ xn , xn1 , xn2 ,  , x1 ]  f [ xn1 , xn2 ,  , x1 , x0 ]
nth order: f [ xn , xn1 , xn1 ,  , x1 , x0 ] 
( xn  x0 )

 In case of equal spacing such as h  x j 1  x j , j  0,1, , n

- 1st order with two points x0 , x1

f n ( x)  b0  b1 ( x  x0 )
b0  f [ x0 ]  f ( x0 )
f [ x1 ]  f [ x0 ] f ( x1 )  f ( x0 )
b1  f [ x1 , x0 ]  
( x1  x0 ) h

2
I   f ( x)dx   f n ( x)dx   b0  b1 ( x  x0 )dx
x1 x1 x1

x0 x0 x0
x1
1
 b0 ( x  x0 )  b1 ( x  x0 ) 2
2 x0

1
 b0 ( x1  x0 )  b1 ( x1  x0 ) 2
2
1
 b0 h  b1h 2
2
1 f ( x1 )  f ( x0 ) 2
 f ( x0 )h  h
2 h
1
 f ( x0 )h   f ( x1 )  f ( x0 )h
2
h
  f ( x0 )  f ( x1 )
2

h
Trapezoidal Formula
I   f ( x0 )  f ( x1 )
2

3
- 2nd order with three points x0 , x1 , x2

f n ( x)  b0  b1 ( x  x0 )  b2 ( x  x0 )( x  x1 )
 b0  b1 ( x  x0 )  b2 ( x  x0 )( x  x0  x0  x1 )
 b0  b1 ( x  x0 )  b2 ( x  x0 ) 2  b2 ( x  x0 )( x1  x0 )

b0  f [ x0 ]  f ( x0 )
f [ x1 ]  f [ x0 ] f ( x1 )  f ( x0 )
b1  f [ x1 , x0 ]  
( x1  x0 ) h
f [ x2 , x1 ]  f [ x1 , x0 ] f ( x2 )  2 f ( x1 )  f ( x0 )
b2  f [ x2 , x1 , x0 ]  
( x2  x0 ) 2h 2
x2 x2
I   f ( x)dx   f n ( x)dx
x0 x0
x
1 1 1 2

 b0 ( x  x0 )  b1 ( x  x0 ) 2  b3 ( x  x0 )3  b2 ( x  x0 ) 2 ( x1  x0 )
2 3 2 x0

1 1 1
 b0 ( x2  x0 )  b1 ( x2  x0 ) 2  b2 ( x2  x0 )3  b2 ( x2  x0 ) 2 ( x1  x0 )  x2  x0  2h
2 3 2
8
 2b0 h  2b1h 2  b2 h3  2b2 h3
3
 f ( x1 )  f ( x0 )  2 2  f ( x2 )  2 f ( x1 )  f ( x0 )  3
 2 f ( x0 )h  2 h   h
 h  3 2h 2 
1
 2 f ( x0 )h  2 f ( x1 )  f ( x0 )h   f ( x2 )  2 f ( x1 )  f ( x0 )h
3
1
  f ( x2 )  4 f ( x1 )  f ( x0 )h
3
h
Simpson’s 1/3 Rule
I   f ( x2 )  4 f ( x1 )  f ( x0 )
3

- 3rd order with four points x0 , x1, x2 , x3

f n ( x)  b0  b1 ( x  x0 )  b2 ( x  x0 )( x  x1 )  b3 ( x  x0 )( x  x1 )( x  x2 )
 b0  b1 ( x  x0 )  b2 ( x  x0 )( x  x1 )  b3 ( x  x0 )( x  x1 )( x  x2 )
 b0  b1 ( x  x0 )  b2 ( x  x0 ) 2  b2 ( x  x0 )( x1  x0 )
 b3 ( x  x0 )( x  x0  x0  x1 )( x  x0  x0  x2 )
 b0  b1 ( x  x0 )  b2 ( x  x0 ) 2  b2 ( x  x0 )( x1  x0 )
 b3 ( x  x0 )( x  x0  h)( x  x0  2h)
 b0  b1 ( x  x0 )  b2 ( x  x0 ) 2  b2 ( x  x0 )( x1  x0 )  b3 ( x  x0 )3  3b3 ( x  x0 ) 2 h  2( x  x0 )h 2

4
b0  f [ x0 ]  f ( x0 )
f [ x1 ]  f [ x0 ] f ( x1 )  f ( x0 )
b1  f [ x1 , x0 ]  
( x1  x0 ) h
f [ x2 , x1 ]  f [ x1 , x0 ] f ( x2 )  2 f ( x1 )  f ( x0 )
b2  f [ x2 , x1 , x0 ]  
( x2  x0 ) 2h 2
f [ x3 , x2 , x1 ]  f [ x2 , x1 , x0 ] f ( x3 )  3 f ( x2 )  3 f ( x1 )  f ( x0 )
b3  f [ x3 , x2 , x1 , x0 ]  
( x3  x0 ) 6h 3
x3 x3
I   f ( x)dx   f n ( x)dx
x0 x0
x
1 1 1 3

b0 ( x  x0 )  b1 ( x  x0 )2  b2 ( x  x0 )( x  x1 )  b1 ( x  x0 )3  b2 ( x  x0 ) 2 ( x1  x0 )
2 3 2

1
 b3 ( x  x0 ) 4  b3 ( x  x0 )3 h  b3 ( x  x0 ) 2 h 2
4 x0

 1 1 1 
 b0 ( x3  x0 )  b1 ( x3  x0 )  b2 ( x3  x0 )  b2 ( x3  x0 ) ( x1  x0 ) 
2 3 2

2 3 2
   x  x  3h
 1  3 0

  b3 ( x3  x0 )  b3 ( x3  x0 ) h  b3 ( x3  x0 ) h
4 3 2 2

 4 
9 9 81
 3b0 h  b1h 2  9b2h3  b2 h3  b3h 4  27b3h 4  9b3h 4
2 2 4
9 2 9 3 9 4
 3b0 h  b1h  b2 h  b3h
2 2 4
9  f ( x1 )  f ( x0 )  2 9  f ( x2 )  2 f ( x1 )  f ( x0 )  3
 3 f ( x0 )h   h   h
2 h  2 2h 2 
9  f ( x3 )  3 f ( x2 )  3 f ( x1 )  f ( x0 )  4
  h
4 6h 3 
9 9
 3 f ( x0 )h   f ( x1 )  f ( x0 )h   f ( x2 )  2 f ( x1 )  f ( x0 )h
2 4
3
  f ( x3 )  3 f ( x2 )  3 f ( x1 )  f ( x0 ) h
8
3
  f ( x3 )  3 f ( x2 )  3 f ( x1 )  f ( x0 )h
8

3h
Simpson’s 3/8 Rule
I   f ( x3 )  3 f ( x2 )  3 f ( x1 )  f ( x0 )
8

5
3. Multiple Application of Newton-Cotes Integration Formula

h
Trapezoidal Formula
Ik   f ( xk )  f ( xk 1 ), x  [ xk , xk 1 ]
2
h
Simpson’s 1/3 Rule
Ik   f ( xk )  4 f ( xk 1 )  f ( xk  2 ), x  [ xk , xk  2 ]
3
3h
Simpson’s 3/8 Rule
I  f ( xk )  3 f ( xk 1 )  3 f ( xk  2 )  f ( xk 3 ), x  [ xk , xk 3 ]
8

(3-1) Trapezoidal

h
Ik   f ( xk )  f ( xk 1 ), x  [ xk , xk 1 ]
2
b
I   f ( x)dx
a
n 1
  Ik
k 0
n 1
h 
    f ( xk )  f ( xk 1 )
k 0  2 
h
  f ( x0 )  2 f ( x1 )  2 f ( x2 )    2 f ( xn 1 )  f ( xn )
2

6
(3-2) Simpson’s 1/3 Rule

h
I 2k   f ( x2k )  4 f ( x2k 1 )  f ( x2k  2 ), x  [ xk , xk  2 ], 2k  0,2,4,6,  , n  2
3
b
I   f ( x)dx
a
n / 2 1
 I
k 0
2k

n / 2 1
h 
   3  f ( x
k 0
2k )  4 f ( x2 k 1 )  f ( x2 k  2 )

h n / 2 1 n / 2 1

 
3
f ( x0 )  4 
k 0
f ( x2 k 1 )  2 
k 0
f ( x2 k  2 )  f ( xn )

h n 1 n 2

  f ( x0 )  4  f ( xn )  2  f ( xk )  f ( xn )
3 k 1, 3, 5 , k  2 , 4 , 6 , 

(3-3) Simpson’s 8/3 Rule

3h
I 3k   f ( x3k )  3 f ( x3k 1 )  3 f ( x3k  2 )  f ( x3k 3 ), x [ xk , xk 3 ], 3k  0,3,6,9,, n  3
8

b
I   f ( x)dx
a
n / 3 1
 I
k 0
3k

n / 2 1
 3h 
   8  f ( x
k 0
3k )  3 f ( x3k 1 )  3 f ( x3k  2 )  f ( x3k  3 )

3h  n / 3 1 n / 2 1

  f ( x0 )  3   f ( x3 k 1 )  f ( x3k  2 )   2  f ( x3k  3 )  f ( xn )
8  k 0 k 0 
h n2 n 3

  f ( x0 )  3   f ( xk )  f ( xk 1 )   2  f ( xk )  f ( xn )
3 k 1, 4 , 7 , k  3, 6 , 9 , 

7
4. Integration With unequal segments

 By using the divided-difference formula, we can drive the related formula

f n ( x)  b0  b1 ( x  x0 )  b2 ( x  x0 )( x  x1 )    bn 1 ( x  x0 )( x  x1 )( x  xn 1 )
 bn ( x  x0 )( x  x1 )( x  xn 1 )( x  xn )

b0  f [ x0 ]
b1  f [ x1 , x0 ]
b2  f [ x2 , x1 , x0 ]

bn  2  f [ xn  2 , x2 , x1 , x0 ]
bn 1  f [ xn 1 , xn  2 , x2 , x1 , x0 ]
bn  f [ xn , xn 1 , xn  2 , x2 , x1 , x0 ]