The University of Manchester · FACULTY OF MATHEMATICS

MATH19661 Chap.6 Numerical Integration and Root Finding

- one subject, every graph, every model, every mark
7 Chapters4-page Bible
Our own words - no uploaded lecturer files
Updated for this semester
Chapter 6 of 11 · MATH19661

Numerical Integration and Root Finding

Numerical Integration and Root Finding deals with integrals and equations that have no exact answer in standard functions. The first half approximates a definite integral by sampling the integrand at equally spaced points.

The trapezium rule joins neighbouring points with straight lines, while Simpson's rule fits parabolas through consecutive triples, so it needs an even number of strips and uses the weight pattern one, four, two, four and so on. Error bounds in terms of the strip width show why Simpson usually wins: its error falls with the fourth power of the width rather than the square, though it needs the fourth derivative to estimate.

The second half solves equations of the form f(x) = 0. Bisection starts from a sign change and halves the bracket each cycle, which is slow but certain for any continuous function. Newton-Raphson follows tangent lines to the axis and usually converges much faster from a reasonable start, but it needs the derivative and can misbehave when the tangent is nearly flat.

Recent January papers open their second question with a four-strip Simpson calculation and follow it with a Newton-Raphson part, usually iterated to a stated accuracy.

In this chapter

What this chapter covers

  • 01

    Strip width and sampling points

  • 02

    Trapezium rule

  • 03

    Simpson's rule with an even number of strips

  • 04

    Error bounds for both rules

  • 05

    Bisection from a sign change

  • 06

    Newton-Raphson iteration from tangent lines

  • 07

    Stopping rules and decimal places

Worked example · free

Newton-Raphson tangent steps towards the root of x³ + x − 3

Q [5 marks]. Use the Newton-Raphson method with x0 = 1 to find the real root of x³ + x − 3 = 0 correct to three decimal places, stating when you stop. These marks are an AskSia practice allocation and are not drawn from a university or published marking scheme.
The orange curve is f(x) = x³ + x − 3.The orange curve is f(x) = x³ + x − 3. Starting from x0 = 2, the green first tangent meets the x-axis at 1.462 and the purple second tangent at 1.248, already close to the root 1.213.x0 = 2x1 = 1.462x2 = 1.248root 1.213first tangentsecond tangentcurve f(x)xy
The orange curve is f(x) = x³ + x − 3. Starting from x0 = 2, the green first tangent meets the x-axis at 1.462 and the purple second tangent at 1.248, already close to the root 1.213.
  • 1Define f(x) = x³ + x − 3 and differentiate: f′(x) = 3x² + 1.
  • 1Write the iteration: the next value is x minus (x³ + x − 3)/(3x² + 1).
  • 1From x0 = 1: f(1) = −1 and f′(1) = 4, so x1 = 1 + 0.25 = 1.25.
  • 1Next, f(1.25) = 0.203125 and f′(1.25) = 5.6875, so x2 = 1.214286; repeating gives x3 = 1.213412.
  • 1One more step gives x4 = 1.213412, which agrees with x3 to three decimal places, so the root is 1.213.
x = 1.213 to three decimal places, reached after x4 agrees with x3. Stopping at x2 = 1.214 would give the wrong third decimal.
Sia tip — List every iterate to at least six decimal places and stop only when two consecutive values agree to the accuracy asked for.
Glossary

Key terms

Strip width
The spacing h between sampling points, equal to the length of the interval divided by the number of strips.
Simpson's rule
A quadrature rule that fits a parabola across each successive pair of strips and weights the ordinates one, four, two, four and so on.
Error bound
An upper limit on the size of the error of a quadrature rule, written in terms of the strip width and a derivative maximum.
Bisection method
A root-finding method that repeatedly halves an interval on which the function changes sign.
Newton-Raphson iteration
A root-finding scheme that moves from each guess to the point where the tangent line meets the horizontal axis.
FAQ

Numerical Integration and Root Finding FAQ

Why must Simpson's rule use an even number of strips?

Each application of Simpson's rule covers two strips with a single parabola through three points. The last point of one pair is the first point of the next, so the strips have to pair off exactly. With an odd count one strip would be left over with no parabola to cover it.

Should my calculator be in degrees or radians?

Radians, whenever the integrand or equation contains a trigonometric function. Recent examiner feedback reports many scripts that computed the ordinates in degrees instead of radians, even though the rest of the method was right. Check the mode before the first value and again if you reset the calculator.

How many Newton-Raphson iterations should I show?

Keep going until two consecutive iterates agree to the number of decimal places the question asks for. Stopping as soon as a value looks right is a common way to lose the final mark; official solutions have required a third iterate even when the second was already close.

When is bisection better than Newton-Raphson?

Bisection never needs a derivative and cannot diverge once a sign change is bracketed, so it is safe when a good starting guess is unknown. It is slow, gaining one halving of the bracket per cycle. A sensible approach is to run a few bisection cycles and then switch to Newton-Raphson.

What does the error bound tell me in practice?

It tells you how small the strip width must be to guarantee a required accuracy. Because the trapezium bound depends on the square of the width and Simpson's on the fourth power, halving the width reduces them by factors of four and sixteen. The bound is usually larger than the actual error.

Study strategy

Exam move

Build a reusable table layout for quadrature with columns for i, the sampling point and the ordinate, and fill it for every Week 5 problem sheet integral before applying any weights. Do the four-strip Simpson calculation from several January papers under a four-minute limit, checking the calculator mode each time.

For root finding, practise rewriting equations into the form f(x) = 0 before differentiating, then run Newton-Raphson until two consecutive iterates agree to one more place than asked. Try one bisection example by hand to feel how slowly it converges, and one Newton-Raphson start near a turning point to see how badly a poor guess behaves. Finish each answer with a sentence stating the result to the requested accuracy.

Working through Numerical Integration and Root Finding in MATH19661? Sia is AskSia’s AI Mathematics tutor — ask any MATH19661 Numerical Integration and Root Finding question and get a clear, step-by-step explanation grounded in how MATH19661 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.

A+Everything unlocked
Unlocks this Bible + all 41 of your subjects at The University of Manchester, plus 1,000+ more Bibles across the AskSia library.
Sia - your MATH19661 tutor, unlimited, worked the way the exam marks it
The full 4-page Bible + practice bank with worked solutions
Chrome extension - sync your LMS so Sia knows your deadlines
Bilingual EN / Chinese on every Bible and every Sia answer
$0.99 Trial
30-day money-back · cancel in one tap · how it works