Numerical Differentiation and

Integration/Integration of Equations/
Romberg Integration

Compline by:
Habtamu G.
 Romberg Integration
- It is used when the function is known and high accuracy is required,
so it is integration to improve the accuracy of the trapezoidal rule.

- Romberg integration combines the trapezoidal rule with Richardson

extrapolation.
Richardson’s Extrapolation
Methods use two estimates of an integral to compute a third, more
accurate approximation which can be represented as:

I = I(h) + E(h) b−a

h=
n
I: the exact value of the integral
I(h): the approximation from n-segment application of the trapezoidal rule
E(h): the truncation error
h: step size
Romberg Integration is a method for combining
two numerical integral estimates to obtain
a third, more accurate value.
I(h )
f (x)
1 I(h 2 )
f (x)

h2 I combined
h1 h2

a b a b

I(h1 ) I(h 2 )
f (x) f (x)
I combined
h1 h1 h2 h2
h2 h 2

a b
b
 f (x) I(h1 ) E(h1 ) f (x) I(h 2 ) E(h 2 )

h1 h2 h2
a b a b

I 1 = I(h1 ) + E(h1 ) I 2 = I(h 2 ) + E(h 2 )

I1 = I 2

Assumed that 𝒇ത ′′is constant regardless of step size, and can

be used to determine that the ratio of the two errors will be
 E(h1 ) h1 2 h1 2
( ) E(h1 )  E(h 2 )( )
E(h2 ) h2 h2
From (1) I(h1 ) − I(h 2 )
E(h 2 ) 
h1 2 1 − ( h 1 )2
I(h1 ) + E(h 2 )( )  I(h 2 ) + E(h 2 )
h2 h2
I  I(h 2 ) + E(h 2 )
I(h1 ) − I(h 2 ) I  I(h 2 ) +
1
[I(h 2 ) − I(h1 )]
I  I(h 2 ) +
1−( )
h1 2 ( h1 ) 2 − 1
h2 h2

For special case where the interval is halved: h = h1

2
2
4 1
I  I(h 2 ) − I(h1 ) We have combined two trapezoidal rules[
3 3 I(h1) of order O(h2) and I(h2) of order O(h2)] to
yield a new estimate of I of order O(h4)
 Example: Integrate the following function from 0 to 0.8
using combined (composite) (Romberg) of trapezoidal rule
a. For one and two segments
b. For two and four segments

f (x) = 0.2 + 25x − 200x 2 + 675x 3 − 900x 4 + 400x 5

Solution
Application of the trapezoidal rule:

segments h Integral
n−1
1 0.8 0.1728 f (x 0 ) + 2 f (x i ) + f (x n )
I = (b − a) i=1
2 0.4 1.0688 2n
4 0.2 1.4848
a)For one and two segments
4 1 4 1
I I(h 2 ) − I(h 1 ) I (1.0688) − (0.1728) = 1.367467
3 3 3 3

I exact = 1.640533  t = 16.6%

(b) For two and four segments

4 1
I (1.4848) − (1.0688) = 1.623467
3 3

I exact = 1.640533
t = 1%
 b
Exercise: The integral of  f (x)dx was evaluated twice
a

using the composite trapezoidal rule. The first with step size of 0.2
gave a value of 1.8 and the second with a step size of 0.4 and gave a
value of 0.95. What will be the value of the integral with O(h4)?

Ans:
I  2.08333
Numerical Differentiation and
Integration/
Gaussian Integration
 Gaussian Integration
Gauss Quadrature
Based on trapezoidal rule curve must pass through
the end points fig (a) results in a large error.

If the fixed points were free to evaluate the area under a straight
line would balance the positive and negative errors as fig. b,
improves estimate of the integral.
Gauss quadrature is the name for one class of the techniques,
particular described in this section are called Gauss- Legendre
formulas.
Before describing the approach, we will show how numerical integration formulas such
as the trapezoidal rule can be derived using the method of undetermined coefficients.
This method will then be employed to develop the Gauss-Legendre formulas.
Derived the trapezoidal rule by integrating a linear
interpolating polynomial and by geometrical reasoning.
To illustrate the approach, the trapezoid rule can be
expressed as:
(*)

where the c’s = constants.

Trapezoidal rule yield exact results when the function
being integrated is a constant or a straight line. Two simple
equations that represent these cases are y = 1 and y = x.
Thus, the following equalities should hold:
or, evaluating the integrals,

These are two equations with two unknowns that can be

solved for

which, when substituted back into Eq. (*), gives

which is equivalent to the trapezoidal rule.

 Two-Point Gauss-Legendre Formula
Like the trapezoidal rule, the objective of Gauss quadrature
is to determine the coefficients of an equation of the form

(**)

Unlike the trapezoidal rule that used fixed end points a and b, the
function arguments x0 and x1 are not fixed at the end points, but are
unknowns.

Fig. Unknown variables x0 and

x1 for integration by Gauss
quadrature
 Two-Point Gauss-Legendre Formula...
Thus, a total of four unknowns that must be evaluated.
a. Two of these conditions by assuming that Eq. (**) fits the integral
of a constant and a linear function exactly.
b. Merely extend this reasoning by assuming that it also fits the
integral of a parabolic (y = x2) and a cubic (y = x3) function.
Derive a linear two-point integration formula that is exact results for
integrating a general third order polynomial,

solved simultaneously for

Yield the two-point Gauss-Legendre formula

(***)

The functional values at x = 1/ 3 and −1/ 3 yields an integral estimate that is third-
order accurate.
The integration limits from −1 to 1 was done to simplify the mathematics.
A simple change of variable can be used to translate other limits of integration into this
form. This is accomplished by assuming that a new variable xd is related to the original
variable x in a linear fashion, as in

(4*)
If the lower limit, x = a, corresponds to xd =−1, these values can be substituted into Eq.
(4*) to yield
(5*)

Similarly, the upper limit, x = b, corresponds to xd = 1, to give

(6*)
Solve simultaneously for

which can be substituted into Eq. (4*) to yield

(7*)

This equation can be differentiated to give

(8*)

Equations (7*) and (8*) can be substituted for x and dx, respectively, in the
equation to be integrated. These substitutions effectively transform the
integration interval without changing the value of the integral. The following
example illustrates how this is done in practice.
 Example : Use Eq. of Gauss-Legendre to evaluate the integral of

between the limits x = 0 to 0.8. The exact value of the integral is

1.640533.
Solution: Before integrating the function, we must perform a change of variable
so that the limits are from −1 to +1. To do this, we substitute a = 0 and b = 0.8
into Eq. to yield

The derivative of this relationship is

Both of these can be substituted into the original equation to yield

The right-hand side is in the form that is suitable for evaluation using Gauss
quadrature. The transformed function can be evaluated at −1/ 𝟑 to be equal to
0.516741 and at 1/ 𝟑 to be equal to 1.305837.

which represents a percent relative error of −11.1 percent.