Integration Using the Gauss Quadrature Rule - Method

2004 Autar Kaw, Loubna Guennoun, University of South Florida, [email protected],

http://numericalmethods.eng.usf.edu

NOTE: This worksheet demonstrates the use of Maple to illustrate the Gauss Quadrature rule of
integration.

Section I: Introduction
Gauss Quadrature rule is another method of estimating an integral. The two point Gauss Quadrature
Rule is an extension of the Trapezoidal Rule approximation where the integral estimate was based
on taking the area under the straight line connecting the function values at the limits of the
integration interval, a and b. However, unlike the Trapezoidal Rule approximation, the two point
Gauss Quadrature rule is based on evaluating the area under a straight line connecting two points on
the curve that are not predetermined as a and b, but as unknowns x1 and x2 .Thus, in the two point
Gauss Quadrature Rule, the integral is approximated as
b
I = f ( x)dx
a

c1 f ( x1 ) + c2 f ( x2 )
There are now four unknowns that must be evaluated x1 , x2 , c1 and c2 . These are found by assuming
that the formula gives exact results for integrating a general third order polynomial,
f ( x) = a0 + a1 x + a2 x 2 + a3 x 3 .
b b
Hence, f ( x)dx = (a
a a
0 + a1 x + a2 x 2 + a3 x 3 )dx

b
x2 x3 x4
= a0 x + a1 + a2 + a3
2 3 4 a
b2 a2 b3 a 3 b4 a4
= a0 (b a ) + a1 + a2 + a3
2 3 4
The formula would then give
b

f ( x)dx = c (a + a1 x1 + a2 x1 + a3 x1 ) + c2 (a0 + a1 x2 + a2 x2 + a3 x2 )
2 3 2 3
1 0
a
Equating the last two equations gives
b2 a2 b3 a 3 b4 a4
a0 (b a) + a1 + a2 + a3
2 3 4
= c1 (a0 + a1 x1 + a2 x1 + a3 x1 ) + c2 (a0 + a1 x2 + a2 x2 + a3 x2 )
2 3 2 3

= a0 (c1 + c2 ) + a1 (c1 x1 + c2 x2 ) + a2 (c1 x1 + c2 x2 ) + a3 (c1 x1 + c2 x2 )

2 2 3 3

The constatnts a0 , a1 , a2 , and a3 are arbitrary

b a = c1 + c2
b2 a2
= c1 x1 + c2 x2
2
b3 a 3
= c1 x1 + c2 x2
2 2

3
b a4
4
= c1 x1 + c2 x2
3 3

4
Solving these equations simultaneously, we have

ba
c1 =
2
ba
c2 =
2
b a 1 b + a
x1 = +
2 3 2
b a 1 b + a
x2 = +
2 3 2

b
Hence, f ( x)dx c f ( x ) + c
a
1 1 2 f ( x2 )

b a b a 1 b + a b a b a 1 b + a
= f + + f +
2 2 3 2 2 2 3 2
This is called the two-point Gauss Quadrature Rule since two points were chosen. For n-points rules
formula, it can be developed using the general form

I c0 f ( x0 ) + c1 f ( x1 ) + ... + cn 1 f ( xn 1 )

NOTE: Weighting factors c and function arguments x used in Gauss Quadrature Rule have already
been defined for any function to be integrated from -1 to 1.
To change the limits of the integral so that they are from -1 to 1, a and b are substituted into
(b + a ) + (b a ) xd
x=
2
This equation can be differentiated to give

ba
dx = dxd
2

These two equations can be substituted for x and dx, respectively, in the equation to be integrated to
obtain the form that is suitable for evaluating the integral using Gauss Quadrature Rule.
The following spreadsheet summarize the weighting factors c and the function arguments x used in
Gauss quadrature Rule up to six points.

> restart;
 A B C
1 Points Weighting Factors Function Arguments
2 2 1.0000000 -0.577350269
3 1.0000000 0.577350269
4 3 0.5555556 -0.774596669
5 0.8888889 0.000000000
6 0.5555556 0.774596669
7 4 0.3478548 -0.861136312
8 0.6521452 -0.339981044
9 0.6521452 0.339981044
10 0.3478548 0.861136312
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
> C:=array(1..4,1..4):
The weighting factor for Gauss Quadrature Rule with one point is
C[1,1]:=2;
The weighting factors for Gauss Quadrature Rule with two points are
C[1,2]:=1;C[2,2]:=1;
The weighting factors for Gauss Quadrature Rule with three points are
C[1,3]:=0.555555556;C[2,3]:=0.888888889;C[3,3]:=0.555555556;
The weighting factors for Gauss Quadrature Rule with four point are
C[1,4]:=0.347854845;C[2,4]:=0.652145155;C[3,4]:=0.652145155;C[4
,4]:=0.347854845;
C1, 1 := 2
C1, 2 := 1
C2, 2 := 1
C1, 3 := 0.555555556
C2, 3 := 0.888888889
C3, 3 := 0.555555556
C1, 4 := 0.347854845
C2, 4 := 0.652145155
C3, 4 := 0.652145155
C4, 4 := 0.347854845
> X:=array(1..4,1..4):
The function argument for Gauss Quadrature Rule with one point is
X[1,1]:=0;
The function arguments for Gauss Quadrature Rule with two points are
X[1,2]:=-0.577350269;X[2,2]:=0.577350269;
The function arguments for Gauss Quadrature Rule with three points are
X[1,3]:=-0.774596669;X[2,3]:=0;X[3,3]:=0.774596669;
The function arguments for Gauss Quadrature Rule with four points are

X[1,4]:=-0.861136312;X[2,4]:=-0.339981044;X[3,4]:=0.339981044;X
[4,4]:=0.861136312;
X1, 1 := 0
X1, 2 := -0.577350269
X2, 2 := 0.577350269
X1, 3 := -0.774596669
X2, 3 := 0
 X3, 3 := 0.774596669
X1, 4 := -0.861136312
X2, 4 := -0.339981044
X3, 4 := 0.339981044
X4, 4 := 0.861136312

Section II: Data

The following simulation will illustrate the Gauss Quadrature Rule of integration. This section is the
only section where the user may interacts with the program.
The user may enter any function f(x), and the lower and upper limit for the function. By entering
these data, the program will calculate the exact value of the integral, followed by the results using
the Gauss Quadrature Rule with 2, 3, and 4 points. The program will also display the true error, the
absolute relative true % error, the approximate error, and the absolute relative approximate % error.

Function in f(x)=0:
> f:=x->300*x/(1+exp(x));
300 x
f := x
1 + ex
The lower limit of the integral (a):
> a:=0.0;
a := 0.
The upper limit of the integral (b):
> b:=10.0;
b := 10.0
This is the end of the user's section. All information must be entered before proceeding to the next
section.

Section III: The exact value of the integral

In this section, the program will evaluate the exact value for the integral of the function f(x)
evaluated at the limits a and b.
> s_exact:=int(f(x),x=a..b);
s_exact := 246.5902935

Section IV: The value of the integral using the Gauss

Quadrature Rule
Conversion of the limits

The integral given above has the limits of [a,b]. It needs to be converted into an integral with
limits [-1,1]
f_new(x) is the new function that will be used for evaluating the integral using the Gauss
Quadrature rule

> f_new:=x->f((b-a)/2*x+(b+a)/2)*(b-a)/2;
1 1 b a
f_new := x f ( b a ) x + + ( b a )
2 2 2 2

One Point
> s_1:=f_new(X[1,1]);
s_1 := 50.19638195
The approximate value of the integral using one-point Gauss quadrature rule is
> AV[1]:=C[1,1]*s_1;
AV1 := 100.3927639
The approximate error (E_a):
> E_a[1]:=undefined;
E_a1 := undefined
The absolute approximate percentage relative error (E_arel):
> E_arel[1]:=undefined;
E_arel1 := undefined

Two Points
> s_1:=f_new(X[1,2]);
s_1 := 341.7623132
> s_2:=f_new(X[2,2]);
s_2 := 4.442768559
The approximate value of the integral using two-points Gauss quadrature rule is
> AV[2]:=C[1,2]*s_1+C[2,2]*s_2;
AV2 := 346.2050818
The approximate error (E_a):
> E_a[2]:=AV[2]-AV[1];
E_a2 := 245.8123179
The absolute approximate percentage relative error (E_arel):
> E_arel[2]:=abs(E_a[2]/AV[2]*100);
E_arel2 := 71.00193811

Three Points
 > s_1:=f_new(X[1,3]);
s_1 := 413.6918962
> s_2:=f_new(X[2,3]);
s_2 := 50.19638195
> s_3:=f_new(X[3,3]);
s_3 := 1.864714594
The approximate value of the integral using three-points Gauss quadrature rule is
> AV[3]:=C[1,3]*s_1+C[2,3]*s_2+C[3,3]*s_3;
AV3 := 275.4837902
The approximate error (E_a):
> E_a[3]:=AV[3]-AV[2];
E_a3 := -70.7212916
The absolute approximate percentage relative error (E_arel):
> E_arel[3]:=abs(E_a[3]/AV[3]*100);
E_arel3 := 25.67167075

Four Points
> s_1:=f_new(X[1,4]);
s_1 := 346.8881973
> s_2:=f_new(X[2,4]);
s_2 := 176.0663493
> s_3:=f_new(X[3,4]);
s_3 := 12.35645101
> s_4:=f_new(X[4,4]);
s_4 := 1.268801886
The approximate value of the integral using four-points Gauss quadrature rule is
> AV[4]:=C[1,4]*s_1+C[2,4]*s_2+C[3,4]*s_3+C[4,4]*s_4;
AV4 := 243.9871154
The approximate error (E_a):
> E_a[4]:=AV[4]-AV[3];
E_a4 := -31.4966748
The absolute approximate percentage relative error (E_arel):
> E_arel[4]:=abs(E_a[4]/AV[4]*100);
E_arel4 := 12.90915496
>
>