BACK Mascot image.
← MA0 0 · Math Prepa for University

Lesson 3

Graphs, and Coordinate Geometry

Mappings, domain and range, the quadratic function and inequalities, the shapes of simple functions and their inverses, and the coordinate geometry of the straight line.

Taught

Functions and Mappings

Mappings

Consider the function ff defined by f:x2x+1f : x \mapsto 2x + 1. If we input the value 22 for xx, ff gives 55 as output, and we say that this function maps 22 to 55, written 252 \mapsto 5. Similarly ff maps 2-2 to 3-3, 1-1 to 1-1 and 00 to 11. Here one value of the independent variable xx maps to just one value of the dependent variable f(x)f(x), and no value of f(x)f(x) is reached from two different values of xx.

Now consider f:xx22x+6f : x \mapsto x^2 - 2x + 6. This function maps 1-1 to 99, 00 to 66, 11 to 55, 22 to 66 and 33 to 99. One value of xx again gives just one value of f(x)f(x), but a value of f(x)f(x) is not necessarily obtained from only one value of xx: both 00 and 22 map to 66.

−2−1012−3−1135x ↦ 2x + 1, one-one−10123569xx² − 2x + 6, many-one
Figure 3.1. Mapping diagrams for the two functions. On the left every output is reached from exactly one input; on the right 66 and 99 are each reached twice.

Definition 3.1 (One-One and Many-One Mappings).

A mapping is one-one if each output arises from exactly one input, and many-one if at least one output arises from more than one input. A mapping which sends some input to more than one output is one-many.

So x2x+1x \mapsto 2x + 1 is a one-one mapping and xx22x+6x \mapsto x^2 - 2x + 6 a many-one mapping.

We may regard a function as a rule for mapping a number aa to a number bb. The strict meaning of the word, however, is restricted to those relationships in which one input value gives rise to just one output value, and a function in the sense of Definition 2.1 has this built in, since it associates to each number a single other number. Both mappings above are functions. The relationship x±xx \mapsto \pm\sqrt{x} is not: input 44 and it gives two outputs, 22 and 2-2. It is one-many, a perfectly good mapping for positive values of xx, and not a function.

Graphs of Functions

Take again f(x)=x22x+6f(x) = x^2 - 2x + 6. For any chosen value of xx the corresponding value of f(x)f(x) can be calculated.

xx1-10011223344
f(x)f(x)99665566991414

Arranged in order of increasing xx, the values of f(x)f(x) show a pattern, and the pattern is easier to see when the results are displayed graphically: we plot the values of f(x)f(x) on a vertical number line against the corresponding values of xx on a horizontal one.

−2−1123451015xy(1, 5)x = 1
Figure 3.2. The curve y=x22x+6y = x^2 - 2x + 6, with its values at the integers from 2-2 to 44 marked.

Figure 3.2 was drawn through values of xx one unit apart. Values taken from a larger range, or closer together, all lie on the same smooth curve, so the curve represents every pair of related values of xx and f(x)f(x). From it we can read off properties of the function.

  1. The lowest point on the curve is where x=1x = 1 and f(x)=5f(x) = 5. Although xx, as the independent variable, may be given any real value, the corresponding value of f(x)f(x) is always greater than 55, or equal to 55 when x=1x = 1. We say that the function has a least value of 55.
  2. The curve is symmetrical about the line x=1x = 1: two values of xx equidistant from 11, of the form 1+a1 + a and 1a1 - a, give the same value of f(x)f(x).
  3. Conversely, every value of f(x)f(x) above the least corresponds to two values of xx, of the form 1+a1 + a and 1a1 - a.

All three can be confirmed algebraically. Completing the square on the right-hand side gives

f(x)=x22x+6=(x1)2+5.f(x) = x^2 - 2x + 6 = (x - 1)^2 + 5 .

For the first, (x1)2(x - 1)^2 is a squared quantity and so is never negative. Its least value is 00, which occurs when x=1x = 1, so the least value that f(x)f(x) can have is 55, at x=1x = 1.

For the second, take two values of xx placed symmetrically on either side of x=1x = 1:

f(1+a)=a2+5,f(1a)=(a)2+5=a2+5,f(1 + a) = a^2 + 5, \qquad f(1 - a) = (-a)^2 + 5 = a^2 + 5,

so values of xx symmetrical about x=1x = 1 give the same value of f(x)f(x).

For the third, the values of xx giving any particular value cc of f(x)f(x) solve x22x+6c=0x^2 - 2x + 6 - c = 0, and by the quadratic formula the roots of this equation are

x=1+c5andx=1c5.x = 1 + \sqrt{c - 5} \qquad\text{and}\qquad x = 1 - \sqrt{c - 5} .

For these to be real we need c5c \geqslant 5, which is the least value once more, and the two values of xx corresponding to one value of f(x)f(x) are symmetrical about x=1x = 1.

Domain and Range

For f(x)=x22x+6f(x) = x^2 - 2x + 6 we have assumed that any real value of xx may be input. A full definition of a function must state the set of permissible input values, so the function illustrated above is fully defined as

f:xx22x+6,xR,f : x \mapsto x^2 - 2x + 6, \quad x \in \RR,

where xRx \in \RR means that xx may be any real number.

Definition 3.2 (Domain and Range).

The set of input values for which a function is defined is its domain. Once the domain is fixed there is a corresponding set of output values, called the range of the function, or its image-set.

The domain is not always the whole of R\RR, as the examples below show. The analysis above also showed that, for the domain xRx \in \RR, the output values were limited: the range of f:xx22x+6f : x \mapsto x^2 - 2x + 6, xRx \in \RR, is f(x)5f(x) \geqslant 5.

Example 3.3.

Sketch the function f:x2x+1f : x \mapsto 2x + 1, xRx \in \RR, and state its range.

The graph is the straight line through (0,1)(0, 1) rising two units for each unit to the right. As xx can be any real number the line is infinite in both directions, so the range of ff is also the whole of R\RR.

Example 3.4.

For the function of the previous example, let AA be the point on the line for which x=2x = 2 and BB the point for which x=1x = -1. Define the function represented by the segment ABAB, and state its range.

The mapping represented by ABAB is x2x+1x \mapsto 2x + 1, but only for values of xx from 1-1 to 22. So the function represented by ABAB is

f:x2x+1,xR,1x2,f : x \mapsto 2x + 1, \quad x \in \RR, \quad -1 \leqslant x \leqslant 2,

and its range is the set of values of f(x)f(x) corresponding to 1x2-1 \leqslant x \leqslant 2, that is 1f(x)5-1 \leqslant f(x) \leqslant 5.

Example 3.5.

State the image-set of the function f(x)=2x+1f(x) = 2x + 1, x{0,1,2,3,4}x \in \{0, 1, 2, 3, 4\}.

For this function there are just five input values, and the corresponding output values form the image-set, which is {1,3,5,7,9}\{1, 3, 5, 7, 9\}. The graphical representation of this function is not a line but the set of five points.

−2−1123−35xyAB−1 ⩽ x ⩽ 2123413579xyx ∈ {0, 1, 2, 3, 4}
Figure 3.3. One rule on two different domains: the segment ABAB of the second example, and the five points of the third.

Remark.

When the domain of a function is not stated, it is taken to be xRx \in \RR.

Problem 3.1.

  1. Sketch the graph of f:xx2f : x \mapsto x^2, xRx \in \RR, and state its range. Then redefine the domain as x0x \geqslant 0 and sketch the graph that now represents ff.
  2. Each of the mappings aa2+5a \mapsto a^2 + 5, b±bb \mapsto \pm\sqrt{b} and ccc \mapsto -\sqrt{c} has for its domain the non-negative real numbers. State which of them are functions.

The Quadratic Function

Functions of similar form usually have properties in common, so their graphs are usually similar in shape. Knowing the common characteristics of a family of functions lets us sketch any one of them without calculating a table of values.

Definition 3.6 (Quadratic Function).

A function of the form f(x)=ax2+bx+cf(x) = ax^2 + bx + c, where aa, bb and cc are constants and a0a \neq 0, is a quadratic function.

The function x22x+6x^2 - 2x + 6 analysed above is one. Completing the square on the right-hand side of the general form gives

f(x)=a(x+b2a)2+4acb24a.f(x) = a\left(x + \frac{b}{2a}\right)^2 + \frac{4ac - b^2}{4a} .

Whatever value xx takes, 4acb24a\dfrac{4ac - b^2}{4a} is constant, equal to KK say, and (x+b2a)20\left(x + \dfrac{b}{2a}\right)^2 \geqslant 0 as it is a squared quantity. So the function has the general form

f(x)=K+a×(zero or a positive quantity).f(x) = K + a \times (\text{zero or a positive quantity}) .
  1. If aa is positive, f(x)f(x) is at least equal to KK: it has a least value of 4acb24a\dfrac{4ac - b^2}{4a}, occurring when x=b2ax = -\dfrac{b}{2a}.
  2. If aa is negative, f(x)f(x) can never be greater than KK: it has a greatest value of 4acb24a\dfrac{4ac - b^2}{4a}, occurring when x=b2ax = -\dfrac{b}{2a}.

So x=b2ax = -\dfrac{b}{2a} is the value of xx corresponding to the greatest or least value of f(x)f(x). Taking two values of xx symmetrical about it, x=b2a±kx = -\dfrac{b}{2a} \pm k,

f(b2a+k)=ak2+K=f(b2ak),f\left(-\frac{b}{2a} + k\right) = ak^2 + K = f\left(-\frac{b}{2a} - k\right),

that is, input values of xx symmetrical about x=b2ax = -\dfrac{b}{2a} give as output the same value of f(x)f(x). From this analysis we deduce that the curve representing f(x)=ax2+bx+cf(x) = ax^2 + bx + c is symmetrical about the line x=b2ax = -\dfrac{b}{2a}, called the axis of the curve, and that it turns at a least value when a>0a > 0 and at a greatest value when a<0a < 0.

xyleast valuea > 0xygreatest valuea < 0
Figure 3.4. The two alternative graphs of a quadratic function, each symmetrical about its dashed axis.

The curve representing any particular quadratic function can now be sketched from this information.

Example 3.7 (Sketching a quadratic function).

Sketch the curve representing f(x)=2x27x4f(x) = 2x^2 - 7x - 4.

Here a=2a = 2 and b=7b = -7, so the axis of the curve is x=b2a=74x = -\dfrac{b}{2a} = \dfrac{7}{4}, and as a>0a > 0 the function has a least value,

4acb24a=4(2)(4)(7)28=32498=818.\frac{4ac - b^2}{4a} = \frac{4(2)(-4) - (-7)^2}{8} = \frac{-32 - 49}{8} = -\frac{81}{8} .

So f(x)f(x) has a least value of 818-\dfrac{81}{8} when x=74x = \dfrac{7}{4}. To locate the curve accurately on the axes we need one more pair of corresponding values, and f(0)f(0) is the easiest to find: f(0)=4f(0) = -4.

−112345−10−55xy−½4−4(7/4, −81/8)x = 7/4
Figure 3.5. The curve y=2x27x4y = 2x^2 - 7x - 4.

A quicker sketch is available when the function factorises. Here

f(x)=2x27x4=(2x+1)(x4).f(x) = 2x^2 - 7x - 4 = (2x + 1)(x - 4) .

The coefficient of x2x^2 is positive, so f(x)f(x) has a least value. When f(x)=0f(x) = 0 the corresponding values of xx are the roots of (2x+1)(x4)=0(2x + 1)(x - 4) = 0, namely x=12x = -\tfrac12 and x=4x = 4, and the average of these, 74\tfrac74, gives the value of xx about which the curve is symmetrical.

Remark.

This quicker method is suitable only when the function factorises.

Problem 3.2.

Find the greatest or least value of each function, and the value of xx at which it occurs. Then sketch the graph of the third, showing its axis of symmetry clearly.

  1. x23x+5x^2 - 3x + 5;
  2. 32xx23 - 2x - x^2;
  3. (1+x)(2x)(1 + x)(2 - x).

Inequalities

Consider the real numbers 55 and 22, for which 5>25 > 2. The introduction of an extra term on both sides leaves the inequality sign unchanged:

5+4>2+4,that is9>6,and54>24,that is1>2.5 + 4 > 2 + 4, \quad\text{that is}\quad 9 > 6, \qquad\text{and}\qquad 5 - 4 > 2 - 4, \quad\text{that is}\quad 1 > -2 .

Multiplication of both sides by a positive number also leaves the sign unchanged: 5×2>2×25 \times 2 > 2 \times 2, that is 10>410 > 4. However, if we multiply both sides by a negative number the inequality is no longer true. Multiplying by 2-2, the left-hand side becomes 10-10 and the right-hand side becomes 4-4, and 10<4-10 < -4. This illustrates the general fact that multiplication or division of both sides by a negative number reverses the inequality sign.

To summarise, if aa and bb are real numbers such that a>ba > b, then

  1. a+k>b+ka + k > b + k for all real values of kk;
  2. ak>bkak > bk for positive values of kk;
  3. ak<bkak < bk for negative values of kk.

These are the rules for manipulating an inequality.

Example 3.8.

Find the range of values of xx satisfying the inequality x3<2x+5x - 3 < 2x + 5.

x3<2x+5x<2x+8adding 3 to both sidesx<8subtracting 2x from both sidesx>8multiplying by 1\begin{aligned} x - 3 &< 2x + 5 \\ x &< 2x + 8 && \text{adding } 3 \text{ to both sides} \\ -x &< 8 && \text{subtracting } 2x \text{ from both sides} \\ x &> -8 && \text{multiplying by } -1 \end{aligned}

Therefore the range of values of xx that satisfies the inequality is x>8x > -8.

Problem 3.3.

Find the range of values of xx that satisfies each inequality.

  1. 2x1<x42x - 1 < x - 4;
  2. 2(x1)>3(x1)2(x - 1) > 3(x - 1);
  3. x+2>4xx + 2 > 4 - x.

Quadratic Inequalities

Definition 3.9 (Quadratic Inequality).

An inequality that involves a quadratic function is a quadratic inequality.

An example is (x2)(2x+1)>0(x - 2)(2x + 1) > 0. The range of values of xx satisfying it can be found graphically. Let f(x)=(x2)(2x+1)f(x) = (x - 2)(2x + 1); the inequality asks for the values of xx at which the curve y=f(x)y = f(x) lies above the xx-axis.

−2−1123−336xy−½2
Figure 3.6. The curve y=(x2)(2x+1)y = (x - 2)(2x + 1), with the parts above the xx-axis drawn boldly.

From the sketch we see that f(x)>0f(x) > 0, where the curve is above the xx-axis, for values of xx greater than 22 and for values less than 12-\tfrac12. Therefore the ranges of values of xx that satisfy (x2)(2x+1)>0(x - 2)(2x + 1) > 0 are

x<12andx>2.x < -\tfrac12 \qquad\text{and}\qquad x > 2 .

Problem 3.4.

Find the range, or ranges, of values of xx that satisfy each inequality.

  1. (x1)(x2)>0(x - 1)(x - 2) > 0;
  2. x24x>5x^2 - 4x > 5;
  3. (32x)(x+5)>0(3 - 2x)(x + 5) > 0.

Problems Involving Quadratic Inequalities

Example 3.10.

Find the range of values of kk for which the equation x2kx+(k+3)=0x^2 - kx + (k + 3) = 0 has real roots.

Real roots, equal or distinct, require a discriminant which is not negative:

(k)24(k+3)0,that isk24k120.(-k)^2 - 4(k + 3) \geqslant 0, \qquad\text{that is}\qquad k^2 - 4k - 12 \geqslant 0 .

Let f(k)=k24k12=(k6)(k+2)f(k) = k^2 - 4k - 12 = (k - 6)(k + 2). Its graph has a least value and meets the axis at k=2k = -2 and k=6k = 6, and it lies on or above the axis outside these two values. So the equation has real roots for

k2andk6.k \leqslant -2 \qquad\text{and}\qquad k \geqslant 6 .

Example 3.11.

Find the set of values of pp for which f(x)=x2+3px+pf(x) = x^2 + 3px + p is greater than zero for all real values of xx.

f(x)f(x) is a quadratic function of xx whose coefficient of x2x^2 is positive, so f(x)f(x) has a least value. So if f(x)>0f(x) > 0 for all xx, the least value of f(x)f(x) has to be greater than zero. Completing the square on the right-hand side gives

f(x)=(x+3p2)2+p9p24,f(x) = \left(x + \frac{3p}{2}\right)^2 + p - \frac{9p^2}{4},

so the least value of f(x)f(x) is p9p24p - \dfrac{9p^2}{4}, and for f(x)>0f(x) > 0 for all xx we need

p9p24>0,that is4p9p2>0,that isp(49p)>0.p - \frac{9p^2}{4} > 0, \qquad\text{that is}\qquad 4p - 9p^2 > 0, \qquad\text{that is}\qquad p(4 - 9p) > 0 .

Let g(p)=p(49p)g(p) = p(4 - 9p). Its graph has a greatest value and meets the axis at p=0p = 0 and p=49p = \tfrac49, and it is positive between them. Therefore f(x)>0f(x) > 0 for all real xx for the set of values of pp given by 0<p<490 < p < \tfrac49.

Example 3.12.

Find the ranges of values of xx for which 2x1<x24<122x - 1 < x^2 - 4 < 12.

There are two inequality relationships here, and we are looking for the values of xx that satisfy both of them.

  1. 2x1<x242x - 1 < x^2 - 4 gives x22x3>0x^2 - 2x - 3 > 0. Let f(x)=x22x3=(x3)(x+1)f(x) = x^2 - 2x - 3 = (x - 3)(x + 1); then f(x)>0f(x) > 0 for x<1x < -1 and x>3x > 3.
  2. x24<12x^2 - 4 < 12 gives x216<0x^2 - 16 < 0. Let g(x)=x216=(x4)(x+4)g(x) = x^2 - 16 = (x - 4)(x + 4); then g(x)<0g(x) < 0 for 4<x<4-4 < x < 4.

Illustrating these ranges on a number line, the values of xx that satisfy both are where the lines overlap:

4<x<1and3<x<4.-4 < x < -1 \qquad\text{and}\qquad 3 < x < 4 .
−5−4−3−2−10123451−5−4−3−2−10123452−5−4−3−2−1012345both
Figure 3.7. The ranges given by the two inequalities, and their overlap.

Some Other Simple Functions

A table of values for each of the following functions gives enough information to deduce the shape of the curve that represents it.

Exponential Functions

Definition 3.13 (Exponential Function).

A function in which the variable appears as an exponent, that is as an index, is an exponential function. The functions 2x2^x, 3x3^x and 10x10^x are all exponential functions of xx.

Consider the function f(x)=2xf(x) = 2^x, for which the following table shows corresponding values of xx and f(x)f(x).

xx10-105-54-43-32-21-10011223344551010
2x2^x11024\tfrac{1}{1024}132\tfrac{1}{32}116\tfrac{1}{16}18\tfrac1814\tfrac1412\tfrac12112244881616323210241024

From this table we see that

  1. f(x)>0f(x) > 0 for all real values of xx;
  2. as xx increases, f(x)f(x) increases at a rapidly accelerating rate;
  3. f(x)=1f(x) = 1 when x=0x = 0;
  4. as xx decreases, through x=10,100,x = -10, -100, \ldots, f(x)f(x) quickly becomes numerically smaller, and we say that as xx approaches minus infinity, f(x)f(x) approaches the value zero. This is written xx \to -\infty, f(x)0f(x) \to 0.

From these observations the sketch of f(x)=2xf(x) = 2^x is drawn.

−4−3−2−11231248xyasymptote
Figure 3.8. The curve y=2xy = 2^x and its asymptote, the xx-axis.

The curve approaches the xx-axis but never actually touches it or crosses it.

Definition 3.14 (Asymptote).

A line which a curve approaches more and more closely, without ever reaching it, is an asymptote to the curve.

So the xx-axis is an asymptote to the curve y=2xy = 2^x. Another way of expressing the behaviour of f(x)f(x) for negative values of xx is that f(x)f(x) approaches a limiting value, or limit, of zero as xx approaches minus infinity. This property is written

limx2x=0.\lim_{x \to -\infty} 2^x = 0 .

Any function of the form axa^x, where a>1a > 1, is represented by a curve similar to that deduced for 2x2^x.

Rational Functions

Definition 3.15 (Rational Function).

A function whose numerator and denominator are both polynomials is a rational function. These are the fractional functions of the last lesson, under their other name.

For example 1x\dfrac{1}{x}, xx2\dfrac{x}{x - 2} and x27x+1\dfrac{x^2 - 7}{x + 1} are rational functions of xx. Consider the function f(x)=1xf(x) = \dfrac{1}{x} and the following table of corresponding values.

xx3-32-21-100112233441010
1x\tfrac{1}{x}13-\tfrac1312-\tfrac121-1undefined1112\tfrac1213\tfrac1314\tfrac14110\tfrac{1}{10}

From this table we see that

  1. for x>0x > 0, f(x)>0f(x) > 0, and as xx \to \infty, f(x)0f(x) \to 0;
  2. for x<0x < 0, f(x)<0f(x) < 0, and as xx \to -\infty, f(x)0f(x) \to 0, so that xx and f(x)f(x) always have the same sign;
  3. for x=0x = 0, f(0)=10f(0) = \dfrac10, which has no finite value and is said to be undefined.

In such circumstances we investigate the behaviour of f(x)f(x) as xx approaches zero. Now xx can approach zero in two ways: it can decrease from positive values towards zero, which is to approach zero from above, or increase from negative values towards zero, which is to approach zero from below.

xx110.10.10.010.010.0010.001
1x\tfrac{1}{x}11101010010010001000
xx1-10.1-0.10.01-0.010.001-0.001
1x\tfrac{1}{x}1-110-10100-1001000-1000

From these tables we see that as xx decreases to zero, f(x)f(x) \to \infty, and as xx increases to zero, f(x)f(x) \to -\infty. From these observations we can draw the sketch representing f(x)=1xf(x) = \dfrac{1}{x}.

−4−3−2−11234−4−224xy
Figure 3.9. The curve y=1xy = \dfrac{1}{x}. Both axes are asymptotes.

This curve has two asymptotes, the horizontal and the vertical axes. All the curves looked at so far have been unbroken, or continuous, but this curve has a break, or discontinuity, at the point where x=0x = 0 and f(0)f(0) is undefined. In general, if f(x)f(x) is undefined for a finite value of xx, x=ax = a say, then the curve representing f(x)f(x) has a discontinuity where x=ax = a.

The function 1x\dfrac{1}{x} does not approach a unique value as xx approaches zero: it goes to \infty or to -\infty depending on whether xx approaches zero from above or from below. In this case we say that limx01x\displaystyle\lim_{x \to 0} \frac{1}{x} does not exist. In general, for limxaf(x)\displaystyle\lim_{x \to a} f(x) to exist with a value kk, say, f(x)f(x) must approach kk both as xx approaches aa from above and as xx approaches aa from below.

Logarithmic Functions

For values of a>0a > 0 with a1a \neq 1, any function of the form logax\log_a x, loga(2x+1)\log_a (2x + 1), and so on, is a logarithmic function, the logarithm being that of Definition 2.21. Consider the function f(x)=log2xf(x) = \log_2 x and the following table of corresponding values.

xx1-10014\tfrac1412\tfrac1211224488
log2x\log_2 xdoes not existundefined2-21-100112233

From this table we see that

  1. when x=1x = -1, f(1)=log2(1)=bf(-1) = \log_2(-1) = b say; but there is no real value of bb for which 2b=12^b = -1, since every power of 22 is positive. So f(1)f(-1) does not exist, and all negative values of xx lead to the same conclusion: log2x\log_2 x does not exist for negative values of xx;
  2. for x>1x > 1, f(x)>0f(x) > 0, and as xx \to \infty, f(x)f(x) \to \infty;
  3. for x=0x = 0, f(0)f(0) is undefined, so we investigate the behaviour of f(x)f(x) as xx approaches zero from above. As xx decreases to zero, f(x)f(x) \to -\infty, and for 0<x<10 < x < 1, f(x)<0f(x) < 0.

From these observations we can sketch the graphical representation of f(x)=log2xf(x) = \log_2 x.

1248−3−2−1123xy
Figure 3.10. The curve y=log2xy = \log_2 x.

Any function of the form logax\log_a x with a>1a > 1 has a graph of similar shape. Note that, while logax\log_a x does not exist for negative values of xx, the value of logax\log_a x can itself be negative.

Simple Variations of Functions

The curve representing f(x)=2xf(x) = 2^x is now familiar, and from it we can obtain the graphs of some simple variations.

  1. g(x)=2x+1g(x) = 2^x + 1. Comparing g(x)=2x+1g(x) = 2^x + 1 with f(x)=2xf(x) = 2^x, for any one value of xx, g(x)g(x) is one unit greater than f(x)f(x). So the curve representing g(x)g(x) is the same shape as that representing f(x)f(x) but raised vertically by one unit. In general, the curve representing f(x)+cf(x) + c is the curve representing f(x)f(x) raised vertically by cc units.
  2. g(x)=2xg(x) = 2^{-x}. For g(x)g(x) an input x=ax = a gives the output 2a2^{-a}, while for f(x)f(x) the input x=ax = -a gives the same output. So g(a)=f(a)g(a) = f(-a), that is g(x)=f(x)g(x) = f(-x) for xRx \in \RR, and the curve representing g(x)g(x) is the same as that representing f(x)f(x) with the negative and positive values of xx transposed. In general, the curve representing f(x)f(-x) is the reflection in the vertical axis of the curve representing f(x)f(x).
  3. g(x)=2xg(x) = -2^x. Here g(x)=f(x)g(x) = -f(x), so the curve representing g(x)g(x) is the same shape as that for f(x)f(x) with the positive and negative values of f(x)f(x) transposed. In general, the curve representing f(x)-f(x) is the reflection in the horizontal axis of the curve representing f(x)f(x).
xyy = 2x + 1xyy = 2xxyy = −2x
Figure 3.11. The three variations of y=2xy = 2^x, each drawn against the faint original.

Problem 3.5.

Write down the values of f(x)=(12)xf(x) = \left(\tfrac12\right)^x corresponding to x=2,4,6x = 2, 4, 6 and to x=2,4,6x = -2, -4, -6. From these values deduce the behaviour of f(x)f(x) as xx \to \infty and as xx \to -\infty, and sketch the graph of f(x)f(x), marking any asymptote. Which variation of 2x2^x is this curve?

Inverse Functions

Consider the mapping f:x2xf : x \mapsto 2x, x{2,3,4}x \in \{2, 3, 4\}. Under this function the domain {2,3,4}\{2, 3, 4\} maps to the image-set {4,6,8}\{4, 6, 8\}.

It is possible to reverse this mapping: we can map each member of the image-set {4,6,8}\{4, 6, 8\} back to the corresponding member of the domain by halving it, 424 \mapsto 2, 636 \mapsto 3, 848 \mapsto 4. Expressed as an algebraic relationship, if x{4,6,8}x \in \{4, 6, 8\} then x12xx \mapsto \tfrac12 x maps 424 \mapsto 2, 636 \mapsto 3, 848 \mapsto 4.

This reverse mapping is a one-one mapping, so it is a function in its own right, and it is called the inverse function of ff. Denoting this inverse function by f1f^{-1}, we see that f1:x12xf^{-1} : x \mapsto \tfrac12 x, x{4,6,8}x \in \{4, 6, 8\}, reverses the mapping f:x2xf : x \mapsto 2x, x{2,3,4}x \in \{2, 3, 4\}. In fact f:x2xf : x \mapsto 2x can be reversed for all real values of xx, so if ff is the function defined by f:x2xf : x \mapsto 2x, xRx \in \RR, then f1f^{-1} is the function which reverses this mapping, defined by

f1:x12x,xR.f^{-1} : x \mapsto \tfrac12 x, \quad x \in \RR .

Now consider the function f:xx2f : x \mapsto x^2, xRx \in \RR. This is a many-one mapping: there are two values of xx which map to one value of f(x)f(x), both 22 and 2-2 mapping to 44, for example. We can reverse this mapping by taking the positive and the negative square root of each member of the image-set, which in algebraic form is the mapping x±xx \mapsto \pm\sqrt{x}. However, this is a one-many mapping, so it is not a function, and we say that f:xx2f : x \mapsto x^2, xRx \in \RR, does not have an inverse function.

If, however, we restrict the domain of ff to x0x \geqslant 0, redefining the function as f:xx2f : x \mapsto x^2, x0x \geqslant 0, then it becomes a one-one mapping. The reverse mapping xxx \mapsto \sqrt{x} is also one-one, so the function f:xx2f : x \mapsto x^2, x0x \geqslant 0, does have an inverse, namely

f1:xx,x0.f^{-1} : x \mapsto \sqrt{x}, \quad x \geqslant 0 .

Definition 3.16 (Inverse Function).

A function ff maps the domain of ff to the image-set of ff. If the reverse mapping, of the image-set of ff to the domain of ff, is a function, it is called the inverse function of ff and is denoted by f1f^{-1}.

Remark.

If ff defines a one-one mapping then f1f^{-1} exists, but if ff defines a many-one mapping then f1f^{-1} does not exist.

The Graphs of Functions and Their Inverses

The graphs of f(x)=2xf(x) = 2x and f1(x)=12xf^{-1}(x) = \tfrac12 x for xRx \in \RR, and of f(x)=x2f(x) = x^2 and f1(x)=xf^{-1}(x) = \sqrt{x} for x0x \geqslant 0, are sketched below. Observing these, we see that in each case the graph of f1(x)f^{-1}(x) is the reflection of the graph of f(x)f(x) in the line y=xy = x.

24682468xyy = 2x and y = ½x12341234xyy = x², x ⩾ 0, and y = √x
Figure 3.12. Two functions and their inverses, each pair reflected in the dashed line y=xy = x.

This is true for the graph of any function ff and its inverse f1f^{-1}. The line y=xy = x is the graph of f:xxf : x \mapsto x, and this function is its own inverse. So if the graph representing a function is known, the graph representing its inverse can be sketched. Even when a function ff does not possess an inverse, the curve representing ff can still be reflected in the line y=xy = x, but in that case the reflected graph does not represent a function.

Example 3.17.

Given the function f(x)=2xf(x) = 2^x, xRx \in \RR, find f1f^{-1} as a function of xx and sketch the graph of f1f^{-1}.

The function ff maps xx to 2x2^x. To find f1f^{-1} we have to reverse this process, that is map values of 2x2^x back to values of xx. If 2x=w2^x = w, say, then taking logarithms to base 22 of each side gives x=log2wx = \log_2 w. Hence the relationship xwx \mapsto w can be expressed as wlog2ww \mapsto \log_2 w, and this is a one-one mapping for w>0w > 0, so it is a function, and it reverses the mapping x2xx \mapsto 2^x. Replacing the variable ww by the variable xx, the inverse of f(x)=2xf(x) = 2^x is

f1(x)=log2x,x>0.f^{-1}(x) = \log_2 x, \quad x > 0 .
−224−224xyy = 2xy = log₂ xy = x
Figure 3.13. The curve y=2xy = 2^x and its inverse y=log2xy = \log_2 x, reflected in the dashed line y=xy = x.

In the same way, for any base a>0a > 0 with a1a \neq 1, the function logax\log_a x from the positive numbers to the real numbers is the inverse of the function axa^x from the real numbers to the positive numbers.

Problem 3.6.

Each of the following functions has the domain xRx \in \RR. Determine which of them have an inverse, and where f1f^{-1} exists express it as a function of xx.

  1. 3x3x;
  2. 2x+12x + 1;
  3. x24x^2 - 4.

Coordinate Geometry

Locating a Point

Graphical methods lend themselves particularly well to the investigation of the geometric properties of many kinds of curves. We restrict ourselves to plane figures, those that can be described fully using only two dimensions, and to represent any figure on a graph we need, as a start, a simple and unambiguous way of describing the position of a point.

Consider the problem of describing the location of a town, Birmingham say. There are many ways in which this can be done, but all require reference to at least one known place and known directions, called a system, or frame, of reference. Within this frame of reference two measurements, or coordinates, are needed to locate the town precisely.

NOB50 km36°52′(i) distance and bearingNOBE30 km40 km(ii) east and north
Figure 3.14. The position of BB described in two alternative ways.

In the first description the system of reference is the fixed point OO and the direction due north from OO, and the coordinates of BB are 5050 km from OO on a bearing of 365236^\circ 52'. In the second the system of reference is the pair of directions due east and due north from the fixed point OO, and the coordinates of BB are 3030 km east of OO and 4040 km north of OO. The two systems most often used for mathematical analysis are basically similar to these two practical systems.

Polar Coordinates

The system of reference is a fixed point OO, called the pole, and a fixed direction from OO, the line OxOx, called the initial line. The coordinates of a point PP are the distance of PP from OO and the angle OPOP makes with OxOx, measured in an anticlockwise sense from OxOx. These coordinates are written as an ordered pair (r,θ)(r, \theta), that is (distance, angle).

Cartesian Coordinates

The system of reference is a fixed point OO, the origin, and a pair of perpendicular lines through OO. It is usual to draw these lines horizontally and vertically; the horizontal line is called the xx-axis and the vertical line the yy-axis. The coordinates of a point PP are the directed distances of PP from OO parallel to the axes: a positive coordinate is a distance measured in the positive direction of the axis, and a negative coordinate is a distance in the opposite direction. The coordinates are given as an ordered pair (a,b)(a, b), with the xx-coordinate, or abscissa, first and the yy-coordinate, or ordinate, second.

xO(2, 30°)30°2polar−2−112−2−11234xy(1, 4)(−2, −1)Cartesian
Figure 3.15. The point (2,30)(2, 30^\circ) in polar coordinates, and the points (1,4)(1, 4) and (2,1)(-2, -1) in Cartesian coordinates.

Coordinate geometry is the name given to the analysis, using graphical methods, of geometric properties. The properties of straight lines and of many curves are most simply found using Cartesian coordinates, so this system of reference is used more frequently than any other. For this analysis we need to refer to three types of points:

  1. fixed points whose coordinates are known, such as the point (4,5)(4, 5);
  2. fixed points whose coordinates are not known numerically, referred to as the points (x1,y1)(x_1, y_1), (x2,y2)(x_2, y_2), and so on, or (a,b)(a, b);
  3. points which are not fixed, called general points, a general point being referred to as the point (x,y)(x, y).

It is conventional to use the letters PP, QQ, RR for general points and AA, BB, CC for fixed points. To avoid distorting the shape of a curve when drawing it on a Cartesian plane, the two axes are graduated using identical scales.

Problem 3.7.

Represent on one diagram the points whose polar coordinates are (1,45)(1, 45^\circ), (3,90)(3, 90^\circ), (1,150)(1, 150^\circ) and (2,200)(2, 200^\circ), and on another the points whose Cartesian coordinates are (4,2)(4, 2), (1,5)(-1, 5), (0,3)(0, 3) and (2,5)(-2, -5).

The Length of the Line Joining Two Points

12341234xyA(1, 2)B(3, 4)N
Figure 3.16. The points A(1,2)A(1, 2) and B(3,4)B(3, 4), with N(3,2)N(3, 2) completing a right-angled triangle.

From the diagram we see that the length of the line joining A(1,2)A(1, 2) and B(3,4)B(3, 4) can be found by Pythagoras, using the point N(3,2)N(3, 2):

AB2=AN2+BN2=(31)2+(42)2=8,soAB=8=22.AB^2 = AN^2 + BN^2 = (3 - 1)^2 + (4 - 2)^2 = 8, \qquad\text{so}\qquad AB = \sqrt{8} = 2\sqrt{2} .

In general, if A(x1,y1)A(x_1, y_1) and B(x2,y2)B(x_2, y_2) are any two points, the point N(x2,y1)N(x_2, y_1) gives AN=x2x1AN = x_2 - x_1 and NB=y2y1NB = y_2 - y_1, and by Pythagoras AB2=AN2+NB2AB^2 = AN^2 + NB^2. Therefore the length of the line joining A(x1,y1)A(x_1, y_1) to B(x2,y2)B(x_2, y_2) is

AB=(x2x1)2+(y2y1)2.AB = \sqrt{(x_2 - x_1)^2 + (y_2 - y_1)^2} .

This formula still holds when some, or all, of the coordinates are negative. For A(2,2)A(-2, 2) and B(3,1)B(3, -1), AN=3(2)=5AN = 3 - (-2) = 5 and NB=12=3NB = -1 - 2 = -3, and AB=25+9=34AB = \sqrt{25 + 9} = \sqrt{34}.

The Midpoint of the Line Joining Two Points

1234123456xyA(1, 1)B(3, 5)M(2, 3)
Figure 3.17. The midpoint M(2,3)M(2, 3) of the line joining A(1,1)A(1, 1) and B(3,5)B(3, 5).

Let MM be the midpoint of the line joining A(1,1)A(1, 1) and B(3,5)B(3, 5). The foot of the perpendicular from MM to the xx-axis lies halfway between the feet of the perpendiculars from AA and BB, so the xx-coordinate of MM is

1+12(31)=12(3+1)=2.1 + \tfrac12(3 - 1) = \tfrac12(3 + 1) = 2 .

Similarly the yy-coordinate of MM is 1+12(51)=12(5+1)=31 + \tfrac12(5 - 1) = \tfrac12(5 + 1) = 3. Therefore MM is the point (2,3)(2, 3).

In general, if MM is the midpoint of the line joining A(x1,y1)A(x_1, y_1) and B(x2,y2)B(x_2, y_2), the xx-coordinate of MM is x1+12(x2x1)=12(x1+x2)x_1 + \tfrac12(x_2 - x_1) = \tfrac12(x_1 + x_2), the arithmetic mean of the xx-coordinates of AA and BB, and likewise its yy-coordinate is the arithmetic mean of the yy-coordinates. So the coordinates of MM are

(x1+x22,  y1+y22).\left( \frac{x_1 + x_2}{2}, \; \frac{y_1 + y_2}{2} \right) .

This formula also holds when some, or all, of the coordinates are negative: the midpoint of the line joining A(3,2)A(-3, -2) and B(1,3)B(1, 3) is (12(3+1),12(2+3))=(1,12)\left(\tfrac12(-3 + 1), \tfrac12(-2 + 3)\right) = \left(-1, \tfrac12\right).

Problem 3.8.

  1. Find the length of the line joining (1,2)(1, 2) and (4,6)(4, 6), and of the line joining (1,4)(-1, -4) and (3,2)(-3, -2).
  2. Find the coordinates of the midpoints of the same two lines.
  3. MM is the midpoint of ABAB. If AA is (5,7)(5, 7) and MM is (0,2)(0, 2), find the coordinates of BB.

Gradient

Definition 3.18 (Gradient).

The gradient of a straight line is a measure of its slope with respect to the xx-axis. It is the increase in the yy-coordinate divided by the increase in the xx-coordinate between one point on the line and another point on the line.

Consider the line passing through the points A(2,3)A(2, 3) and B(6,1)B(6, 1). From AA to BB the yy-coordinate decreases by 22, that is increases by 2-2, and the xx-coordinate increases by 44, so the gradient of ABAB is 24=12\dfrac{-2}{4} = -\dfrac12. From BB to AA the gradient is

increase in yincrease in x=24=12,\frac{\text{increase in } y}{\text{increase in } x} = \frac{2}{-4} = -\frac12,

so it does not matter in which order the two points are considered, provided they are considered in the same order when calculating the increases in both xx and yy.

Consider now the straight line through A(1,2)A(1, 2) and B(4,3)B(4, 3). From AA to BB the increase in the yy-coordinate is 11 and the increase in the xx-coordinate is 33, so the gradient of ABAB is 13\dfrac13. If CC and DD are any two other points on the same line, the right-angled triangle they make with lines parallel to the axes is similar to the one made by AA and BB, so it gives the same ratio: the gradient of a line may be found from any two points on the line.

From the two examples we see that the gradient of a line may be positive or negative. A positive gradient indicates an uphill slope with respect to the positive direction of the xx-axis, that is a line which makes an acute angle with the positive sense of the xx-axis. A negative gradient indicates a downhill slope, a line which makes an obtuse angle with it.

xyriserunpositive gradientxyfallrunnegative gradient
Figure 3.18. A positive gradient rises to the right; a negative gradient falls.

In general, the gradient of the line passing through A(x1,y1)A(x_1, y_1) and B(x2,y2)B(x_2, y_2) is

the increase in the y-coordinatethe increase in the x-coordinate=y2y1x2x1.\frac{\text{the increase in the } y\text{-coordinate}}{\text{the increase in the } x\text{-coordinate}} = \frac{y_2 - y_1}{x_2 - x_1} .

As the gradient of a straight line is the increase in yy divided by the increase in xx from one point on the line to another, the gradient measures the increase in yy per unit increase in xx, that is the rate of increase of yy with respect to xx.

Parallel Lines

If l1l_1 and l2l_2 are parallel lines, they are equally inclined to the positive direction of the xx-axis, so the gradient triangles of the two lines are similar and give the same ratio: parallel lines have equal gradients.

Perpendicular Lines

Consider two perpendicular lines whose gradients are m1m_1 and m2m_2. Draw through the origin a line OSOS parallel to the first and a line OROR parallel to the second, drop SS to TT on the xx-axis and RR to QQ on the yy-axis. Because OSOS and OROR are perpendicular, the triangle OQROQR is a copy of the triangle OTSOTS turned through a right angle, so the triangles are similar and

STOT=QROQ.\frac{ST}{OT} = \frac{QR}{OQ} .
xySTRQO
Figure 3.19. Perpendicular lines OSOS and OROR, and the similar triangles OTSOTS and OQROQR.

But the gradient of OSOS is STOT=m1\dfrac{ST}{OT} = m_1, and as OROR rises to the left the gradient of OROR is OQQR=m2-\dfrac{OQ}{QR} = m_2. Since STOT\dfrac{ST}{OT} and OQQR\dfrac{OQ}{QR} are reciprocals,

m1m2=STOT×(OQQR)=1.m_1 m_2 = \frac{ST}{OT} \times \left(-\frac{OQ}{QR}\right) = -1 .

So the product of the gradients of perpendicular lines is 1-1, or, if one line has a gradient mm, the gradient of any line perpendicular to it is 1m-\dfrac{1}{m}.

Example 3.19 (A point on a median).

Show that the point (67,0)\left(-\tfrac67, 0\right) is on the median through AA of triangle ABCABC, where AA, BB, CC are the points (2,4)(2, 4), (2,3)(-2, 3), (1,2)(1, -2). If also the point (a,b)(a, b) is on this median, find a relationship between aa and bb.

If ADAD is the median through AA, then DD is the midpoint of BCBC, that is the point (12,12)\left(-\tfrac12, \tfrac12\right). If E(67,0)E\left(-\tfrac67, 0\right) is on ADAD, the gradients of ADAD and AEAE should be equal. Now

gradient of AD=4122+12=75,gradient of AE=402+67=75,\text{gradient of } AD = \frac{4 - \tfrac12}{2 + \tfrac12} = \frac75, \qquad \text{gradient of } AE = \frac{4 - 0}{2 + \tfrac67} = \frac75,

so EE is on the median ADAD. A condition that P(a,b)P(a, b) should be on ADAD is that the gradient of APAP equals the gradient of ADAD:

4b2a=75,so205b=147a,that is7a5b+6=0.\frac{4 - b}{2 - a} = \frac75, \qquad\text{so}\qquad 20 - 5b = 14 - 7a, \qquad\text{that is}\qquad 7a - 5b + 6 = 0 .

Problem 3.9.

  1. Find the gradients of the lines passing through (1,3)(-1, -3) and (2,1)(-2, 1), and through (3,2)(3, -2) and (1,4)(-1, 4).
  2. Determine, by comparing gradients, whether the points (0,1)(0, -1), (1,1)(1, 1) and (2,3)(2, 3) are collinear, that is whether they lie on the same straight line.
  3. Determine whether ABAB is parallel or perpendicular to CDCD, where AA, BB, CC, DD are (0,1)(0, -1), (1,1)(1, 1), (1,5)(1, 5), (1,1)(-1, 1).

Equations and Regions

The Cartesian system of reference provides a means of defining the position of any point in a plane, and this plane is called the xyxy plane. In general xx and yy are independent variables, each able to take any value independently of the value of the other, unless some restriction is placed on them.

Consider the set of points for which x=2x = 2. As the value of yy is not restricted, these points all lie on the line parallel to the yy-axis passing through (2,0)(2, 0). So the equation x=2x = 2 defines this line in the xyxy plane; x=2x = 2 is called the equation of the line, which is briefly referred to as “the line x=2x = 2”.

Now consider the set of points for which x>2x > 2. All points to the right of the line x=2x = 2 have an xx-coordinate greater than 22, so the inequality x>2x > 2 defines the region of the xyxy plane to the right of the line. Similarly x<2x < 2 defines the region to its left.

12345−2−1123xyx > 2
Figure 3.20. The region x>2x > 2, with its boundary line drawn broken.

Remark.

The region defined by x>2x > 2 does not include the line x=2x = 2. When a region does not include the points on its boundary lines, these are drawn as broken lines; when it does include them, they are drawn as solid lines.

Now consider the function f(x)=(x3)(x+1)f(x) = (x - 3)(x + 1), and the curve representing it drawn on the xyxy plane. If PP, QQ and RR are points on a line x=x1x = x_1, with QQ on the curve, the yy-coordinate of QQ is f(x1)f(x_1). The yy-coordinate of PP, above QQ, is greater than that of QQ, and the yy-coordinate of RR, below QQ, is less. This argument applies for all values of x1x_1. Therefore the inequality y>(x3)(x+1)y > (x - 3)(x + 1) defines the set of points in the region above the curve, and the inequality y<(x3)(x+1)y < (x - 3)(x + 1) defines the region below it, whereas only for points on the curve is y=(x3)(x+1)y = (x - 3)(x + 1). This last is called the equation of the curve, and the curve is often referred to simply as the curve y=(x3)(x+1)y = (x - 3)(x + 1).

−2−11234−4−2246xyPQR
Figure 3.21. The region y>(x3)(x+1)y > (x - 3)(x + 1), with the points PP, QQ, RR on one vertical line.

An equation such as x27x+3=0x^2 - 7x + 3 = 0 contains only one variable, and its solution comprises a finite set of values of xx. An equation containing two variables, such as y=(x3)(x+1)y = (x - 3)(x + 1), has as its solution an infinite set of ordered pairs (x,y)(x, y). If AA is the solution set of the equation y=f(x)y = f(x), the elements of AA are the coordinates (x,y)(x, y) of all points on the curve y=f(x)y = f(x), and conversely the coordinates of points not on the curve are not elements of AA. So for a point PP on the curve and a point QQ off it, PAP \in A but QAQ \notin A.

In general, if f(x)f(x) is any function of xx, then in the xyxy plane

  1. y=f(x)y = f(x) defines the curve representing f(x)f(x), and is called the equation of that curve;
  2. y>f(x)y > f(x) and y<f(x)y < f(x) define the regions of the plane above and below that curve.

Example 3.20.

Determine whether the points (5,11)(5, 11) and (2,20)(-2, -20) are on the curve y=(x4)(x+6)y = (x - 4)(x + 6).

Substituting 1111 for yy in the left-hand side of the equation gives 1111, and substituting 55 for xx in the right-hand side gives (54)(5+6)=11(5 - 4)(5 + 6) = 11. The two sides are equal, so (5,11)(5, 11) is a member of the solution set of y=(x4)(x+6)y = (x - 4)(x + 6) and is on the curve.

Substituting 20-20 for yy in the left-hand side gives 20-20, and substituting 2-2 for xx in the right-hand side gives (24)(2+6)=24(-2 - 4)(-2 + 6) = -24. So the left-hand side is greater than the right-hand side, and (2,20)(-2, -20) is not a point on the curve: it lies above it.

Example 3.21.

Draw a sketch to show the region of the xyxy plane defined by the inequalities 0x20 \leqslant x \leqslant 2, y0y \geqslant 0, yx2y \leqslant x^2.

The relationship 0x20 \leqslant x \leqslant 2 contains two inequalities, x0x \geqslant 0 and x2x \leqslant 2, which must be considered separately. Taking each inequality in turn, we shade out the region that it excludes.

  1. y0y \geqslant 0 is the line y=0y = 0 and the region above the xx-axis, so we shade out the region below the xx-axis.
  2. yx2y \leqslant x^2 is the curve y=x2y = x^2 and the region below it, so we shade out the region above the curve.
  3. x0x \geqslant 0 is the yy-axis and the region to its right, so we shade out the region to the left of the yy-axis.
  4. x2x \leqslant 2 is the line x=2x = 2 and the region to its left, so we shade out the region to the right of this line.

Combining these four, the unshaded region, including the boundary lines, is the set of points that satisfies all the given inequalities.

121234xyy = x²x = 2
Figure 3.22. The region 0x20 \leqslant x \leqslant 2, y0y \geqslant 0, yx2y \leqslant x^2. Here the wanted region is shaded, rather than the regions excluded.

The Equation of a Particular Curve

So far in our work on functions and graphs we have begun with a function and deduced from its properties the curve that represents it. The reverse process, in which we begin with a curve given geometrically and deduce its equation, is as follows.

Consider the circle whose centre is the point C(4,2)C(4, 2) and whose radius is 22. Any point PP on the circumference of this circle is such that PC=2PC = 2; any point QQ inside the circle satisfies CQ<2CQ < 2; and any point RR outside the circle satisfies CR>2CR > 2. The distance of any point (x,y)(x, y) from CC is (x4)2+(y2)2\sqrt{(x - 4)^2 + (y - 2)^2}, so the coordinates (x,y)(x, y) of PP must satisfy the equation

(x4)2+(y2)2=4.(x - 4)^2 + (y - 2)^2 = 4 .
2461234xyCPQR
Figure 3.23. The circle with centre C(4,2)C(4, 2) and radius 22, with PP on it, QQ inside and RR outside.

This equation defines the set of points on the circumference of the circle, and so is the equation of the circle. Similarly the coordinates of QQ satisfy the inequality (x4)2+(y2)2<4(x - 4)^2 + (y - 2)^2 < 4, so this inequality defines the region inside the circle, and the inequality (x4)2+(y2)2>4(x - 4)^2 + (y - 2)^2 > 4 defines the region outside it.

The Straight Line

Straight lines play an important part in any geometric analysis. A straight line may be defined in many ways, for example as

  1. the line which passes through the origin and has a gradient of 12\tfrac12, or
  2. the line which passes through the points (2,1)(2, 1) and (4,2)(-4, -2).

For the first, if P(x,y)P(x, y) is a point on the line other than the origin, then the gradient of OPOP is 12\tfrac12. The gradient of OPOP is y0x0=yx\dfrac{y - 0}{x - 0} = \dfrac{y}{x}, so the coordinates of PP satisfy

yx=12,or2y=x,\frac{y}{x} = \frac12, \qquad\text{or}\qquad 2y = x,

and 2y=x2y = x is the equation of the line.

Remark.

For any point Q(x,y)Q(x, y) above PP, y>12xy > \tfrac12 x, that is 2y>x2y > x. Therefore the inequality 2y>x2y > x defines the region above the line, and similarly 2y<x2y < x defines the region below it.

For the second, a point P(x,y)P(x, y) is on the line through A(2,1)A(2, 1) and B(4,2)B(-4, -2) exactly when the gradient of PAPA equals the gradient of ABAB. The gradient of PAPA is y1x2\dfrac{y - 1}{x - 2} and the gradient of ABAB is 1(2)2(4)=12\dfrac{1 - (-2)}{2 - (-4)} = \dfrac12, so the coordinates of PP satisfy

y1x2=12,or2y=x.\frac{y - 1}{x - 2} = \frac12, \qquad\text{or}\qquad 2y = x .

These apparently different definitions give the same line. It is conventional to use integers for coefficients whenever possible.

Consider the more general case of the line whose gradient is mm and which passes through the origin. For a point P(x,y)P(x, y) on this line other than the origin, the gradient of OPOP is mm, so the coordinates of PP satisfy yx=m\dfrac{y}{x} = m, or y=mxy = mx; the final equation also includes the origin.

Generalising even further to cover any straight line, consider the line whose gradient is mm and which cuts the yy-axis at a directed distance cc from the origin. The number cc is called the intercept on the yy-axis. With AA the point (0,c)(0, c), a point P(x,y)P(x, y) is on the line exactly when the gradient of APAP is mm, so the coordinates of PP satisfy

ycx0=m,ory=mx+c.\frac{y - c}{x - 0} = m, \qquad\text{or}\qquad y = mx + c .
xycP(x, y)gradient m
Figure 3.24. The line y=mx+cy = mx + c, with intercept cc on the yy-axis and gradient mm.

This is called the standard form of the equation of a straight line. It follows that

  1. an equation of the form y=mx+cy = mx + c represents a straight line with gradient mm and intercept cc on the yy-axis;
  2. any equation involving a linear relationship between xx and yy, that is ax+by+c=0ax + by + c = 0 where aa and bb are constants not both zero, is the equation of a straight line.

Example 3.22.

Write down the gradient of the line 3x4y+2=03x - 4y + 2 = 0, and find the equation of the line through the origin which is perpendicular to the given line.

Writing 3x4y+2=03x - 4y + 2 = 0 in standard form gives y=34x+12y = \tfrac34 x + \tfrac12, so the gradient of the given line is 34\tfrac34. So the gradient of the perpendicular line is 43-\tfrac43. The required line passes through the origin, that is it has zero intercept on the yy-axis, so its equation is

y=43x,that is4x+3y=0.y = -\tfrac43 x, \qquad\text{that is}\qquad 4x + 3y = 0 .

Example 3.23.

Sketch the line x2y+3=0x - 2y + 3 = 0.

This line can be located accurately in the xyxy plane once two points on the line are known. The intercepts on the axes can be found by inspection: x=0x = 0 gives y=32y = \tfrac32, and y=0y = 0 gives x=3x = -3. So the line passes through (0,32)\left(0, \tfrac32\right) and (3,0)(-3, 0).

The Line with Gradient mm Through the Point (x1,y1)(x_1, y_1)

If P(x,y)P(x, y) is any point on the line with gradient mm passing through A(x1,y1)A(x_1, y_1), then the gradient of APAP is mm. Therefore the coordinates of PP satisfy

yy1xx1=m,that isyy1=m(xx1).\frac{y - y_1}{x - x_1} = m, \qquad\text{that is}\qquad y - y_1 = m(x - x_1) .

Example 3.24.

Find the equation of the line with gradient 13-\tfrac13 passing through (2,1)(2, -1).

Substituting 13-\tfrac13 for mm, 22 for x1x_1 and 1-1 for y1y_1 gives the equation of the line as

y(1)=13(x2),that isx+3y+1=0.y - (-1) = -\tfrac13(x - 2), \qquad\text{that is}\qquad x + 3y + 1 = 0 .

Alternatively, as any straight line has an equation y=mx+cy = mx + c, the equation of this line can be written y=13x+cy = -\tfrac13 x + c. As the point (2,1)(2, -1) lies on the line, its coordinates satisfy the equation, so 1=23+c-1 = -\tfrac23 + c and c=13c = -\tfrac13. Therefore the equation is y=13x13y = -\tfrac13 x - \tfrac13, or x+3y+1=0x + 3y + 1 = 0.

Remark.

The worked examples necessarily contain a lot of explanation, but this should not mislead the reader into thinking that solutions need be equally long. The temptation to overwork a problem should be avoided, particularly in coordinate geometry, where problems are basically simple. With a little practice, either of the methods above gives the equation of a line directly.

The Line Through (x1,y1)(x_1, y_1) and (x2,y2)(x_2, y_2)

When x1x2x_1\neq x_2, the gradient of the line through (x1,y1)(x_1, y_1) and (x2,y2)(x_2, y_2) is y2y1x2x1\dfrac{y_2 - y_1}{x_2 - x_1}, so by the previous section its equation is

yy1=y2y1x2x1(xx1).y - y_1 = \frac{y_2 - y_1}{x_2 - x_1}(x - x_1) .

For example, the line through (1,2)(1, -2) and (3,5)(3, 5) has equation

y+2=5(2)31(x1),that is7x2y11=0.y + 2 = \frac{5 - (-2)}{3 - 1}(x - 1), \qquad\text{that is}\qquad 7x - 2y - 11 = 0 .

Intersection

If two curves cut at a point AA, then AA is called a point of intersection of the curves. The coordinates of AA satisfy both equations, so AA can be found by solving the equations simultaneously.

Example 3.25.

Find the point of intersection of the lines y3x+1=0y - 3x + 1 = 0 and y+x2=0y + x - 2 = 0.

Subtracting the first equation from the second gives 4x3=04x - 3 = 0, so x=34x = \tfrac34, and then y=2x=54y = 2 - x = \tfrac54. Therefore (34,54)\left(\tfrac34, \tfrac54\right) is the point of intersection.

Example 3.26.

Find the points of intersection AA and BB of the circle x2+y23x+2=0x^2 + y^2 - 3x + 2 = 0 and the line y=x1y = x - 1.

The coordinates of both AA and BB satisfy both equations. Solving them simultaneously by substituting x1x - 1 for yy in the equation of the circle,

x2+(x1)23x+2=0,that is2x25x+3=0,that is(2x3)(x1)=0,x^2 + (x - 1)^2 - 3x + 2 = 0, \qquad\text{that is}\qquad 2x^2 - 5x + 3 = 0, \qquad\text{that is}\qquad (2x - 3)(x - 1) = 0,

so x=32x = \tfrac32 or x=1x = 1. Substituting these in y=x1y = x - 1 gives y=12y = \tfrac12 and y=0y = 0. Therefore AA and BB are the points (32,12)\left(\tfrac32, \tfrac12\right) and (1,0)(1, 0).

In general, the coordinates of the points of intersection of two curves y=f(x)y = f(x) and y=g(x)y = g(x) can be found from the simultaneous solution of the equations y=f(x)y = f(x) and y=g(x)y = g(x).

Example 3.27.

Find the equation of the line through (1,2)(1, 2) which is perpendicular to the line 3x7y+2=03x - 7y + 2 = 0.

Writing 3x7y+2=03x - 7y + 2 = 0 in standard form gives y=37x+27y = \tfrac37 x + \tfrac27, showing that the given line has a gradient of 37\tfrac37. So the required line has gradient 73-\tfrac73 and passes through (1,2)(1, 2). Using yy1=m(xx1)y - y_1 = m(x - x_1) gives its equation as

y2=73(x1),that is7x+3y13=0.y - 2 = -\tfrac73(x - 1), \qquad\text{that is}\qquad 7x + 3y - 13 = 0 .

Note that the line perpendicular to 3x7y+2=03x - 7y + 2 = 0 has an equation 7x+3y13=07x + 3y - 13 = 0: the coefficients of xx and yy have been transposed, and the sign between the xx and yy terms has changed. In fact, given the line ax+by+c=0ax + by + c = 0, any line perpendicular to it has an equation

bxay+k=0,bx - ay + k = 0,

and this property of perpendicular lines can be used to shorten the working of problems.

Example 3.28 (The circumcentre of a triangle).

AA, BB and CC are the points (0,4)(0, 4), (2,3)(2, 3) and (2,1)(-2, -1). Find the circumcentre of triangle ABCABC.

The circumcentre of a triangle is the point of intersection of the perpendicular bisectors of its sides.

ACAC has gradient 4(1)0(2)=52\dfrac{4 - (-1)}{0 - (-2)} = \dfrac52, and its midpoint is (1,32)\left(-1, \tfrac32\right). Therefore the perpendicular bisector of ACAC has gradient 25-\tfrac25 and passes through (1,32)\left(-1, \tfrac32\right), so its equation is

y32=25(x+1),that is4x+10y11=0.(1)y - \tfrac32 = -\tfrac25(x + 1), \qquad\text{that is}\qquad 4x + 10y - 11 = 0 . \tag{1}

Similarly the gradient of ABAB is 3420=12\dfrac{3 - 4}{2 - 0} = -\dfrac12, and its midpoint is (1,72)\left(1, \tfrac72\right). Therefore the perpendicular bisector of ABAB has gradient 22 and passes through (1,72)\left(1, \tfrac72\right), so its equation is

y72=2(x1),that is4x2y+3=0.(2)y - \tfrac72 = 2(x - 1), \qquad\text{that is}\qquad 4x - 2y + 3 = 0 . \tag{2}

Subtracting (2)(2) from (1)(1) gives 12y14=012y - 14 = 0, so y=76y = \tfrac76, and then 4x=2y3=234x = 2y - 3 = -\tfrac23 gives x=16x = -\tfrac16. Therefore the circumcentre of triangle ABCABC is the point (16,76)\left(-\tfrac16, \tfrac76\right).

Problem 3.10.

  1. Find the equation of the line passing through (1,3)(-1, 3) and (4,3)(-4, -3).
  2. Find the equation of the line through (1,2)(1, -2) perpendicular to 2x3y+6=02x - 3y + 6 = 0.
  3. Draw a sketch showing the region of the xyxy plane defined by y<2y < 2 and y>(x2)(x+2)y > (x - 2)(x + 2).

Remark (Summary).

If AA and BB are the points (x1,y1)(x_1, y_1) and (x2,y2)(x_2, y_2), then

  1. the length of ABAB is (x2x1)2+(y2y1)2\sqrt{(x_2 - x_1)^2 + (y_2 - y_1)^2};
  2. the midpoint of ABAB is the point (12(x1+x2),12(y1+y2))\left(\tfrac12(x_1 + x_2), \tfrac12(y_1 + y_2)\right);
  3. when x1x2x_1\neq x_2, the gradient of ABAB is y2y1x2x1\dfrac{y_2 - y_1}{x_2 - x_1}. A vertical line has no finite gradient and has an equation of the form x=constantx=\text{constant}.

The equation y=mx+cy = mx + c defines the straight line with gradient mm and intercept cc on the yy-axis. The inequality y>mx+cy > mx + c defines the region of the xyxy plane above that line, and y<mx+cy < mx + c the region below it.

If lines l1l_1 and l2l_2 have equations y=m1x+c1y = m_1 x + c_1 and y=m2x+c2y = m_2 x + c_2, then l1l_1 and l2l_2 are parallel if m1=m2m_1 = m_2, and, when both gradients are finite and non-zero, perpendicular if m1m2=1m_1 m_2 = -1. A horizontal line is perpendicular to a vertical line. The equation of any line perpendicular to ax+by+c=0ax + by + c = 0 is of the form bxay+k=0bx - ay + k = 0.

Exercises

Questions marked with an examining board are taken from past A-level papers: JMB is the Joint Matriculation Board, U of L the University of London, C Cambridge, and AEB the Associated Examining Board.

Exercise 3.1.

In Utopia, assume income is non-negative. Income tax on earnings is calculated as follows: the first £10,000 is tax free, the next £10,000 is taxed at 5%5\%, and the remaining income is taxed at 10%10\%.

  1. Taking income as input and tax payable as output, state whether these rules for calculating tax constitute a function. If they do, state the implied domain and range.
  2. If £II is income and £TT is the tax payable, express the mapping ITI \mapsto T as formulae of the form T=f(I)T = f(I), stating the values of II for which each is valid.

Exercise 3.2.

State the inverse function, with its domain, of each of the following functions, or explain why it has none.

  1. f:x12x3f : x \mapsto \tfrac12 x - 3, xRx \in \RR;
  2. f:x(x1)(x3)f : x \mapsto (x - 1)(x - 3), xRx \in \RR, x2x \geqslant 2;
  3. f:x(x2)(x+2)f : x \mapsto (x - 2)(x + 2), xRx \in \RR;
  4. f:x10xf : x \mapsto 10^x, xRx \in \RR. Draw sketch graphs of ff and f1f^{-1} on the same set of axes, and describe how one graph can be obtained from the other.

Exercise 3.3.

Let f(x)=11xf(x) = \dfrac{1}{1 - x}.

  1. For what value of xx is f(x)f(x) undefined? Describe the behaviour of f(x)f(x) as xx approaches this value from above and from below.
  2. Write down limxf(x)\displaystyle\lim_{x \to \infty} f(x) and limxf(x)\displaystyle\lim_{x \to -\infty} f(x).
  3. Use this information to sketch the graph of ff, marking the asymptotes clearly.

Exercise 3.4.

Find the ranges of values of kk for which the equation x2+(k3)x+k=0x^2 + (k - 3)x + k = 0 has

  1. real distinct roots;
  2. roots of the same sign. (JMB)

Exercise 3.5.

If xx is real and x2+(2k)x+12k=0x^2 + (2 - k)x + 1 - 2k = 0, show that kk cannot lie between certain limits, and find these limits. (JMB)

Exercise 3.6.

Show that, if x2>k(x+1)x^2 > k(x + 1) for all real xx, then 4<k<0-4 < k < 0. (C)

Exercise 3.7.

Find the condition that must be satisfied by kk in order that the expression 2x2+6x+1+k(x2+2)2x^2 + 6x + 1 + k(x^2 + 2) may be positive for all real values of xx. (JMB)

Exercise 3.8.

  1. If x=2x = 2 is a root of the equation ax2+2(2a5)x+8=0ax^2 + 2(2a - 5)x + 8 = 0, find the possible value, or values, of aa and the corresponding value, or values, of the other root.
  2. Find the range, or ranges, of possible values of the real number aa if ax2+2(2a5)x+8>0ax^2 + 2(2a - 5)x + 8 > 0 for all real values of xx. (C)

Exercise 3.9.

Determine, for each of the expressions f(x)=x2+4x6f(x) = x^2 + 4x - 6 and g(x)=x28x+2g(x) = -x^2 - 8x + 2, the range, or ranges, of values of xx for which it is positive. Give your answers correct to two places of decimals, and explain briefly the reasons for your answers. (C)

Exercise 3.10.

  1. State the range of values of xx for which 2x2+5x122x^2 + 5x - 12 is negative.
  2. The value of the constant aa is such that the quadratic function f(x)=x2+4x+a+3f(x) = x^2 + 4x + a + 3 is never negative. Determine the nature of the roots of the equation af(x)=(x+2)(a1)a f(x) = (x + 2)(a - 1), and deduce the value of aa for which this equation has equal roots. (AEB, 1973)

Exercise 3.11.

By eliminating xx and yy from the equations

1x+1y=1,x+y=a,yx=m,\frac{1}{x} + \frac{1}{y} = 1, \qquad x + y = a, \qquad \frac{y}{x} = m,

where a0a \neq 0, obtain a relation between mm and aa. Given that aa is real, determine the ranges of values of aa for which mm is real. (JMB)

Exercise 3.12.

If a>0a > 0, show that the quadratic expression ax2+bx+cax^2 + bx + c is positive for all real values of xx when b2<4acb^2 < 4ac. Hence find the range of values of pp for which the quadratic function

f(x)=4x2+4px(3p2+4p3)f(x) = 4x^2 + 4px - (3p^2 + 4p - 3)

is positive for all real values of xx. Illustrate your result by making sketch graphs of f(x)f(x) for each of the cases p=0p = 0 and p=1p = 1. (U of L)

Exercise 3.13.

  1. If aa is a positive constant, find the set of values of xx for which a(x2+2x8)a(x^2 + 2x - 8) is negative. Find the value of aa if this function has a least value of 27-27.
  2. Find two quadratic functions of xx which are zero at x=1x = 1, which take the value 1010 when x=0x = 0, and which have a greatest value of 1818. Sketch the graphs of these two functions. (U of L)

Exercise 3.14.

Find the set of values of kk for which f(x)=3x25x+kf(x) = 3x^2 - 5x + k is greater than unity for all real values of xx. Show that, for all kk, the least value of f(x)f(x) occurs when x=56x = \tfrac56, and find kk if this least value is zero. (U of L)

Exercise 3.15.

The roots of the equation 9x2+6x+1=4kx9x^2 + 6x + 1 = 4kx, where kk is a real constant, are denoted by α\alpha and β\beta.

  1. Show that the equation whose roots are 1α\dfrac{1}{\alpha} and 1β\dfrac{1}{\beta} is x2+6x+9=4kxx^2 + 6x + 9 = 4kx.
  2. Find the set of values of kk for which α\alpha and β\beta are real.
  3. Find also the set of values of kk for which α\alpha and β\beta are real and positive. (U of L)

Exercise 3.16.

  1. The points A(1,3)A(1, 3), B(5,7)B(5, 7), C(4,8)C(4, 8) and D(a,b)D(a, b) form a rectangle ABCDABCD. Find aa and bb.
  2. The points A(1,5)A(1, 5), B(4,1)B(4, -1) and C(2,4)C(-2, -4) form triangle ABCABC. Show that the triangle is right-angled, and find its area.

Exercise 3.17.

ABCDABCD is a quadrilateral, where AA, BB, CC and DD are the points (3,1)(3, -1), (6,0)(6, 0), (7,3)(7, 3) and (4,2)(4, 2). Show that the diagonals bisect each other at right angles, and hence find the area of ABCDABCD.

Exercise 3.18.

A circle of radius two units, with its centre at the origin, cuts the xx-axis at AA and BB and cuts the positive yy-axis at CC. Show that ABAB subtends a right angle at CC. If D(a,b)D(a, b) is a point on the circumference of the circle, find a relationship between aa and bb.

Exercise 3.19.

A point P(a,b)P(a, b) is equidistant from the yy-axis and from the point (4,0)(4, 0). Find a relationship between aa and bb.

Exercise 3.20.

Find the equation of the perpendicular from the point A(5,3)A(5, 3) to the line 2xy+4=02x - y + 4 = 0. Hence find the distance of AA from the line.

Exercise 3.21.

The equation of a circle is (x1)2+(y1)2=4(x - 1)^2 + (y - 1)^2 = 4.

  1. Find the coordinates of AA and BB, the points of intersection of the line x+y=2x + y = 2 and the circle.
  2. Show that the point C(1,3)C(1, 3) is on the circumference of the circle. Find the midpoint MM of ABAB, and show that MC=MA=MBMC = MA = MB.
  3. What can you deduce about the line ABAB?

Exercise 3.22.

A line is drawn through the point A(1,2)A(1, 2) to cut the line 2y=3x52y = 3x - 5 at PP and the line x+y=12x + y = 12 at QQ, with PP between AA and QQ. If AQ=2APAQ = 2AP, find the coordinates of PP and QQ. (U of L)

Check Yourself

 

Fresh questions on the whole chapter — none of them is worked out above. Do each on paper first; the box only tells you whether you got there.

Answers are checked in your browser, as often as you like. Nothing is sent anywhere and nothing is kept but your own work. A formula may be written with the symbols themselves or with ~ & | -> <-> ^, and \and, \or, \to expand as you type.

Exercise 3.23.

Each mapping below has domain xRx \in \RR. Which one is one-one?

answer one of these

Exercise 3.24.

What is the range of f(x)=(x3)2+2f(x) = (x - 3)^2 + 2, xRx \in \RR?

answer one of these

Exercise 3.25.

What is the axis of symmetry of the curve y=x26x+1y = x^2 - 6x + 1?

answer one of these

Exercise 3.26.

What is the greatest value of 5(x+1)25 - (x + 1)^2?

answer one of these

Exercise 3.27.

Which values of xx satisfy 2x<6-2x < 6?

answer one of these

Exercise 3.28.

Which values of xx satisfy (x+1)(x3)<0(x + 1)(x - 3) < 0?

answer one of these

Exercise 3.29.

Which line is an asymptote to the curve y=3xy = 3^x?

answer one of these

Exercise 3.30.

Where does the curve y=1x2y = \dfrac{1}{x - 2} have a discontinuity?

answer one of these

Exercise 3.31.

How is the curve y=3x2y = 3 - x^2 obtained from the curve y=x2y = x^2?

answer one of these

Exercise 3.32.

What is the inverse of f(x)=3x2f(x) = 3x - 2, xRx \in \RR?

answer one of these

Exercise 3.33.

What is the length of the line joining (3,4)(3, -4) to (7,2)(-7, 2)?

answer one of these

Exercise 3.34.

What is the midpoint of the line joining (1,3)(-1, -3) to (3,5)(3, -5)?

answer one of these

Exercise 3.35.

What is the gradient of a line perpendicular to the line joining (1,5)(-1, 5) and (2,3)(2, -3)?

answer one of these

Exercise 3.36.

What is the equation of the line through the origin perpendicular to 3x2y+4=03x - 2y + 4 = 0?

answer one of these

Exercise 3.37.

What is the equation of the line with gradient 11 passing through the point (h,k)(h, k)?

answer one of these

Exercise 3.38.

At which points do the curves y=x2y = x^2 and y=x(2x)y = x(2 - x) intersect?

answer one of these