3 Sequences, functions and graphs
- Syllabus
- 2017
- Section
- 3
- Level
- Higher

A sequence is an ordered list. A term-to-term rule tells how to get from one term to the next; a position-to-term rule gives any term directly from its position n.
| Rule type | Example | How to generate |
|---|---|---|
| term-to-term | start at 3, then multiply by 2 | 3,6,12,24,… |
| position-to-term | un=n2 | 1,4,9,16,… |
| named pattern | odd numbers | 1,3,5,7,… |
| named pattern | powers of 2 | 1,2,4,8,… |
Identify the starting position, apply the stated rule exactly, and label the terms. For a position rule, substitute n=1,2,3,… rather than repeatedly changing the previous value.
Check that every generated term obeys the same definition. A list may look familiar but still follow a different rule after its displayed terms.
A finite list alone does not determine one unique sequence. Use the stated rule or context; do not invent a pattern only because it fits the first few terms.
To continue a sequence, compare consecutive terms and find an operation that is applied consistently. State both the operation and any starting information needed.
| Consecutive comparison | Likely rule |
|---|---|
| constant difference | add or subtract that value |
| constant ratio | multiply or divide by that value |
| alternating changes | repeat the change cycle |
| changing first differences | inspect second differences or another structure |
For 5,9,13,17,…, each difference is 4, so the next terms are 21,25 and the term-to-term rule is ‘add 4’.
For 1,2,4,8,…, each term is twice the previous term, so the next terms are 16,32 and the rule is ‘multiply by 2’.
Do not describe 1,2,4,8 as ‘add 1, add 2, add 4’ if the intended invariant is multiply by 2. Prefer the simplest consistent rule supported by the task.
An arithmetic sequence has a constant difference d. Its nth term is linear: un=dn+c, where c is chosen so that the formula reproduces the first term.
| Step | Action |
|---|---|
| difference | calculate d from consecutive terms |
| first draft | write dn |
| adjustment | compare d with the first term to find c=a−d |
| verify | substitute n=1,2,3 |
| use | substitute a position, or solve the formula for n |
For 7,11,15,19,…, d=4. The sequence 4n begins 4,8,12,16, so add 3: un=4n+3.
To test whether 83 is a term, solve 4n+3=83, giving n=20. Because 20 is a positive integer, 83 is a term.
The common difference is the coefficient of n, not usually the complete nth term. Always include the adjustment and verify at least the first term.
Every arithmetic term satisfies un=a+(n−1)d, where a is the first term and d is the common difference. Two independent term facts can determine both unknowns.
| Given fact | Equation |
|---|---|
| second term is 7 | a+d=7 |
| fifth term is 19 | a+4d=19 |
| terms up and uq known | subtract to get (q−p)d=uq−up |
If u2=7 and u5=19, subtract a+d=7 from a+4d=19: 3d=12, so d=4 and then a=3.
Rebuild the stated terms: u2=3+4=7 and u5=3+16=19. This catches an off-by-one error in the multiplier of d.
The nth term contains (n−1)d, not nd, because the first term is reached before any common-difference steps have been taken.
The formula un=a+(n−1)d gives any term of an arithmetic sequence directly from its first term a, common difference d, and position n.
| Goal | Set up |
|---|---|
| find a term | substitute its position for n |
| find a position | set a+(n−1)d equal to the given value |
| compare terms | use uq−up=(q−p)d |
| write linear form | expand to un=dn+(a−d) |
For a=4 and d=3, u50=4+49(3)=151. The multiplier is 49 because there are 49 steps from the first term to the 50th.
If 4+(n−1)3=100, then 3n+1=100 and n=33, so 100 is the 33rd term. A non-positive or non-integer result would mean it is not a term.
This formula applies to arithmetic sequences only. A constant ratio or changing difference requires a different model.
An arithmetic series is the sum of the terms of an arithmetic sequence. Pairing first and last terms gives Sn=2n(a+l)=2n[2a+(n−1)d].
| Information available | Efficient form |
|---|---|
| first term a and last term l | Sn=2n(a+l) |
| first term a and difference d | Sn=2n[2a+(n−1)d] |
| a range of terms | subtract cumulative sums or re-index carefully |
For 4+7+10+13+⋯ with 50 terms, a=4, d=3, so S50=250[2(4)+49(3)]=25(155)=3875.
The last term is l=a+(n−1)d. Writing the series forwards and backwards makes each pair total a+l; there are n such pairs across two copies of the series.
Do not substitute the last term as n. First determine how many terms are being summed, especially when the sum starts after the first term.
A function maps each allowed input to exactly one output. Different inputs may share an output, but one input cannot be assigned two different outputs.
| Idea | Meaning |
|---|---|
| input | an allowed element of the domain |
| output | the value assigned by the function |
| mapping | the rule connecting each input to its output |
| image | the output produced by a particular input |
A mapping diagram represents a function if every input has exactly one outgoing arrow. An output may receive several arrows, and some listed outputs may receive none.
The rule x↦x2 maps both −3 and 3 to 9; this is still a function because each input has only one output.
‘Exactly one output per input’ does not mean ‘exactly one input per output’. The latter extra condition is needed for an inverse to be a function on the full range.
The notation f(x) names the output of function f when the input is x. The mapping form f:x↦3x−2 and the equation f(x)=3x−2 describe the same rule.
| Request | Action |
|---|---|
| find f(5) | substitute x=5 into the rule |
| find f(a+1) | replace every x by (a+1) |
| solve f(x)=k | set the rule equal to k and solve |
| write mapping form | state f:x↦ followed by the rule |
If f(x)=3x−2, then f(5)=3(5)−2=13 and f(a+1)=3(a+1)−2=3a+1.
Functions named f and g can use different rules. Keep the function name attached to the correct definition throughout a multi-part problem.
f(x) is not f multiplied by x. It is a single notation for the output produced by input x.
The domain is the set of allowed inputs; the range is the set of outputs actually produced. Algebraic restrictions and stated intervals can limit either set.
| Feature | Domain control |
|---|---|
| denominator | exclude values making it zero |
| even root | require the radicand to be non-negative |
| stated interval | keep only inputs inside it |
| inverse function | its domain is the original range |
For f(x)=1/(x−2), x=2 makes the denominator zero, so the domain excludes 2. Also f(x) can never equal 0, so the range excludes 0.
On a graph, project the curve onto the x-axis for the domain and onto the y-axis for the range, respecting open endpoints, holes and asymptotes.
A value excluded from the domain is an input restriction; it is not automatically excluded from the range. Analyse inputs and outputs separately.
A composite applies functions in sequence: fg(x)=f(g(x)), so g acts first. An inverse function reverses the original mapping and satisfies f−1(f(x))=x on the permitted domain.
| Task | Method |
|---|---|
| find fg(x) | substitute the whole expression g(x) into f |
| evaluate fg(a) | find g(a), then apply f |
| find f−1 | write y=f(x), solve for x, then swap labels |
| verify inverse | simplify both f−1(f(x)) and f(f−1(x)) |
If f(x)=2x+1 and g(x)=x2, then fg(x)=2x2+1, while gf(x)=(2x+1)2. Order matters.
For f(x)=3x−5, set y=3x−5 and solve x=(y+5)/3, so f−1(x)=(x+5)/3.
A many-to-one rule such as x2 needs a restricted domain, for example x≥0, before its inverse x is a function.
The superscript −1 denotes an inverse, not a reciprocal: generally f−1(x)=1/f(x).
A graph tells a story about how one quantity changes with another. Read the axis labels, units and scale before interpreting its shape.
| Feature | Distance–time meaning | Speed–time meaning |
|---|---|---|
| horizontal segment | stopped | constant speed |
| steeper segment | faster travel | faster acceleration or deceleration |
| rising segment | moving away | speed increasing |
| falling segment | returning | speed decreasing |
To answer a question, locate the given value on one axis, move to the graph, then read the corresponding value from the other axis. For an interval, compare its two endpoints.
A negative gradient on a distance-from-home graph means returning towards home; distance travelled itself has not become negative.
Rectangular Cartesian coordinates use two perpendicular number lines: the horizontal x-axis and vertical y-axis. Their intersection is the origin (0,0).
| Convention | Meaning |
|---|---|
| (x,y) | horizontal coordinate first, vertical coordinate second |
| positive x | right of the origin |
| negative x | left of the origin |
| positive y | above the origin |
| negative y | below the origin |
Check the scale on each axis separately: one square need not represent one unit, and the two axes may use different scales.
(3,−2) and (−2,3) are different points. Never reverse the coordinate order to match the direction you move.
To plot (x,y), start at the origin, move horizontally to x, then vertically to y. To read a point, project it to the x-axis first and the y-axis second.
| Quadrant | Sign of x | Sign of y |
|---|---|---|
| I | + | + |
| II | − | + |
| III | − | − |
| IV | + | − |
A point on the x-axis has y=0; a point on the y-axis has x=0. Such points are not in any quadrant.
After plotting, read the point back from the axes. This catches swapped coordinates and incorrect signs.
Geometrical facts can fix a point's horizontal coordinate, vertical coordinate or both. Translate each fact into a coordinate constraint before calculating.
| Geometrical fact | Coordinate consequence |
|---|---|
| vertical alignment | same x-coordinate |
| horizontal alignment | same y-coordinate |
| reflection in x-axis | (x,y)→(x,−y) |
| reflection in y-axis | (x,y)→(−x,y) |
| translation by (ba) | (x,y)→(x+a,y+b) |
Mark known coordinates, use shape properties such as equal sides, parallel lines or perpendicular diagonals, and solve only for coordinates not already fixed.
A diagram may not be drawn to scale. Coordinates must follow stated properties and axis values, not visual appearance.
For endpoints A(x1,y1) and B(x2,y2), the midpoint is M=((x1+x2)/2,(y1+y2)/2).
For A(−5,4) and B(3,−2), M=((−5+3)/2,(4−2)/2)=(−1,1).
The midpoint lies halfway in both directions, so average the two x-coordinates and independently average the two y-coordinates.
Do not average an x-coordinate with a y-coordinate, and keep negative values inside brackets when adding.
A straight-line conversion graph represents a constant rate between two quantities. If zero converts to zero, the line passes through the origin.
| Task | Action |
|---|---|
| draw | choose accurate conversion pairs, plot them, join with a straight line |
| convert across | start at the known axis, move to the line, then to the other axis |
| check | confirm direction, units and sensible magnitude |
| extend | use the same constant gradient only while the rate remains fixed |
If 10 dollars converts to 60 krone, the rate is 6 krone per dollar. Thus 40 dollars corresponds to 240 krone.
A conversion graph can be read in either direction, but the numerical rate must be inverted when the direction is reversed.
For a straight line, gradient=Δy/Δx=rise/run. Use two points on the line and measure changes in the same direction.
| Line direction left to right | Gradient |
|---|---|
| rises | positive |
| falls | negative |
| horizontal | zero |
| vertical | undefined |
Choose two clear, widely separated points on the line. Form a right-angled gradient triangle and divide the vertical change by the horizontal change.
Gradient is not y/x unless one chosen point is the origin. In general it is a change divided by a change.
In y=mx+c, m is the gradient and c is the y-coordinate where the line crosses the y-axis, so the intercept is (0,c).
| Task | Method |
|---|---|
| read equation | identify coefficient m and constant c |
| write from graph | find gradient, then read the y-intercept |
| write from gradient and intercept | substitute directly into y=mx+c |
| rearrange | make y the subject before reading m,c |
For 2y=−7x+10, divide every term by 2: y=−3.5x+5. The gradient is −3.5 and the intercept is (0,5).
The x-intercept is not c. Set y=0 and solve separately if an x-intercept is required.
Choose suitable x-values, calculate each y-value accurately, plot the coordinate pairs, then join them with the correct shape.
| Function | Expected graph |
|---|---|
| y=mx+c | straight line |
| x=k | vertical line |
| y=c | horizontal line |
| y=ax2+bx+c | smooth parabola |
| ax+by=c | rearrange to straight-line form when useful |
For a quadratic, use enough points around the turning point and look for symmetry. For a linear graph, two correct points determine the line, but a third point checks arithmetic.
Join quadratic points with a smooth curve, not straight line segments. Do not force a straight line through points from a non-linear rule.
A graph family is recognised from its intercepts, symmetry, turning points, end behaviour, asymptotes and periodicity—not from one isolated point.
| Family | Signature features |
|---|---|
| quadratic | one turning point; parabolic shape |
| cubic | opposite end directions; up to two turning points |
| reciprocal k/x | two branches; axes are asymptotes |
| sine/cosine | smooth periodic waves |
| tangent | repeating branches with vertical asymptotes |
For polynomial graphs, calculate a table including intercept regions and turning behaviour. For trigonometric graphs in degrees, mark key angles and repeat using the correct period.
A reciprocal or tangent graph must not be drawn through a vertical asymptote. Separate branches never join across an undefined input.
A transformation changes known points of y=f(x) without rebuilding the whole value table. Changes outside f act vertically; changes inside f act horizontally.
| New graph | Point mapping from (x,y) | Effect |
|---|---|---|
| y=f(x)+a | (x,y)→(x,y+a) | up by a |
| y=af(x) | (x,y)→(x,ay) | vertical scale factor a |
| y=f(x+a) | (x,y)→(x−a,y) | left by a |
| y=f(ax) | (x,y)→(x/a,y) | horizontal scale factor 1/a |
Transform several defining points—intercepts, vertices and turning points—then preserve the original curve's connections and shape.
Inside changes act in the opposite horizontal direction: f(x+3) moves the graph 3 units left, not right.
Compare a distinctive point on the original and transformed graphs. Decide whether x-coordinates or y-coordinates changed, then test the corresponding transformation rule on another point.
| Observation | Likely algebra |
|---|---|
| every y increases by a | f(x)+a |
| every x decreases by a | f(x+a) |
| every y is multiplied by a | af(x) |
| every x is divided by a | f(ax) |
Use invariant features: vertical translations keep x-coordinates of turning points; horizontal transformations keep their y-coordinates.
Describe exactly one coherent transformation unless the graph genuinely shows a combination. A visual guess must be verified with coordinates.
A curve has a changing gradient. Its gradient at one point is estimated by the gradient of the tangent touching the curve there.
| Step | Action |
|---|---|
| 1 | draw a tangent that matches the curve's local direction |
| 2 | choose two well-separated points on the tangent |
| 3 | calculate Δy/Δx |
| 4 | include a sign and sensible precision |
Use points on the tangent, not necessarily points on the curve. A large gradient triangle reduces the effect of reading error.
A chord through two curve points gives an average gradient over an interval; it is not automatically the gradient at the named point.
At an intersection, two graphs have the same x and y. Therefore the intersection x-coordinates solve y2−y1=0.
| Task | Action |
|---|---|
| prepare | write each side as a graph y=y1 and y=y2 |
| draw | plot both on the same axes |
| solve | read every intersection's x-coordinate |
| report | give estimates to precision supported by the scale |
To solve x2−5x=x−7, plot y=x2−5x and y=x−7, or rearrange consistently to use an already drawn curve and a suitable line.
Do not report the intersection's y-coordinate when the equation asks for x. Check for all intersections in the shown domain.
For distinct points (x1,y1) and (x2,y2), m=(y2−y1)/(x2−x1).
Between (9,−4) and (5,8), m=(8−(−4))/(5−9)=12/(−4)=−3.
Subtract coordinates in the same order on top and bottom. Reversing both orders gives the same gradient.
If x2=x1, the denominator is zero and the vertical line has undefined gradient—not gradient zero.
Parallel lines have equal gradients. For two non-vertical perpendicular lines, their gradients satisfy m1m2=−1.
| Step | Action |
|---|---|
| 1 | rearrange the given line to identify its gradient |
| 2 | keep that gradient for parallel, or use the negative reciprocal for perpendicular |
| 3 | substitute the given point into y=mx+c to find c |
| 4 | verify both the point and gradient relationship |
A line perpendicular to y=−4x+5 has gradient 1/4. Through (3,7) it satisfies 7=(1/4)(3)+c, so c=25/4.
Changing only the sign is not enough for a perpendicular gradient. Take the reciprocal and change the sign.
A rate of change compares how quickly one quantity changes with another. A straight line has constant gradient; a curve usually has a gradient that varies from point to point.
| Situation | Rate of change |
|---|---|
| displacement against time | velocity |
| velocity against time | acceleration |
| curve y=f(x) | gradient dy/dx |
| tangent at a point | instantaneous rate there |
For a very small change in x, the corresponding change in y is approximately (dy/dx)Δx. Differentiation gives this local gradient exactly for an algebraic function.
An average rate over an interval is the gradient of a chord; an instantaneous rate is the gradient of a tangent at one point.
Use the power rule: d(axn)/dx=anxn−1 for an integer power n. Differentiate every term separately, and constants differentiate to zero.
| Term | Derivative |
|---|---|
| 5x3 | 15x2 |
| −2x | −2 |
| 7 | 0 |
| 4/x2=4x−2 | −8x−3 |
If y=4x3+5x2+2x−7, then dy/dx=12x2+10x+2.
The exponent decreases by 1 after multiplying by the old exponent. Do not merely reduce the exponent or differentiate a constant as 1.
The derivative gives the gradient function. Substitute an x-value for a gradient at a point; solve dy/dx=0 to find stationary points.
| Task | Method |
|---|---|
| gradient at x=a | find dy/dx, then substitute a |
| stationary x-values | solve dy/dx=0 |
| stationary coordinates | substitute each x into the original y |
| negative gradient | solve dy/dx<0 |
For y=x3−x2−8x+12, dy/dx=3x2−2x−8=(3x+4)(x−2). Thus stationary points occur at x=−4/3 and x=2.
Solving dy/dx=0 gives only the x-coordinates. Use the original function—not the derivative—to find the corresponding y-coordinates.
A stationary point is a local maximum when the graph rises then falls, and a local minimum when it falls then rises.
| Gradient change through point | Classification |
|---|---|
| positive to negative | local maximum |
| negative to positive | local minimum |
| no sign change | stationary point of inflection |
Sketch or inspect the general shape around each stationary point. The curve's direction on both sides decides whether the point is a maximum or minimum.
‘Maximum’ or ‘minimum’ may mean local unless a domain is given. Endpoints must also be checked when finding an absolute extreme on a closed interval.
A horizontal tangent alone does not prove maximum or minimum; a stationary point can be an inflection if the gradient keeps the same sign.
For displacement s(t), velocity is v=ds/dt and acceleration is a=dv/dt=d2s/dt2. Keep units consistent: metres, seconds, m/s and m/s².
| Request | Equation to use |
|---|---|
| instantaneously at rest | set v=0 |
| given acceleration | set a to that value |
| displacement at a time | substitute into s(t) |
| optimise a quantity | differentiate model, solve derivative =0, classify |
If s=2t3−5t2+6t−5, then v=6t2−10t+6 and a=12t−10. Setting a=5 gives t=1.25 s.
Reject solutions outside the stated physical domain, such as negative time, impossible length or a value making a denominator zero.
Velocity can be negative because it includes direction; speed is ∣v∣ and cannot be negative.