Pure: Numerical Methods
Numerical methods are techniques for finding an approximate solution to an equation that cannot be solved exactly by algebra, using a sequence of closer and closer estimates. In A Level Maths this covers locating a root with a change of sign, refining it by iteration or the Newton-Raphson method, and estimating areas under curves with the trapezium rule.
Before you start
Make sure you're comfortable with these topics first:
Method
- To locate a root in an interval [a,b], evaluate f(a) and f(b); if f is continuous and the values have opposite signs, a root lies between a and b (the change of sign test).
- To use iteration, rearrange f(x) = 0 into the form x = g(x), then apply x_(n+1) = g(x_n) repeatedly, starting from the given x0, rounding each answer to the number of decimal places asked.
- To use the Newton-Raphson method, differentiate f(x) to find f'(x), then apply x_(n+1) = x_n - f(x_n)/f'(x_n) repeatedly from the given starting value.
- To confirm a root correct to a stated number of decimal places, evaluate f at the value rounded up and rounded down (e.g. 1.3275 and 1.3285 for 1.328 to 3dp) and check for a change of sign.
- To use the trapezium rule, split the interval into equal strips of width h, list the y-values at each x-value, and apply (h/2)[y0 + yn + 2(sum of the middle y-values)].
- To decide whether the trapezium rule over- or under-estimates the true area, consider whether the curve is convex or concave on that interval: convex curves give an overestimate, concave curves give an underestimate.
Common mistakes
MistakeClaiming a root from a change of sign without stating that the function is continuous on the interval.
InsteadState continuity as well as opposite endpoint signs. Also remember that an even-multiplicity root may not produce a sign change.
MistakeRounding each iterate heavily before using it to calculate the next value.
InsteadStore full calculator precision between iterations and round only the displayed iterates or the final answer as requested.
MistakeDoubling every ordinate in the trapezium rule, including the two endpoint ordinates.
InsteadUse each endpoint once and every interior ordinate twice: (h/2)[y_0 + y_n + 2(sum of interior ordinates)].
Worked examples
Example 1
Show that f(x) = x^3 - 4x + 1 = 0 has a root between x = 0 and x = 1. Using the iteration x_(n+1) = (x_n^3 + 1)/4 with x0 = 0, find x1, x2 and x3, each to 4 decimal places, and hence show the root is x = 0.254 correct to 3 decimal places.
- Evaluate f(0) = 1 and f(1) = 1 - 4 + 1 = -2; opposite signs, so a root lies in (0,1).
- Rearrange 4x = x^3 + 1 to get x = (x^3 + 1)/4, so x_(n+1) = (x_n^3 + 1)/4.
- Substitute x0 = 0: x1 = (0^3 + 1)/4 = 0.2500.
- Substitute x1 = 0.2500: x2 = (0.2500^3 + 1)/4 = 0.2539 (4dp).
- Substitute x2 = 0.2539: x3 = (0.2539^3 + 1)/4 = 0.2541 (4dp).
- Check f(0.2535) = 0.00229 (>0) and f(0.2545) = -0.00152 (<0); change of sign confirms the root is x = 0.254 correct to 3dp.
Example 2
Let f(x) = e^(-x) - x. Starting with x_0 = 0.6, use two Newton-Raphson iterations to estimate the root. Confirm the root correct to 3 decimal places.
- Differentiate: f'(x) = -e^(-x) - 1, so x_(n+1) = x_n - (e^(-x_n) - x_n)/(-e^(-x_n) - 1).
- Using x_0 = 0.6 gives x_1 = 0.566949910..., so x_1 = 0.566950 to 6 decimal places.
- Using the unrounded x_1 gives x_2 = 0.567143284..., so x_2 = 0.567143 to 6 decimal places.
- For 3 decimal places, test the rounding boundaries: f(0.5665) = 0.001008... > 0 and f(0.5675) = -0.000559... < 0.
- Since f is continuous and changes sign between these boundaries, the root rounds to 0.567.
- Final answer: the Newton-Raphson estimate is x = 0.567143 to 6 decimal places, so the root is x = 0.567 correct to 3 decimal places.
Example 3
A continuous non-negative curve has the following table of ordinates: x = -1, -0.5, 0, 0.5, 1, 1.5; y = 2, 3, 0, 4, 5, 2 respectively. Use every ordinate to estimate the area under the curve over this interval by the trapezium rule. Count the strips and explain the interior weights by writing the sum of the areas of adjacent trapezia. Can this table alone tell you whether the estimate is too large or too small?
- There are six ordinates but five strips, one between each adjacent pair of x-values. The common strip width is h = (1.5 - (-1))/5 = 0.5.
- Each strip contributes h/2 times the sum of its two endpoint heights. Adding the five areas gives (0.5/2)[(2 + 3) + (3 + 0) + (0 + 4) + (4 + 5) + (5 + 2)].
- Each interior ordinate belongs to two neighbouring trapezia. Collect the terms as (0.5/2)[2 + 2 + 2(3 + 0 + 4 + 5)]: list each interior ordinate once inside the sum, with coefficient two outside. Each endpoint has coefficient one.
- The interior sum is 3 + 0 + 4 + 5 = 12. The bracket is 2 + 2 + 2(12) = 28, so the estimate is 0.25 x 28 = 7 square units. The zero interior ordinate still has coefficient two.
- The table gives heights only at the listed x-values. It does not describe how the curve bends between them, so it cannot alone establish whether the trapezia lie above or below the curve.
- Final answer: the trapezium-rule area estimate is 7 square units, using five strips of width 0.5. Whether this is an overestimate or an underestimate cannot be decided from this table alone.
Practice questions
Q1[1 mark]State the condition on f(x) required for the change of sign test to guarantee a root of f(x) = 0 lies in the interval [a,b].Show answer
Answer: f must be continuous on [a,b] (and f(a), f(b) must have opposite signs).
Q2[2 marks]Show that f(x) = x^3 + x - 3 has a root between x = 1 and x = 2.Show answer
Answer: f(1) = -1, f(2) = 7; opposite signs, so a root lies in (1,2).
Q3[1 mark]The equation x^3 + x - 3 = 0 can be rearranged to x = (3 - x)^(1/3). Using x0 = 1.5, find x1, giving your answer to 4 decimal places.Show answer
Answer: x1 = 1.1447 (substitute x0 = 1.5 into (3 - x)^(1/3)).
Q4[3 marks]The curve y = x^2 + 1 is defined for 0 <= x <= 3, with y = 1, 2, 5, 10 at x = 0, 1, 2, 3. Use the trapezium rule with all 3 strips to estimate the integral of x^2 + 1 from 0 to 3.Show answer
Answer: 12.5 (h = 1: (1/2)[1 + 10 + 2(2+5)] = 12.5).
Q5[3 marks]Let f(x) = x^2 - 7. Using the Newton-Raphson method with x0 = 2, find x1.Show answer
Answer: x1 = 2.75 (x1 = 2 - f(2)/f'(2) = 2 - (-3)/4).
Q6[3 marks]Let f(x) = x^3 - 2x - 5. Using the Newton-Raphson method with x0 = 2, find x1 and x2, each to 4 decimal places.Show answer
Answer: x1 = 2.1000, x2 = 2.0946 (using f'(x) = 3x^2 - 2).
Exam-style questions
Written in the style of a A Level Maths exam paper, with a full mark scheme.
Exam tips
- Write the iteration formula, the starting value and a labelled sequence of iterates so each substitution can be followed.
- To confirm a root to a stated decimal place, test the two rounding boundaries and state the sign change and continuity.
- For the trapezium rule, list equally spaced x-values and check the strip width before substituting the ordinates.
Q1[3 marks]The curve y = 1/(x+1) is defined for 0 <= x <= 4. The table gives values of y at x = 0, 1, 2, 3 and 4: y = 1, 0.5, 0.3333, 0.25, 0.2. Use the trapezium rule with all 4 strips to find an estimate for the integral of 1/(x+1) from 0 to 4, giving your answer to 4 decimal places.Show mark scheme
Tick each line you got.
Nothing ticked yet - 3 available
Q2[5 marks]Let f(x) = x^3 - 6x + 2. (a) Show that f(x) = 0 has a root between x = 0 and x = 1. (2) (b) Using the iteration x_(n+1) = (x_n^3 + 2)/6 with x0 = 0.5, find x1 and x2, each to 4 decimal places. (3)Show mark scheme
Tick each line you got.
Nothing ticked yet - 5 available
Q3[4 marks]The equation x^2 - 10 = 0 has a positive root, r. The iteration x_(n+1) = (x_n + 10/x_n)/2 is used to estimate r, starting from x0 = 3. (a) Find x1 and x2, each to 4 decimal places. (3) (b) State what happens to the sequence x0, x1, x2, ... as n increases. (1)Show mark scheme
Tick each line you got.
Nothing ticked yet - 4 available
This topic in real past papers
We have tagged some boards' past papers question by question (not every board or year yet). This topic comes up in:
- OCR A Level Maths past papers: 14 questions across 14 papers, latest tagged June 2024
- AQA A Level Maths past papers: 6 questions across 4 papers, latest tagged November 2021
- WJEC A Level Maths past papers: 2 questions across 2 papers, latest tagged Summer 2024
Free printable worksheet
Want more practice on paper? Download the Pure: Numerical Methods worksheet pack - 9 pages of exam-style questions with a full mark scheme. Free to download, no card or account. Print it for personal and classroom use.
This topic is chapter 9 of A Level Maths Pure Essentials, the whole course as one free printable PDF. Its worksheet is also chapter 9 of A Level Maths Pure Topic Workbook 1 and chapter 9 of A Level Maths 10-Minute Tests 1, free printable practice books.
Other versions of this worksheet: