The University of Sydney

MOOC Introduction to Calculus

Notes for ‘Curve sketching’

Important Ideas and Useful Facts:

(i) Curve sketching, general principles: Curve sketching involves drawing or sketching (often
in one’s mind’s eye) the curve corresponding to the graph of a function y = f (x) in the
xy-plane. In the sketch,

– the scales on the x and y-axes do not need to be the same;

– positioning of axes is flexible and chosen to highlight or enhance the main features
of the curve;
– aim for simplicity and clarity, and avoid including any unnecessary details;
– focus on the main qualitative features;
– extra details such as coordinates of important points can be added according to the
particular problem one is trying to solve.

(ii) Curve sketching, checklist: The following is a natural checklist of features to look for and
include if possible when sketching the graph of a function y = f (x):

– find the y-intercept, f (0), where the curve crosses the y-axis, assuming x = 0 is in
the domain of the function;
– find any straightforward x-intercepts, where f (x) = 0, that is, where the curve crosses
the x-axis (but be aware that typically this may not be easy, and indeed the purpose
of creating a sketch may be to locate the approximate positions of x-intercepts, to
assist in finding solutions or approximate solutions to a difficult equation);
– look for asymptotic behaviour, seeing what happens as x → ∞ or x → −∞, leading
to possible horizontal or oblique asymptotes, and whether there are any vertical
asymptotes, near which y → ∞ or y → −∞;
– look for where the curve is increasing (y ′ > 0) or decreasing (y ′ < 0), to help
locate turning points and local or global maxima or minima, information for which
is typically stored or summarised in the sign diagram for the derivative;
– look for where the curve is concave up (y ′′ > 0) or concave down (y ′′ < 0) and points
of inflection, where concavity changes, information for which is typically stored or
summarised in the sign diagram for the second derivative.

Examples:

1. Sketch the curve y = f (x) = 2x3 − 9x2 + 12x − 1.

Solution: The y-intercept is y(0) = −1. (Finding the x-intercepts looks difficult, but we
may obtain information about them as a consequence of the rest of this curve sketching
exercise.)

1
The asymptotic behaviour is not very interesting. The cubic is dominated by the 2x3
term, so we have

lim f (x) = ∞ and lim f (x) = −∞ ,

x→∞ x→−∞

which are abbreviations for saying that f (x) gets arbitrarily large and positive as x does,
and f (x) gets arbitrarily large (in magnitude) and negative as x does. In particular, there
are no horizontal asymptotes.
You may be curious to know how we are sure there is no oblique asymptotic behaviour.
Suppose, by way of contradiction that there exists some oblique asymptote with equation

y = mx + c

for some constants m and c. Certainly m can’t be zero, for otherwise we would have a
horizontal asymptote, and we have already seen that this is not the case. This hypothetical
asymptote would have to have the property that the curve y = f (x) gets arbitrarily close
to it as x gets large, which would imply that the ratio of mx + c with f (x) should tend
to 1. But, using our usual tricks for evaluating limits,
m
mx + c mx + c x2
+ xc3
lim = lim = lim
x→∞ f (x) x→∞ 2x − 9x2 + 12x − 1
3 x→∞ 2 − 9
+ x122 − x13
x
0+0
= = 0,
2−0+0−0
which of course does not equal 1. A similar contradiction is reached by taking the limit
as x → −∞, so there is no chance of there being an oblique asymptote.
We have
y ′ = 6x2 − 18x + 12 = 6(x2 − 3x + 2) = 6(x − 1)(x − 2) ,
so that y ′ = 0 when x = 1, 2. Further

y ′′ = 12x − 18 = 6(2x − 3) ,

so y ′′ = 0 when x = 3
2
. We get the following sign diagrams for the first and second
derivatives:

3
x 1 2 x 2

y′ + 0 − 0 + y ′′ − 0 +

The diagram for y ′ indicates that there is a turning point at x = 1, with local maximum

f (1) = 2 − 9 + 12 − 1 = 4 ,

and a turning point at x = 2, with local minimum

f (2) = 16 − 36 + 24 − 1 = 3 .

2
 The diagram for y ′′ indicates that there is a point of inflection at x = 32 , and the corre-
sponding y-value is
( ) ( )3 ( )2 ( )
3 3 3 3 7
f = 2 −9 + 12 −1 = .
2 2 2 2 2
All of these important features are highlighted in the sketch below. Notice that, in fact,
there is only one x-intercept, something that was not at all evident from the rule of the
function, and it is forced to lie somewhere between 0 and 1, and the sketch suggests quite
close to 0.

y = f (x)
4
7
2
3

3
1 2
−1 2

x2 − 2
2. Sketch the graph of the rational function y = g(x) = .
x−1
Solution: The y-intercept is g(0) = 2. Weyhave
= −xconsidered
2 this curve before and exploited
the fact that the rule can be expressed in different ways:
√ √
x2 − 2 (x + 2)(x − 2) 1 1
g(x) = = = x+1− = x+1+ .
x−1 x−1 x−1 1−x
√
The x-intercepts are clearly ± 2. The line x = 1 is a vertical asymptote, and the line
1
y = x + 1 is an oblique asymptote. Put h(x) = 1−x . We have also seen before that the
derivative is
1
g ′ (x) = 1 + h′ (x) = 1 +
(1 − x)2
with the following sign diagram:

x 1
g ′ (x) + u +

3
The derivative can also be put in the following form:
1
g ′ (x) = 1 + h′ (x) = 1 + = 1 + (1 − x)−2 = 1 + (x − 1)−2 .
(1 − x)2
The second derivative therefore becomes
d d( ) d( )
g ′′ (x) = (1) + (x − 1)−2 = (x − 1)−2 .
dx dx dx
d
We have been told that dx (xn ) = nxn−1 for any exponent n, and it follows that
d( )
(x − 1) n
= n(x − 1)n−1 ,
dx
since the curve y = (x − 1) is obtained from the curve y = xn by a horizontal shift one
n

unit to the right, so the slope of tangent lines is unaffected. We can apply this fact in our
case, when n = −2, to conclude that
d( ) −2
g ′′ (x) = (x − 1)−2 = (−2)(x − 1)−3 = .
dx (x − 1)3
It follows that g ′′ (1) is undefined and the sign of g ′′ (x) is the opposite of the sign of x − 1,
giving the following sign diagram:

x 1
y ′′ + u −

This pattern of concave up, followed by concave down, is matched by the concavity
behaviour in the sketch below.

y = g(x)

√ 1
− 2
−1 √
1 2

y =x+1
oblique asymptote

4
3. As a (difficult) variation of the previous example, sketch the graph of the rational function
1 x2
y = f (x) = x + 1 + = .
x−1 x−1
Solution: The y-intercept is f (0) = 0 and y = 0 when x = 0, so there is only one x-
intercept and the curve passes through the origin. As in the previous example, the line
x = 1 is a vertical asymptote and the line y = x + 1 is an oblique asymptote. The
derivative becomes, similar to the previous example in the first three steps,
( )
′ d d 1 d( −1
)
f (x) = (x + 1) + = 1+ (x − 1) = 1 − (x − 1)−2
dx dx x − 1 dx
1 (x − 1) − 1
2
x2 − 2x + 1 − 1 x(x − 2)
= 1− = = = .
(x − 1)2 (x − 1) 2 (x − 1)2 (x − 1)2
Differentiating again, we get
′′ d d( −2
) 2
f (x) = (1) − (x − 1) = −(−2)(x − 1)−3 = .
dx dx (x − 1)3
From these formulae for f ′ (x) and f ′′ (x) we get the following sign diagrams:

x 0 1 2 x 1
f ′ (x) + 0 − u − 0 + f ′′ (x) − u +

Thus there is a turning point at x = 0, with local maximum y = 0, and a turning point
at x = 2, with local minimum y = 4. The curve is concave down for x < 1 and concave
up for x > 1, and we get the following sketch:

y = f (x)

−1 1 2
y =x+1
oblique asymptote