Mathematics Revision Notes 2009

RI(JC) Mathematics Society
Pure Mathematics Page
Functions 3
Graphing Techniques 8
Equations & Inequalities 16
Arithmetic & Geometric Progressions 20
Summation, Recurrence & Induction 25
Differentiation 32
Integration 43
Differential Equations 54
Power Series 61
Binomial Expansion 66
Vectors 72
Complex Numbers 84
Statistics Page
Permutations & Combinations 99
Probability 103
Distributions 109
Sampling & Testing 118
Correlation & Regression 124
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
2
Acknowledgements
- Members of RI(JC) Mathematics Society, batch of 2009, for their time and effort in
creating these notes;
o Bi Ran
o Hu Yi Jie
o Kan Wai Keong Desmond (Editor)
o Koo Cheng Wai (Editor)
o Kung Ying Xiang
o Liew Xiao Tian
o Ong Chin Hwee
o Samuel Lim Yong Peng
o Tan Chor Hiang
o Tan Jun Xian Gabriel
o Tan Wei Jie
- Mr. Kok Chee Kean, teacher-in-charge, for his guidance and support;
- Members of RJC Mathematics Society, batch of 2008, for pioneering this project;
- and you, the reader, for supporting our work.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
3
Functions
A function f is a relation which maps each element in its domain
f
D to one and only one
element in its range
f
R , e.g.
2
: 1 f x x + ÷ , xe¯, where
2
1 x + is the rule of f and ¯ is
the domain of f . Use the Vertical Line Test to determine if a given relation f is a function.
Replace
f
D and
f
R below with the actual domain and range of f in interval notation.
One-one Functions
Use the Horizontal Line Test to determine if a given function is one-one.
e.g.
2
: 1 f x x + ÷ , xe¯ and
2
: 1 g x x + ÷ , 0 x > .
Since the horizontal line 2 y = , where
2 [1, )
f
R e = · , does not cut the graph of f
at one and only one point, f is not one-one.
OR
Since . ) . ) 1 1 2 f f ÷ = = ,
where 1,1
f
D ÷ e =  , f is not one-one.
Since any horizontal line y k = , where
| ) 1,
g
k R e = · , cuts the graph of g at one
and only one point, g is one-one.
g is a restriction of f if f and g have the same rule and
g f
D D _ , i.e. g is f with its
domain restricted. Even if f is not one-one, a restriction of f which is one-one can be
defined. Use turning points or points where f is undefined to show that f is not one-one
or restrict its domain.
y
( ) y f x =
2 y =
x
0
1
2
any line y k = ,
[1, )
g
k R e = ·
( ) y g x =
y
0
1
k
x
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
4
Inverse Functions
1
f
÷
, the inverse function of f , is a function such that for all
f
x D e and
f
y R e ,
. )
1
f y x
÷
= if and only if . ) f x y = .
-
1
f
÷
exists if and only if f is one-one.
- Domain of
1
f
f R
÷
=
- Range of
1
f
f D
÷
=
To find the rule of
1
f

given the rule of f :
1. Let . ) y f x = .
2. Express x in terms of y , such that . ) x g y = , where . ) g y is a function of y .
3. Then . ) . )
1
f x g x
÷
= .
The rules of
1
ff
÷
and
1
f f
÷
are always x , but they usually have different domains.
1 1
ff f
D D
÷ ÷
= , but
1
f
f f
D D
÷
= .
Composite Functions
gf is the composite function of f followed by g .
- gf exists if and only if
f g
R D _ .
- Domain of
f
gf D =
- Range of
g
gf R = , with the domain of g restricted to
f
R .
FAQ
Q: Sketch the graph of f .
A: Substitute the rule of f , then sketch the graph using GC.
Q: What is the domain/range of f ?
A: Observe from the rule of f by substituting values, or from the graph using GC.
Q: What is the minimum value of k such that f , with its domain restricted to ( , ) k  ,
is one-one?
A: Find the minimum value of k such that f , with its domain restricted, passes the
Horizontal Line Test. Usually, x k = will be a turning point or point where f is
undefined on the graph of f .
Q:
Solve . ) . )
1 ÷
= f x f x .
A:
Solve . ) f x x = or . )
1
f x x
÷
= instead, since the graphs of f and
1
f
÷
must intersect on
the line y x = . (Example 2)
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
5
Q:
Solve . ) . )
1 1 ÷ ÷
= ff x f f x .
A:
The solution set is the intersection of the domains of f and
1
f
÷
, since both
1
ff
÷
and
1
f f
÷
have rule x .
Q:
Sketch the graph of
1 ÷
ff (or
1 ÷
f f ).
A:
The graph of
1
ff
÷
is the section of the line y x = , where
1
f
f
x D R
÷
e = . Similarly, the
graph of
1
f f
÷
is the section of the line y x = , where
f
x D e .
Q:
Sketch the graphs of ( ) f x and
1
( )
÷
f x (on a single diagram).
A:
Show their geometrical relationship: the graphs of f and
1
f
÷
are reflections of each
other about the line y x = .
Example 1 [NJC06/CT/Q9 (Modified)]
A mapping f is given by
2
1
: 1,
( 1)
f x x
x
÷ e
÷
÷  .
State the largest possible domain of f in the form . ) ,b ÷· , where b e , such that the
inverse function of f exists. Hence define
1
f
÷
in a similar form.
The functions g and h are defined as follows:
. ) : ln( 2), 1,1 g x x x + e ÷ ÷
2
: 2 1, h x x x x
+
÷ ÷ e ÷ 
Determine whether the composite function gh exists.
Give the rule and domain of the composite function hg and find its range.
Solution Comments
(a) Largest possible ( ,1)
f
D = ÷· 
1
f
÷
exists if and only if f is one-one.
Observe a discontinuity at 1 x = by
sketching the graph of f on the GC or
from
2
1
( 1) x ÷
. Restrict
f
D based on the
discontinuity instead of looking for
turning points.
. ) 1
1,
f
f
R D
÷
= = ÷ ·
Use the GC to find
f
R .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
6
Let . )
. )
2
1
1
1
y f x
x
= = ÷
÷
, . ) ,1 x e ÷· .
. )
2
1
1
1
y
x
+ =
÷
. )
2 1
1
1
x
y
÷ =
+
1
1
1
x
y
÷ = ±
+
1
1
1
x
y
= ±
+
1
1
1
x
y
= ÷
+
or
1
1
1 y
+
+
(rejected)
Use
f
D to decide which of the 2
expressions for x to accept and which
to reject. Here, . ) ,1
f
D = ÷· , so 1 x < ,
and
1
0
1 y
>
+
for all y e .
1
1
: 1
1
f x
x
÷
÷
+
÷ , . ) 1, x e ÷ · 
Express
1
f
÷
in a similar form, and write
down both its rule and domain.
(b)(i)
. ) . )
2
2
2 1 1 2 h x x x x = ÷ ÷ = ÷ ÷
. ) . ) . ) 1 , 2,
h
R h = · = ÷ ·
( ) h x has a minimum point at 1 x = , and
its domain
+
 includes 1 .
Since . ) 2, ( 1,1 ) ÷ · _ ÷
/
,
h g
R D _
/
, hence
the composite function gh does not exist.

gh exists if and only if
h g
R D _ .
(b)(ii)
. ) 1,1
hg g
D D = ÷ =
. ) . ) . ) . ) . )
ln 2 hg x h g x h x = = +
. ) . ) . )
2
: ln 2 2ln 2 1 hg x x x + ÷ + ÷ ÷ ,
( 1,1) xe ÷ 
You do not need to show hg exists,
since the question already assumes it.
. ) . ) . )
1 , 1 (0, ln3)
g
R g g = ÷ = . ) g x is strictly increasing on ( 1,1) ÷ .
hg h
R R = with
h
D restricted to
g
R
. ) . ) ) | ) 1 , 0 2 1 ,
hg
R h h = = ÷ ÷


In the restricted (0, ln 3)
h g
D R = = ,
( ) h x decreases from 0 to 1 , reaches its
minimum at . ) 1 2 h = ÷ , then increases
from 1 to ln3 . Since . ) 0 (ln3) h h > ,
(0) h is the larger endpoint of
hg
R .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
7
Example 2 [HCI06/Promo/Q11 (Modified)]
The function f is defined by
1
: , 0
1
x
f x x
e
>
+
÷
By considering the domains of
1
ff
÷
and
1
f f
÷
, solve . ) . )
1 1
ff x f f x
÷ ÷
= .
Show the graphs of ( ) f x , . )
1
f x
÷
and . )
1
ff x
÷
on a single sketch.
Solution Comments
(i)
1 1
1
0,
2
f
ff f
D D R
÷ ÷
´ ]
= = =

]
' ]
1
[0, )
f
f f
D D
÷
= = ·
For . )
1 1
( ) ff x f f x
÷ ÷
= ,
| )
1 1
0, 0, 0,
2 2
x
´ ] ´ ]
e · · =

] ]
' ] ' ]
1
0
2
x < < 
The rules of
1
ff
÷
and
1
f f
÷
are always x , but
they may have different domains.
Hence, . ) . )
1 1
ff x f f x
÷ ÷
= if and only if x is
in the domain of both
1
ff
÷
and
1
f f
÷
, i.e.:
1 1
ff f f
f f
R x D D D
÷ ÷
= e · ·
(ii) Sketch the graphs on the GC first. When the
rule of
1
( ) f x
÷
is unknown, key in
DrawInv(
1
Y ) to draw the inverse of the graph
of
1
Y .
GC syntax:
Set
1
1/ (1 ^ ( )) Y e X = + , then:
[2ND][PRGM] 8:DrawInv
[VARS] Y-VARS Function…
1
Y
[ENTER]
0
1
2
y
x 1
2
1 1
,
2 2
´ }
]
' ¹
y x =
1
( ) y f x
÷
=
( ) y f x =
1
( ) y ff x
÷
=
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
8
Graphing Techniques
Graphs of Conic Sections
The dotted lines represent lines of symmetry or asymptotes, not the axes.
Ellipse
Circle
Parabola
Hyperbola
r
( , ) h k
y k =
x h =
2 2 2
( ) ( ) y r x k h    
y k =
2
( ) y b x h k   
( , ) h k
( , ) h k
2
( ) x b y k h   
x h =
a
b
( , ) h k
y k =
x h =
2 2
2 2
( (
1
) ) y k
b
x h
a




2 2
2 2
( (
1
) ) y k
b
x h
a




2 2
2 2
( (
1
) ) x h
a
y k
b




Mathematics Revision Notes 2009
RI(JC) Mathematics Society
9
Properties of Hyperbolas
For both forms of hyperbolas above,
- Centre: . ) , h k
- Horizontal line of symmetry: x h =
- Vertical line of symmetry: y k =
- Oblique asymptote (sloping upwards): . )
b
y k x h
a
÷ = ÷
- Oblique asymptote (sloping downwards): . )
b
y k x h
a
÷ = ÷ ÷
If the x term comes first in the equation of a hyperbola, i.e.
2 2
2 2
( ) ( )
1
h k
a b
÷ ÷
÷ =
x y
, then the
graph looks like a handwritten x (with left/right parts) and always intersects the x axis.
Sketching Graphs of Conic Sections
- There is a special case for each conic where its centre or vertex is at the origin (i.e. 0 h = ,
0 k = ). To draw a conic with centre/vertex ( , ) h k , translate this special case by replacing
x with x h ÷ and y with y k ÷ .
- If the equation of a conic is given in the general form
2 2
0 Ax By Cx Dy E      ,
complete the square to convert it into the standard forms given in the table above.
- When the equation of a conic is given in general form, to identify the conic:
Circle Ellipse Hyperbola Parabola
Is there both
2
x term and
2
y term?
Are the coefficients of
2
x term
and
2
y term of same sign?
Are the coefficients the same value?
Is there either
2
x term or
2
y term?
Yes
Yes
No
No
Yes
No Yes
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
10
Graphs of Rational Functions
+
=
+
ax b
y
cx d
, where 0 c = , numerator and denominator have no common factor.
- Using long division, express y as
q
p
cx d
+
+
, where p and q are constants.
- Vertical asymptote:
d
x
c
= ÷
o When 0 cx d + = ,
q
cx d +
is undefined, hence y is undefined.
- Horizontal asymptote:
a
y p
c
= =
o As x ÷ ±· , 0
q
cx d
÷
+
, hence y p ÷ .
2
+ +
=
+
ax bx c
y
dx e
, where , 0 a d = , numerator and denominator have no common factor.
- Using long division, express y as
r
px q
dx e
+ +
+
, where p , q , and r are constants.
- Vertical asymptote:
e
x
d
= ÷
o When 0 dx e + = ,
r
dx e +
is undefined, hence y is undefined.
- Oblique asymptote: y px q = +
o As x ÷ ±· , 0
r
dx e
÷
+
, hence y px q ÷ + .
Let the initial graph be of ( ) y f x = , and let a be a positive constant.
Translating Graphs
Translate by a units in the: In the equation Equation after transformation
positive y -direction Replace y by y a ÷
( ) y a f x ÷ =
i.e. . ) y f x a = +
negative y -direction Replace y by y a +
( ) y a f x + =
i.e. . ) y f x a = ÷
positive x -direction Replace x by x a ÷
( ) y f x a = ÷
negative x -direction Replace x by x a + ( ) y f x a = +
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
11
Stretching & Reflecting Graphs
Stretch parallel to the y -axis by
factor a, with x -axis invariant
Replace y by
y
a
( )
y
f x
a
=
i.e. ( ) y af x =
Stretch parallel to the x -axis by
factor a, with y -axis invariant
Replace x by
x
a
x
y f
a
´ }
=
]
' ¹
Reflect in the x -axis ( 0 y = ) Replace y by y ÷
. ) y f x ÷ =
i.e. ( ) y f x = ÷
Reflect in the y -axis ( 0 x = )
Replace x by x ÷
( ) y f x = ÷
Transformations Involving Modulus
Reflect the part of the graph
of = ( ) where ( ) < 0
in the -axis.
Delete the part of the graph
of = ( ) where < 0.
Reflect the part of the graph
of = ( ) where > 0 in
the -axis.
Other Transformations
Assume the same graph of . ) = y f x as above, which is drawn with a dashed line below.
Other transformations include:
- '( ) y f x =
-
1
( )
y
f x
=
-
2
( ) y f x =
o Related to ( ) y f x = or ( ) y f x = ÷
. )
= y f x
y
x
0
y
x
x
0
= ( ) y f x
y
x
0
= ( ) y f x
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
12
Graph
Graph of . ) = y f x
Transformed graph
Stationary point x -intercept
Increasing Positive
Decreasing Negative
Concave upwards Increasing
Concave downwards Decreasing
Maximum gradient Maximum point
Minimum gradient Minimum point
Positive Positive
Negative Negative
Increasing Decreasing
Decreasing Increasing
Approaches · or ÷· Approaches 0
Approaches 0 Approaches · or ÷·
Maximum point Minimum point
Minimum point Maximum point
1. Discard the part of . ) y f x = where . ) 0 f x < .
2. Look at x -intercepts of . ) y f x = to decide the
kind of tangent on the transformed graph.
- 1 distinct real root: vertical tangent
- 2 equal real roots: sharp tangent
- >2 equal real roots: horizontal tangent
3. This is the graph of . ) = y f x .
4. Find the graph of . ) =÷ y f x by reflecting the
graph of . ) y f x = in the x -axis.
5. Combine the graphs of . ) y f x = and
. ) y f x =÷ to get the graph of . )
2
= y f x .
= '( ) y f x y
x
0
y
x
0
1
=
( )
y
f x
y
x
0
2
= ( ) y f x
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
13
Sketching a Transformed Graph
1. Work from the left of the graph ( x = ÷·) to the right ( x = · ).
2. Identify critical features on the original graph.
3. On the new graph, label all corresponding coordinates of points, equations of asymptotes,
stationary points etc. which are given on the original graph.
- For points, write ' , e.g. use ' A to denote point A after transformation.
4. Check your answer.
- Use substitutions, e.g. when ( ) 10 f x = ,
1
0.1
( )
y
f x
= = .
- Use GC to graph normally or use its Conics application.
Sequences of Transformations Involving Modulus
Example: Transform the graph of ( ) y f x = to that of
. )
2 3 y f x = ÷
Solution:
. )
( ) (2 ) (2 ) (2 3 ) 2 3 f x f x f x f x f x ÷ + ÷ ÷ ÷ ÷ ÷ ÷
When x is involved, replace x by x last, to get a graph symmetric about the y -axis.
Example: Transform the graph of ( ) y f x = to that of
. )
2 3 y f x = ÷
Solution:
. ) . ) . ) . )
( ) 2 2 2 3 f x f x f x f x f x ÷ ÷ + ÷ ÷ ÷ ÷
Here 2 3x ÷ is inside the modulus, hence replace x by x first, then transform x to 2 3 x ÷ .
General Tips
- Generally, perform transformations from the innermost to outermost brackets.
- Write the exact phrasing of the transformations when the question asks you to “describe”
them, e.g. write “translate by 3 units…”, not “translate for 3 units…”.
Example [YJC/I/Q8]
The curve C has equation
2
1
, .
2 3
x kx
y k
x
+
+ +
= e
+

(i) Obtain the equations of the asymptotes of C .
(ii) Find the value of k for which the x -axis is tangent to C .
(iii) Sketch the case for 3 k = . Hence, using a graphical method, find the range of values
of x that satisfy the inequality
2
3 1 4 6 x x x + + > + .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
14
Solution Comments
(i)
2
1
2 3
3 13 6
2 2 4 4(2 3)
x kx
y
x
x k k
x
+ +
=
+
÷
= + ÷ +
+
Use long division to get a quotient of
3
2 2 4
x k
+ ÷
and a remainder of
13 6
4
k ÷
.
Equations of asymptotes:
3
2
x = ÷ and
3
2 2 4
x k
y = + ÷ . 
Graphs of
2
ax bx c
y
dx e
+ +
=
+
usually have 2
asymptotes:
1. To find the vertical asymptote, set the x -
term in the denominator 0 = , i.e. 2 3 0 x + = .
2. To find the oblique asymptote, take y ÷±·
such that
13 6
0
4(2 3)
k
x
÷
÷
+
and
3
2 2 4
x k
y ÷ + ÷ .
(ii) Where x -axis is tangent to C ,
2
0 1 0 y x kx = ¬ + + =
Since there can only be
1 value of x where 0 y = ,
. ) . )
2
4 1 1 0 D k = ÷ =
2
4 k =
2 k = or 2 ÷ (rejected k
+
e  )
2 k = 
For a quadratic equation
2
0 ax bx c + + = , 0 a = ,
discriminant
2
4 D b ac = ÷ .
0 D > : 2 distinct real roots.
0 D = : 1 (repeated) real root.
0 D < : No real roots.
Always try using the discriminant with quadratic
equations in Graphing Techniques questions.
Differentiation is tedious and will result in a
complicated expression. Furthermore, if the
question states “algebraically”, differentiation
cannot be used.
Use the GC to plot the graph for 2 k = to check
your answer.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
15
(iii)
When 3 k = ,
2
3 1
2 3
x x
y
x
+ +
=
+
.
2
3 1 4 6 x x x + + > +
2
3 1 2 2 3 x x x + + > +
2
2
3 1
3 1
2
2 3 2 3
x x
x x
x x
+ +
+ +
= >
+ +
2 y >
2 y < ÷ or 2 y >
From the graph,
5.79 x < ÷ or
3
1.79
2
x ÷ < < ÷
or
3
1.21
2
x ÷ < < ÷ or 2.79 x > 
Substitute 3 k = into the equations of the graph and its asymptotes, and sketch it using
the GC. Label all axis intercepts (to 3 s.f.) and asymptotes.
“Graphical method” suggests:
1. Manipulate given inequality until 1 side, say LHS, resembles equation of graph.
2. Sketch graph of RHS on GC and your diagram.
3. The solution usually involves points of intersection of the 2 graphs.
- e.g. the blue marks on the x -axis in the diagram above.
Exclude the values of x for which the function is undefined, e.g.
3
2
x = above.
We can divide both sides of the inequality by 2 3 x + because it is always positive.
3
2
x = ÷
0
0.382 ÷
0.333
x
y 2
3 1
2 3
x x
y
x
+ +
=
+ 3
2 4
x
y = +
2.62 ÷
2 y =
2 y = ÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
16
Equations & Inequalities
General Approach
1. Zero the inequality (bring all the terms to one side, such that the other side is 0).
2. Combine the terms to form a single fraction.
3. Factorize the terms as far as possible.
- Use
. )
. )
. ) . ) 0 0
f x
f x g x
g x
< · < to convert the inequality to a product of factors.
Since . )
2
g x ]
]
is always positive, multiplying it on both sides preserves the
inequality sign.
4. Eliminate factors which are either always positive or always negative.
- Complete the square, e.g. for all xe, . )
2
2
4 5 2 1 1 0 x x x ÷ + = ÷ + > > .
5. Draw number line and graph and shade the regions where the inequality holds.
6. Write the range of values of x (e.g. 1 2 x < < ).
- To write the set of values, use set-builder notation, NOT interval notation.
o e.g. { :1 2} x x e < < ¯ , NOT (1, 2]
- Check for any values (especially endpoints) to be excluded
o e.g. 2 x = when the inequality contains
1
2 x ÷
o e.g. whether the inequality is strict ( <) or not ( <).
Using the GC
- Find intersection between 2 graphs: [2ND][TRACE] 5:intersect
o Find roots of a graph: [2ND][TRACE] 2:zero
- Find stationary points of a graph: [2ND][TRACE] 3:minimum (or 4:maximum)
- Solve system of linear equations: [APPS] 5:PlySmlt2
FAQ
- Solve a given inequality, using algebraic or graphical approaches.
- Using the solution to an original inequality, solve a new inequality.
o Substitute x with some function of x , ( ) f x (usually x ÷ ,
1
x
, ln x ,
x
e , etc.) in
the original inequality, such that it becomes the new inequality. Then, substitute
x with the same function ( ) f x in the original solution set to obtain the new
solution set.
- Model a problem with a system of linear equations and solve the system.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
17
General Tips
- In general, do not cross-multiply inequalities! Cross-multiplying only works if the
denominators on both sides of the inequality are either both positive or both negative.
- Use the GC if the question does not specify “exact” values. But be careful when using the
GC directly: some solution intervals are not visible at certain zoom levels! (Example 1)
- Given an inequality involving modulus, e.g.
. ) f x a < where a is a positive constant,
o Divide it into cases, where . ) 0 f x < and . ) 0 f x > .

. ) . ) f x a a f x a < ¬ ÷ < <

. ) ( ) f x a f x a > ¬ < ÷ or ( ) f x a >
o OR Square both sides to get . )
2
2
f x a ] <
]
, removing the modulus.
- Remember to flip the sign of the inequality when you:
o Multiply or divide both sides by a negative number
 e.g. 2 2 a b a b < ¬ ÷ > ÷
o Take the reciprocal of both sides, if they are either both positive or both negative
 e.g.
1 1
a b
a b
< ¬ > , where , 0 a b > (or , 0 a b < )
- Apply your numerical answer to the context of the question, if appropriate.
Example [N2006/II/Q1]
Solve the inequality
2
9
1
9
x
x
÷
<
÷
.
Solution Comments
2
9
1
9
x
x
÷
<
÷
Use the General Approach outlined above.
2
9
1 0
9
x
x
÷
÷ <
÷
1. Zero the inequality.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
18
. )
2
2
9 9
0
9
x x
x
÷ ÷ ÷
<
÷
2
2
0
9
x x
x
÷
<
÷
2. Combine the terms to form a single fraction.
. ) . )
2 2
9 0 x x x ÷ ÷ <
. ) . ) . ) 1 3 3 0 x x x x ÷ + ÷ <
3. Factorize the terms as far as possible.
. ) . ) . ) 1 3 3 0 x x x x ÷ + ÷ >
Make the coefficient of the highest power of x ,
here
4
x , positive, so that the graph in Step 5
will start as positive from the right.
Remember to flip the sign of the inequality
when multiplying both sides by 1 ÷ .
Skip Step 4, since the terms are already linear.
5. Draw number line and graph and shade the
regions where inequality holds.
- Mark the roots 3 ÷ , 0 , 1, 3 of the equation
. ) . ) . ) 1 3 3 y x x x x = ÷ + ÷ .
- The graph has x -intercepts at the roots.
- If ( ) f x is a polynomial with the coefficient
of the highest power of x being positive,
the graph of ( ) f x will start as positive
from the right.
- The graph has a “wavy” shape.
3 x < ÷ or 0 1 x < < or 3 x >  Write the range of values of x , checking for
any values (especially endpoints) to be
excluded.
Since the original inequality has denominator
. ) . )
2
9 3 3 x x x ÷ = + ÷ , exclude 3 x = ÷ and
3 x = where the inequality is undefined.
Do NOT use commas: e.g. 3 x < ÷ , 0 1 x < < ,
3 x > . Write out “or” between every 2 adjacent
inequalities.
x
3 ÷ 3 0 1
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
19
Alternative method from this step
(Sketch the graph of
2
9
1
9
x
y
x
÷
= ÷
÷
)
From the GC,
3 x < ÷ , 0 1 x < < , or 3 x >
Alternative method
Since the question doesn’t ask for exact values,
zero the inequality, then use the GC directly.
Always zoom in to check places where the
graph is near the x -axis, where the inequality
is near equality.
e.g. the solution interval 0 1 x < < is not clearly
visible at some zoom levels, e.g. ZStandard and
ZDecimal.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
20
Arithmetic & Geometric Progressions
Below, AP denotes arithmetic progression and GP denotes geometric progression.
For an AP
1 2 3
, , , u u u . with first term
1
a u = and common difference d :
- nth term, . ) 1
n
u a n d = + ÷
- Sum of first n terms,
. ) . ) 2 1
2 2
n n
n n
S a n d a u ] = + ÷ = +
]
- If , , a b c form an AP, then 2 b a c b b a c ÷ = ÷ · = +
For a GP
1 2 3
, , , u u u . with first term
1
a u = and common ratio r :
- nth term,
1 n
n
u ar
÷
=
- Sum of first n terms,
. )
1
(1 )
1 1
n
n
n
a r
a r
S
r r
÷
÷
= =
÷ ÷
- Sum to infinity,
1
a
S
r
·
=
÷
( S
·
exists only if 1 r < )
- If , , a b c form a GP, then
2
b c
b ac
a b
= · =
For any sequence
1 2 3
, , , u u u .
-
1 k k k
u S S
÷
= ÷ , where 0 k > .
-
1 1 k k m m k
u u u S S
+ ÷
+ +.+ = ÷ , where 0 m k > > .
FAQ
Q: Find various properties of an AP or GP, given other properties of it.
A: 1. Write down a, d (or r ), n, any k th term
k
u , any sum of first k terms
k
S which is
given in the question.
2. Use the various AP & GP formulae to solve.
Q: Show that a sequence is an AP.
A: Show that the difference is common, i.e.
1 k k
u u
÷
÷ is constant for all k .
Q: Show that a sequence is a GP.
A:
Show that the ratio is common, i.e.
1
k
k
u
u
÷
is constant for all k .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
21
Q: Model and solve a problem using an AP or GP.
A: 1. Identify if the sequence given is an AP or GP.
2. Write down a, d (or r ), n, any k th term
k
u , any sum of first k terms
k
S which is
given in the question.
3. Use the various AP & GP formulae to solve.
Q: Find the least/most n such that some condition holds.
A: 1. Form an inequality in terms of n.
2. Solve manually OR use GC’s table of values/graph.
Example 1 [JJC/II/Q1]
An arithmetic progression of positive terms is such that twice the sum of the first nine terms
is equal to the sum of the next nine terms. Let
n
T denote the n th term of the arithmetic
progression. Given that
1
T , 20 ,
16
T forms a geometric progression, find the first term and the
common difference of the arithmetic progression.
Solution Comments
Let the arithmetic progression have first
term a, common difference d , and sum of
first n terms
n
S .
Define all variables that will be used. Interpret
the question in the context of the defined
variables: find a and d .
9 18 9
2S S S = ÷
9 18
3S S =
Always write sums of consecutive terms in
terms of differences of sums of first k terms:
10 11 18 18 9
T T T S S + +.+ = ÷
. )
9 18
3 2 8 (2 17 )
2 2
a d a d × + = +
27 108 18 153 a d a d + = +
5 a d =
Simplify the first sentence in the question to
an equation in terms of a and d , which we
want to find.
16
1 16
1
20
400
20
T
TT
T
= ¬ =
. ) 15 400 a a d + =
Simplify the other statement in the question to
another equation in terms of a and d .
5
a
d =
. )
2
3 4 400 a a a a + = =
2
100 a =
Now substitute the first equation, 5 a d = , into
the second equation. Simultaneous equations
are a common feature in AP & GP questions.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
22
10 a = or 10 ÷ (rejected 0 a > as the
arithmetic progression is of positive terms)
10 a = , 2 d = 
Look for conditions given in the question
whenever you must choose a correct value.
Example 2 [N97/I/Q15 (Part)]
A bank has an account for investors. Interest is added to the account at the end of each year at
a fixed rate of 5% of the amount in the account at the beginning of that year. A man decides
to invest $x at the beginning of one year and then a further $x at the beginning of the
second and each subsequent year. He also decides that he will not draw any money out of the
account, but just leave it, and any interest, to build up.
(i) Show that, at the end of n years, when the interest for the last year has been added, he
will have a total of
. )
$21 1.05 1
n
x ÷ in his account.
(ii) After how many complete years will he have, for the first time, at least $12x in his
account?
Solution Comments
(i) Let
n
u be the amount of money in $ in the
man’s account at the end of n years.
Define
n
u as this to avoid writing the
same phrase repeatedly. Remember to
specify the units (in $).
1
1.05 u x =
. ) . )
2
2
1.05 1.05 1.05 1.05 u x x x = + = +
. ) . )
. )
2
3
3 2
1.05 1.05 1.05
1.05 1.05 1.05
u x x
x
= + +
= + +
.
. )
1
1.05 1.05 1.05
n
u x
÷
= + +.+
n
n
Evaluate successive values of
k
u , i.e.
1
u ,
2
u ,
3
u , and observe a pattern for
n
u . Here, “replace” 3 with n.
Do not simplify
2
1.05 ,
3
1.05 etc., or
the pattern will be difficult to observe.
. )
1
1.05 1.05 1
1.05 1
n
x
÷
÷
=
÷
. )
1
21 1.05 1
n
x
÷
= ÷  (shown)
Apply the formula for the sum of the
first n terms of a GP to simplify
n
u .
(ii) 12
n
u x >
. )
1
21 1.05 1 12
n
x x
÷
÷ >
. )
1
21 1.05 1 12

÷ > ( 0 x > )
1
12 11
1.05 1
21 7

> + =
“At least” means use > instead of >.
Manipulate the inequality to solve for
n.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
23
. )
11
1 ln1.05 ln
7
n
´ }
÷ >
]
' ¹
11
ln
7
9.26
ln1.05
n
´ }
]
' ¹
> = (3 s.f.)
least integer value of 0 1 n = 
When n is in the exponent, take ln on
both sides. Here, the sign of the
inequality of preserved when dividing
both sides by ln1.05 0 > .
Switch the sign when dividing both
sides by ln x , where 0 1 x < < .
Example 3 [VJC06/J1CT/Q9b (Modified)]
At the beginning of the year, Xin Fu deposited $100,000 with a bank that pays 10% interest
per annum at the end of each year. After the interest is credited, he immediately withdraws
$12,000. Likewise, Xin Fu will again withdraw $12,000 at the end of each subsequent year,
immediately after the bank’s interest has been credited.
Express Xin Fu’s bank account balance in $ after his nth withdrawal in terms of n.
Solution Comments
Let
n
u be Xin Fu’s bank account balance in
$ after his nth withdrawal.
Define
n
u clearly, and specify the units ($).
. )
1
1.1 100000 12000 u = ÷
. ) . )
2 1
2
1.1 12000
1.1 100000 12000 1.1 1
u u = ÷
= ÷ +
. ) . )
3 2
3 2
1.1 12000
1.1 100000 12000 1.1 1.1 1
u u = ÷
= ÷ + +
.
. ) 1.1 100000 u =
n
n
. )
1 2
12000 1.1 1.1 1
n ÷ ÷
÷ + +.+
n
Evaluate successive values of
k
u , i.e.
1
u ,
2
u ,
3
u , and observe a pattern for
n
u . Here,
“replace” 3 with n, 2 with 1 n ÷ , etc.
Do not simplify
2
1.1 ,
3
1.1 etc., or the
pattern will be difficult to observe.
. )
. )
1 1.1 1
1.1 100000 12000
1.1 1
n
n
]
÷
] = ÷
÷
]
]
. ) 1.1 100000 120000(1.1 1)
n n
= ÷ ÷
. )
120000 20000 1.1
n
= ÷
. )
20000 6 1.1
n
= ÷ 
Apply the formula for the sum of the first
n terms of a GP to simplify
n
u .
Count the number of terms carefully. Here,
1 2
(1.1 1.1 1)
n n ÷ ÷
+ +.+ is a sum of powers
of 1.1 ranging from 0 to 1 n ÷ , so it has n
terms (not 1 n ÷ ).
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
24
Alternative methods
Write the recurrence relation:
1
1.1 12000
n n
u u
÷
= ÷
Now,
- Use repeated substitution.
- OR Use induction, especially if the result
to be proven:
. )
20000 6 1.1
n
n
u = ÷ is
already given.
Refer to Example 2 in Summation, Recurrence
& Induction.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
25
Summation, Recurrence & Induction
Summation
- Definition:
. ) . ) . ) 1 ( )
n
r m
f r f m f m f n
=
= + + +.+
¯
, where m n < .
- For constants a and b , functions of r ( ) f r and ( ) g r and m n < ,
o
. ) . )
n n
r m r m
af r a f r
= =
=
¯ ¯
o
. ) . ) . ) ( )
n n n
r m r m r m
af r bg r a f r b g r
= = =
] ± = ±
]
¯ ¯ ¯
o Sum of AP: general term has the form a br + .
o Sum of GP: general term has the form
r
ab .
o
. )
1
1
1
2
n
r n n
=
= +
¯
r
o
. ) . )
2
1
1
1 2 1
6
n
r n n n
=
= + +
¯
r
o
. ) . )
2
2
1
2
3 2
1
1 1
1 1
4 2
n n
r
r n n n n r
= =
´ }
´ }
= + = + =
] ]
' ¹
' ¹
¯ ¯
r
- Convert sums with 1 r = to use those formulas above showing r starting from 1.
o
. ) . ) . )
1
1
1
n n m
r m
f r f r f r
=
÷
= =
= ÷
¯ ¯ ¯
r r
- Use method of difference to find
r
u
¯
when consecutive terms cancel out. (Example 1)
Recurrence
- For a sequence
1 2
, , , u u . a recurrence relation defines it recursively. Each term
n
u is
defined as a function of the preceding terms
1 2 1
, , ,
n
u u u
÷
. .
- AP can be expressed as
1 n n
u u d
÷
= + , where the common difference d is a constant.
- GP can be expressed as
1 n n
u ru
÷
= , where the common difference r is a constant.
Mathematical Induction
Step 1: Base case. Aim: Show that
1
P is true.
Let
n
P be the statement
“…”
When 1 n = ,
L.H.S. = …
R.H.S. = …
L.H.S. = R.H.S.
1
P is true.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
26
Step 2: Inductive step. Aim: Show that
1 k
P
+
is true whenever
k
P is true.
Assume
k
P is true for some k
+
e ,
i.e. …
To prove
1 k
P
+
is true,
i.e. …
Now, L.H.S. = … = R.H.S.
i.e.
1 k
P
+
is true whenever
k
P is true.
Since
1
P is true, by Mathematical Induction,
n
P is true for all n
+
e .
- The assumption that
k
P is true for some k
+
e is the induction hypothesis.
- Make the necessary changes if the question specifies a base value of n other than 1.
- Know where you’re going when performing the inductive step:
1. Convert L.H.S. of
1 k
P
+
to involve L.H.S. of
k
P .
2. Since we assume
k
P is true, substitute L.H.S. of
k
P with R.H.S. of
k
P .
o The induction hypothesis must be used somewhere in the induction.
3. Convert the expression to R.H.S. of
1 k
P
+
(proven).
Example 1 [RJC08/Assignment 7/Q4 (Modified)]
Express
. ) . )
2

1 2 ( 3)
r
r r r + + +
in partial fractions.
Hence, find the sum of the first n terms of the series
2 4 6
2 3 4 3 4 5 4 5 6
+ + +.
× × × × × ×
Find the sum to infinity of this series.
Solution Comments
Let
. ) . ) . )
2
1 2 3 1 2 3
r A B C
r r r r r r
= + +
+ + + + + +
for some constants , , A B C .
. ) . ) . ) . ) 2 2 3 1 3 r A r r B r r = + + + + +
. ) 1 ( 2) C r r + + +
Recall the partial fractions decomposition for
linear factors in the denominator (in MF15).
Substitute appropriate values of r to find
, , A B C .
Check that the solution allows for cancelling
out terms using method of difference in the
later part: 1 4 3 0 ÷ + ÷ = .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
27
When 1 r = ÷ , 2 2 1 A A ÷ = ¬ = ÷
When 2 r =÷ , 4 4 B B ÷ = ÷ ¬ =
When 3 r = ÷ , 6 2 3 C C ÷ = ¬ = ÷
. ) . ) . )
2
1 2 3
r
r r r + + +
1 4
2 3
3
1 r r r
= ÷ + ÷
+ + +

Required sum
. ) . )
1
2

1 2 ( 3)
n
r
r
r r r
=
=
+ + +
¯
Observe that the terms of the given series are
of the form
. ) . )
2
1 2 ( 3)
r
r r r + + +
, where
1, 2, r = ..
1
1 4 3
1 2 3
n
r
r r r
=
´ }
= ÷ + ÷
]
+ + +
' ¹
¯
1 4 3
2 3 4
= ÷ + ÷
1 4 3
3 4 5
÷ + ÷
1 4 3
4 5 6
÷ + ÷
.
.
.
1 4 3
1 1 n n n
÷ + ÷
÷ +
1 4 3
1 2 n n n
÷ + ÷
+ +
1 4 3
1 2 3 n n n
÷ + ÷
+ + +
If a sum cannot be found directly or you have
decomposed the term into partial fractions,
think of the method of difference.
Write out the lines where 1, 2,3 r = at the top,
and 2, 1, r n n n = ÷ ÷ at the bottom. Always
do this for the first 3 values and last 3 values
of r.
The terms should be in terms of n , not r.
Show the cancelling out of terms clearly,
leaving the fractions in their original form
(do not simplify them).
1 4 1 3 4 3
2 3 3 2 2 3 n n n
= ÷ + ÷ ÷ + ÷
+ + +
1 1 3
2 2 3 n n
= + ÷
+ +

Write out the remaining terms and simplify.
Sum to infinity
1
1 4 3
1 2 3
r
r r r
·
=
´ }
= ÷ + ÷
]
+ + +
' ¹
¯
1
1 4 3
lim
1 2 3
n
n
r
r r r
÷·
=
]
´ }
= ÷ + ÷
] ]
+ + +
' ¹
]
¯
1 1 3
lim
2 2 3
n
n n
÷·
´ }
= + ÷
]
+ +
' ¹
1
2
= 
The sum to infinity is the limit as n÷· of
the sum to n terms. As n÷·,
1
0
2 n
÷
+
and
3
0
3 n
÷
+
. Hence, the sum to infinity is
1
2
.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
28
Example 2 (Refer to first paragraph of Example 3 in Arithmetic & Geometric Progressions.)
Let
n
u be Xin Fu’s bank account balance in $ after his n th withdrawal.
(i) State an expression for
n
u in terms of
1 n
u
÷
.
(ii) Hence, find
n
u in terms of n .
(iii) Prove your result in part (ii) for all 1 n > by mathematical induction.
Solution Comments
(i)
1
1.1 12000
n n
u u
÷
= ÷ 
(ii)
. )
1
1.1 100000 12000 98000 u = ÷ =
. )
. )
. ) . )
. )
. )
1
2
2
3
3
1.1 12000
1.1 1.1 12000 12000
12000 1
1.1 1.1 12000 12000 1.1 1
12000 1.1 1
12000 1.1 1
n n
n
n
n
u u
u
u
÷
÷
÷
÷
÷
÷
÷ ÷
= ÷
= ÷ ÷
= ÷ +
= ÷ ÷ +
= ÷ + +
= ÷ + +.+
2
2
3
3
1
2
2 1
1.1
1.1
1.1
1.1
.1 1.1 1
.
n
n
n n
u
u
u
Use repeated substitution. The recurrence
relation expresses
n
u in terms of
1
.
n
u
÷
Use
the relation repeatedly to express
n
u in
terms of
2 n
u
÷
,
3 n
u
÷
etc.
Observe a pattern and express
n
u in terms
of
1
u (which can be evaluated). Here, the
general form for
n
u is:
12000( 1)
÷
÷
÷ +.+
1
1 1. . 1 1
k k
n k
u
1
1 1
n k
u u n k k n
÷
= ¬ ÷ = ¬ = ÷ .
To express
n
u in terms of
1
u , substitute
1 k n = ÷ into the general form.
. )
. )
1
1
1 1.1 1
1.1 98000 12000
1.1 1
n
n
÷
÷
]
÷
] = ÷
÷
]
]
. ) . )
1 1
1.1 98000 120000 1.1 1
n n ÷ ÷
= ÷ ÷
. )
1
120000 22000 1.1

= ÷
. )
120000 20000 1.1
n
= ÷
. )
20000 6 1.1
n
= ÷ 
2 3
(1.1 1.1 1)
n n ÷ ÷
+ +.+ is a sum of powers
of 1.1 ranging from 0 to 2 n ÷ , so it has
1 n ÷ terms (not 2 n ÷ ).
(iii) Let
n
P be the statement

. )
20000 6 1.1
n
n
u = ÷ ”.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
29
When 1 n = ,
L.H.S.
1
98000 u = =
R.H.S.
. )
1
20000 6 1.1 98000 = ÷ =
L.H.S. = R.H.S.
1
P is true.
Show how the value 98000 is calculated.
Assume
k
P is true for some k
+
e ,
Assume the induction hypothesis.
i.e.
. )
20000 6 1.1
k
k
u = ÷ .
Replace n by k in the statement of
n
P .
To prove
1 k
P
+
is true,
i.e.
. )
1
1
20000 6 1.1
k
k
u
+
+
= ÷ ,
Replace n by 1 k + in the statement of
n
P .
Now, L.H.S.
1
1.1 12000
k k
u u
+
= = ÷ Use the given recurrence relation,
1
1.1 12000
k k
u u
+
= ÷ , to convert L.H.S. of
1 k
P
+
to involve L.H.S. of
k
P .
. )
1.1 20000 6 1.1 12000
k
]
= ÷ ÷
]
(by induction hypothesis)
. )
1
132000 12000 20000 1.1
k+
= ÷ ÷
. )
1
120000 20000 1.1
k+
= ÷
. )
1
20000 6 1.1
k+
= ÷
= R.H.S. (proven)
i.e.
1 k
P
+
is true whenever
k
P is true.
Since
1
P is true, by Mathematical
Induction,
n
P is true for all n
+
e .
Use the induction hypothesis that
k
P is
true, i.e.
. )
20000 6 1.1
k
k
u = ÷ , to substitute
L.H.S. of
k
P with R.H.S. of
k
P .
Convert the expression to the R.H.S. of
1 k
P
+
. Manipulate the expression to be more
and more similar to R.H.S. of
1 k
P
+
.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
30
Example 3 [N01/I/13b]
Use induction to prove that
. ) . ) . ) . ) . ) . ) . )
2
2
3 1! 7 2! 13 3! 1 ! 1 ! 1 n n n n n + + +.+ + + = + ÷
Solution Comments
. ) . ) . ) . ) . )
2
3 1! 7 2! 13 3! 1 ! n n n + + +.+ + +
. ) . )
2
1
1 !
n
r
r r r
=
= + +
¯
Rewrite the L.H.S. of the given equation as a
sum, for easier manipulation.
. ) . )
2
1 ! n n n + + shows the general form.
Check that it works for the first 3 terms
given.
Let
n
P be the statement

. ) . ) . ) . )
2
2
1
1 ! 1 ! 1
n
r
r r r n n
=
+ + = + ÷
¯
”.
When 1 n = ,
L.H.S.
. ) . )
2
1 1 1 1! 3 = + + =
R.H.S. . ) . )
2
1 1 1! 1 3 = + ÷ =
L.H.S. = R.H.S.
1
P is true.
Show how the value 3 is calculated.
Assume
k
P is true for some k
+
e ,
Assume the induction hypothesis.
i.e.
. ) . ) . ) . )
2
2
1
1 ! 1 ! 1
k
r
r r r k k
=
+ + = + ÷
¯
.
Replace n by k in the statement of
n
P .
To prove
1 k
P
+
is true,
i.e.
. ) . ) . ) . )
1
2
2
1
1 ! 2 1 ! 1
k
r
r r r k k
+
=
] + + = + + ÷
]
¯
,
Replace n by 1 k + in the statement of
n
P .
Now, L.H.S.
. ) . )
1
2
1
1 !
k
r
r r r
+
=
= + +
¯
. ) . )
2
1
1 !
k
r
r r r
=
= + +
¯
. ) . ) . )
2
1 1 1 1 ! k k k
]
] + + + + + +
]
]
. ) . ) . ) . )
2 2
1
1 ! 3 3 1 !
k
r
r r r k k k
=
] = + + + + + +
]
¯
Convert L.H.S. of
1 k
P
+
to involve L.H.S. of
k
P .
When L.H.S. of
n
P involves a sum of n
terms, extract the ( 1 k + )th term from the
sum of ( 1 k + ) terms in L.H.S. of
1 k
P
+
to
obtain the sum of k terms in L.H.S. of
k
P .
. ) . ) . ) . )
2
2
1 ! 1 3 3 1 ! k k k k k ] = + ÷ + + + +
]
Use the induction hypothesis that
k
P is true
to substitute L.H.S. of
k
P with R.H.S. of
k
P .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
31
. ) . ) 1 1 ! 1 k k ] = + + ÷
]
. ) . )
2
3 3 [ 1 !] k k k + + + +
. ) . )
2
4 4 1 ! 1 k k k ] = + + + ÷
]
. ) . )
2
2 1 ! 1 k k ] = + + ÷
]
= R.H.S. (proven)
i.e.
1 k
P
+
is true whenever
k
P is true.
Since
1
P is true, by Mathematical Induction,
n
P is true for all n
+
e .
Convert the expression to the R.H.S. of
1 k
P
+
.
Manipulate the expression to be more and
more similar to R.H.S. of
1 k
P
+
.
. ) 1 ! k + is the product of all positive integers
up to ( 1 k + ), and ! k is the product up to k .
Hence:
. ) . ) . ) 1 ! 1 ! k k k + = +
. ) 1 !
1
!
k
k
k
+
= +
In general:
. ) . ) . ) ! 1 2 1 ( !) m m m m k k ] = ÷ ÷ . +
]
. ) . ) . )
!
1 2 1
!
m
m m m k
k
= ÷ ÷ . +
where 0 m k > > .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
32
Differentiation
Basic Formulae
Let u and v be functions of x . Let y be a function of u .
. )
d du dv
u v
dx dx dx
± = ±
. )
d du dv
uv v u
dx dx dx
= + (Product Rule)
2
du dv
v u
d u
dx dx
dx v v
÷
´ }
=
]
' ¹
(Quotient Rule)
dy dy du
dx du dx
= × (Chain Rule)
Derivatives of Standard Functions
Let n be an integer and a be a constant.
Unless otherwise stated, the derivatives are valid for all xe.
NOT in MF15:
. ) = y f x
. ) = '
dy
f x
dx
Range of validity/Comments
n
x
1 n
nx
÷
Memorize
2
1 1 d
dx x x
´ }
= ÷
]
' ¹
and
1
2
d
x
dx x
= .
x
e
x
e
x
a ln
x
a a
ln x
1
x
0 x >
sin x cosx
cosx sin x ÷
tan x
2
sec x
cosecx cosec cot x x ÷
cot x
2
cosec x ÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
33
In MF15:
. ) = y f x
. ) = '
dy
f x
dx
Range of validity/Comments
secx sec tan x x
1
sin x
÷
2
1
1 x ÷
1 x <
1
cos x
÷
2
1
1 x
÷
÷
1 x <
1
tan x
÷
2
1
1 x +
-
1
1
sin
sin
x
x
÷
= , while
1
cosec
sin
x
x
= .
-
1
1
cos
cos
x
x
÷
= , while
1
sec
cos
x
x
= .
-
1
1
tan
tan
x
x
÷
= , while
1
cot
tan
x
x
= .
Techniques of Differentiation
- Let . ) f x be any function and . ) ' f x be its derivative given in the table above. To find
the derivative of
. ) . )
f g x , use the Chain Rule with . ) u g x = and
. ) . )
( ) y f g x f u = = :
(Example 1)
dy dy du
dx du dx
= ×
. ) . ) . ) . ) . )
' '
d
f g x f g x g x
dx
= ×
- Implicit differentiation: If an equation involving x and y can be expressed as
. ) y f x = , differentiating explicitly gives
. )
dy d
f x
dx dx
= . However, if the equation cannot
be easily expressed as ( ) y f x = , we can find
dy
dx
by differentiating the equation term-by-
term, with respect to x . (Example 2)
Use the Chain Rule to differentiate a function of y , say . ) g y , with respect to x .
. ) ( )
d d dy
g y g y
dx dy dx
= ×
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
34
- Parametric differentiation: When a curve involving and is described by equations
of the form ( ) x f t = and ( ) y g t = , with t in some given range of values, each value of
the parameter t gives a point with coordinates . ) ( , ) ( ), ( ) x y f t g t = on the curve.
(Example 3)
o Use the Chain Rule to find
dy
dx
, after calculating
dx
dt
and
dy
dt
.
dy dy dt dy dx
dx dt dx dt dt
= × = ÷
Applications of Differentiation
- Properties of . ) f x from its Derivatives
For a range of values
< < a x b, . ) f x is:
Only if for all < < a x b: i.e. Gradient (
dy
dx
) is:
Strictly increasing 0
dy
dx
> Positive
Strictly decreasing 0
dy
dx
< Negative
Concave upwards
2
2
0
d y
dx
>
Increasing
Concave downwards
2
2
0
d y
dx
< Decreasing
o If for a range of values a x b < < , the graph of . ) f x looks like a U ( ) or a part
of it, then . ) f x is concave Upwards. If it looks like an n (
·
) or a part of it, then
. ) f x is concave downwards.
- First Derivative Test: To determine the nature of a stationary point at x  = , evaluate
dy
dx
at a x value slightly less than  and a x value slightly more than , and observe
whether these values of
dy
dx
are positive or negative.
o Use it when
2
2
d y
dx
is difficult to find (
dy
dx
is difficult to differentiate) or
2
2
0
d y
dx
= .
o In the following table, write the actual values of 
÷
,  , 
+
and
dy
dx
at these
values of x .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
35
x

÷


+

÷


+

÷


+

÷


+
dy
dx
0 > 0 = 0 < 0 < 0 = 0 > 0 > 0 = 0 > 0 < 0 = 0 <
Graph
Nature of
stationary
point
Maximum
turning point
Minimum
turning point
Stationary point of inflexion
- Second Derivative Test: To determine the nature of a stationary point at x  = , evaluate
2
2
d y
dx
at x  = and observe whether it is positive or negative.
o In the following table, write the actual values of  and
2
2
d y
dx
at x  = .
2
2
d y
dx
at x  
0  0  0 
Nature of
stationary
point
Maximum
turning point
Minimum
turning point
No conclusion
possible
- Tangents & Normals
Q: Find the equation of a straight line.
A:
Given the gradient m and a point on the line with coordinates . )
1 1
, x y ,
1 1
( ) y y m x x ÷ = ÷
Q: Find the gradient, m, of a straight line.
A:
Given 2 points on the line with coordinates . )
1 1
, x y and . )
2 2
, x y ,
2 1
2 1
y y
m
x x
÷
=
÷
Given that the line is the tangent to a curve . ) y f x = at x  = ,
dy
m
dx
= , where x  =
Given that the line is the normal to a curve . ) y f x = at x  = ,
1
dy
m
dx
= ÷ ÷ , where x  =
since
1 2
1 m m = ÷ , where
1
m and
2
m are the gradients of 2 perpendicular lines.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
36
Q: Find the axis intercepts of a curve.
A: x -intercepts: Let 0 y = in the equation of the curve and solve for x .
y -intercepts: Let 0 x = in the equation of the curve and solve for y .
Q: Find the value of x at which the tangent is horizontal (parallel to the x -axis).
A:
Let 0
dy
dx
= and solve for x .
Q: Find the value of x at which the tangent is vertical (parallel to the y -axis).
A:
Let 1 0
dx dy
dy dx
= ÷ = and solve for x .
Q:
Find the distance, d , between 2 points with coordinates . )
1 1
, x y and . )
2 2
, x y .
A:
. ) . )
2 2
2 1 2 1
d x x y y = ÷ + ÷
- Maximization, Minimization & Rates of Change
o Always show that a stationary point you have found is indeed a maximum point
(or minimum point), using either Derivative Test to determine its nature.
o Remember to include units in your final answer.
o Make sure that there is only 1 variable in the expression before differentiating the
expression with respect to that same variable.
o Use the Chain Rule to connect rates of change.
o Similar triangles and trigonometric ratios are useful in forming expressions.
Example 1
Differentiate
x
a with respect to x .
Solution Comments
ln
x
x a
a e =
(since
ln k
k e = for all k e )
ln x a
e = (since log log
r
a a
b r b = )
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
37
Let ln u x a = ,
u
y e = .
ln x x a
d d dy
a e
dx dx dx
= =
By the Chain Rule,
. ) . )
ln
ln
u
u
dy dy du
dx du dx
d d
e x a
du dx
e a
= ×
= ×
=
The details are unnecessary for actual working. Only write:
ln
ln
l
n
n
l
x x a
x a
x
d d
a e
dx dx
e a
a a
=
=
= 
. )
ln
ln
x a
e a =
ln
x
a a = 
Substitute u back with ln x a .
Example 2 [NJC06/Midyear]
The equation of a curve C is given by
3 2 3
2 y yx x k + + = , where k is a constant.
Find
dy
dx
in terms of x and y .
Solution Comments
3 2 3
2 y yx x k + + =
Differentiating with respect to x ,
Whenever an equation involving x and y
cannot be easily expressed as
. ) y f x = , use
implicit differentiation.
2 2 2
3 2 6 0
dy dy
y x xy x
dx dx
+ + + =
Using the Chain Rule,
3 3 2
3
d d dy dy
y y y
dx dy dx dx
= × =
Using the Product Rule,
2 2 2
d d d
yx x y y x
dx dx dx
= +
2
2
dy
x xy
dx
= +
. )
2 2 2
3 2 6 0
dy
y x xy x
dx
+ + + =
. ) . )
2 2 2
3 2 6
dy
y x xy x
dx
+ = ÷ +
. )
2
2 2 2 2
2 3
2 6
3 3
x y x
dy xy x
dx y x y x
+
+
= ÷ = ÷
+ +

Group the terms containing
dy
dx
by factoring
it out, then shift
dy
dx
to one side and factorize
further.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
38
Example 3 [SRJC08/Prelim/I/9i]
The parametric equations of a curve are given by
. )
2
1
2
4
t
x e t = ÷ and
t
y e = , where 0 1 t < < .
Find
dy
dx
and show that
2
2
0
d y
dx
< for 0 1 t < < .
Solution Comments
t
dy
e
dt
=
. )
2 2
1 1 1
2 ( 1)
4 2 2
t t
dx
e e
dt
= ÷ = ÷
Whenever equations for x and y are given
in terms of a parameter t , use parametric
differentiation.
Find
dy
dt
and
dx
dt
first.
dy dy dx
dx dt dt
= ÷
2
2
1
t
t
e
e
=
÷

Use the Chain Rule.
2
2 2
2
1
t
t
d y d e
dx dx e
´ }
=
]
÷
' ¹
2
2
d y d dy
dx dx dx
´ }
=
]
' ¹
2
2
1
t
t
d e dt
dt e dx
´ }
= ×
]
÷
' ¹
Use the Chain Rule to differentiate the
expression involving t with respect to t .
2
1
2
1
t
t
d e
dx dt e
dt
´ }
= ×
]
÷ ´ }
' ¹
]
' ¹
1 dt
dx dx
dt
=
´ }
]
' ¹
. ) . )
. )
2 2
2 2
2
1 2
2
2
1
1
t t t t
t
t
e e e e
e
e
]
÷ ÷
´ }
]
=
]
] ÷
' ¹
÷
]
. )
. )
2
3
2
4
1
1
t t
t
e e
e
]
]
= ÷ ÷
]
÷
]
. )
. )
2
3
2
4 1
1
t t
t
e e
e
+
= ÷
÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
39
For all t e , 0 1 t < < ,
1
t
e > and
2
1
t
e > .
Try to find an inequality satisfied by terms in
2
2
d y
dx
to prove that
2
2
0
d y
dx
< for 0 1 t < < . Use
the given range of values of t .
Since 1 0
t
e > > ,
2
1 2 0
t
e + > > and
2
1 0
t
e ÷ > ,
. )
. )
. )
2
2
3 2
2
1
0 4
1
t t
t
e e
d y
dx
e
]
+
]
= ÷ <
]
÷
]
 (shown)
For all xe, 0 x > , 1
x
e > .
Show that the factors in the numerator and
denominator of
. )
. )
2
3
2
1
1
t t
t
e e
e
+
÷
are all positive,
hence the whole fraction must be positive.
Example 4 [N1998/I/7]
A spherical balloon is being inflated and, at the instant when its radius is 3 m, its surface
area is increasing at the rate of 2 m
2
s
-1
. Find the rate of increase, at the same instant, of
(i) the radius,
(ii) the volume.
[The formulae for the surface area and volume of a sphere are
2
4 A r  = and
3
4
3
V r  = .]
Solution Comments
(i) Let A be the surface area of the balloon in
m
2
.
Let r be the radius of the balloon in m.
Define all variables that will be used, with
their units.
2
dA
dt
=
2
4 A r  =
Write down all given rates of change, and
any relations between the variables.
When 3 r = ,
8 24
dA
r
dr
  = =
Substitute 3 r = only in
dA
dx
, and not in
2
4 A r  = , since r is a variable which
takes a specific value of 3 in the question,
not a constant.
Alternatively, find the expression for
dr
dt
first,
1
4
dr
dt r 
= , then substitute 3 r = .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
40
1
1
2
24
1
12
dr dr dA
dt dA dt
dA dA
dr dt


= ×
´ }
= ÷ ×
]
' ¹
= ×
=
Identify what you need to find and what is
given:
dr
dt
given
dA
dt
and
dA
dr
. Connect
them using the Chain Rule. You can think
of the dA as being “cancelled out”.
Hence, the radius is
increasing at
1
12
m s
-1
. 
Remember to include units in your final
answer. Do not write units in
dr
dt
, since r
and t are already in terms of m and s
respectively.
(ii) Let V be the volume of the balloon in m
3
.
3
4
3
V r  =
When 3 r = ,
2
4 36
dV
r
dr
  = =
36 2 72
dV dV dr
dt dr dt
  = × = × =
Hence, the volume is
increasing at 72 m
3
s
-1
. 
Example 5 [SRJC08/Prelim/I/6]
The diagram below shows the cross-section of a cylinder of
radius x that is inscribed in a sphere of fixed internal radius R .
Show that
. )
2 2 2 2 2
16 A x R x  = ÷ , where A is the curved surface
area of the cylinder.
Prove that, as x varies, the maximum value of A is obtained
when the height of the cylinder is equal to its diameter.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
41
Solution Comments
Draw a line across the rectangular cross-
section of the cylinder. Introducing right-
angled triangles is useful because it allows
the use of Pythagoras’ Theorem and
trigonometric ratios.
Let h be the height of the cylinder. Define all variables that will be used, and
are not already defined in the question.
By Pythagoras’ Theorem,
. ) . )
2 2
2
2 2 R h x = +
2 2
2 h R x = ÷
The curved surface area of a cylinder is its
circumference × height 2 2 rh xh   = = .
Since the expression to be proven is in
terms of R and x only, express h in terms
of R and x first.
. )
2 2
2 2
2
2 2
4
A xh
x R x
x R x



=
= ÷
= ÷
. )
2 2 2 2 2
6 1 A x R x  = ÷  (shown)
2 2
4
dA d
x R x
dx dx
 = ÷
2 2
2 2
2
4
2
x
R x x
R x

]
´ }
÷
= ÷ +
] ]
÷ ]
' ¹
]
2
2 2
2 2
4
x
R x
R x

´ }
= ÷ ÷
]
÷
' ¹
Use the Product Rule. The question states
A is to be maximized as x varies, so find
the appropriate derivative
dA
dx
.
Since R is a constant,
2
0
d
R
dx
= .
R
R
2R
2x
h
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
42
. )
2 2
2 2
4
2 R x
R x

= ÷
÷
When 0
dA
dx
= ,
2 2
2 0 R x ÷ =
Factorize
dA
dx
by removing terms which
you know are non-zero, such as constants
and . )
2
f x ]
]
, . ) f x , where
. ) 0 f x = for
all x . e.g. x R = here ( 0 x R h = ¬ = ),
so
2 2
0 R x ÷ > .
This allows cancelling out of terms when
dA
dx
is equated to 0 .
2
2
2
R
x =
2
R
x = or
2
R
÷ (rejected 0 x > )
2
R
x =
To prove A is maximum when 2 h x = ,
first let 0
dA
dx
= , then solve for x .
Hence,
2
R
x = maximizes A .
x
2
R
÷
´ }
]
' ¹
2
R
2
R
+
´ }
]
' ¹
dA
dx
0 > 0 = 0 <
Always verify the nature of the stationary
point. Here,
dA
dx
is difficult to differentiate,
so use the First Derivative Test instead.
Since there is only 1 value of x which
gives a stationary point, it must be the
“maximum value” to be proven.
To carry out the test, choose some 0 R > ,
e.g. 1 R = .
When
2
R
x = , 2 R x = ,
2 2
2 h R x = ÷
. )
2
2
2 2x x = ÷
2
2 x =
2x = or 2x ÷ (rejected 0 x > and 0 h > )
2 h x = = diameter of cylinder  (proven)
Express R in terms of x , then substitute
for R in the equation
2 2
2 h R x = ÷ .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
43
Integration
Basic Formulae
Let u and v be functions of x .
. ) u v dx udx vdx + = +
¦ ¦ ¦
. ) . )
d
f x dx f x c
dx
= +
¦
Integration reverses differentiation.
. ) . )
d
f x dx f x
dx
=
¦
Differentiation reverses integration.
If
. ) f x is a function which is defined and continuous for all a x b < < and
. ) . ) f x dx F x c = +
¦
then the definite integral of
. ) f x from a to b with respect to x is:
. ) . ) . ) . )
b b
a a
f x dx F x F b F a ] = = ÷
] ¦
Integration of Standard Functions
Let n be an integer and a be a constant.
Unless otherwise stated, the derivatives below hold for all xe.
Remember to include the arbitrary constant c for indefinite integrals.
NOT in MF15:
. ) f x . )
¦
f x dx
Range of
validity
. ) f x . )
¦
f x dx
n
x
1
1
n
x
n
+
+
1 n = ÷ cosx sin x
1
x
ln x
sin x cosx ÷
x
e
x
e
2
sec x
tan x
x
a
ln
x
a
a
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
44
In MF15:
. ) f x . )
¦
f x dx
Range of
validity
. ) f x . )
¦
f x dx
Range of
validity
2 2
1
x a +
1
1
tan
x
a a
÷
´ }
]
' ¹
tan x . ) ln sec x
1
2
x  <
2 2
1
a x ÷
1
sin
x
a
÷
´ }
]
' ¹
x a < cot x . ) ln sin x
0 x  < <
2 2
1
x a ÷
1
ln
2
x a
a x a
÷ ´ }
]
+
' ¹
x a > cosecx . ) ln cosec cot x x ÷ +
0 x  < <
2 2
1
a x ÷
1
ln
2
a x
a a x
+ ´ }
]
÷
' ¹
x a < tan x . ) ln sec tan x x +
1
2
x  <
If ( ) f x dx
¦
involves a logarithm ( | ¦
ln ( ) g x , where ( ) g x is a function of x ), place a
modulus around the expression within the logarithm, i.e. write ln ( ) g x . This removes the
need to specify the range of validity.
- e.g. cot ln sin x x c = +
¦
Techniques of Integration
- Let
. ) f x be any function and
. ) F x be its integral given in the table above. To find the
integral of
. ) . ) . ) ' f g x g x × , where
. ) g x is a function of x , use the fact that by the
Chain Rule from differentiation,
. ) . ) . ) . ) . ) '
d
F g x f g x g x
dx
= × .
Hence,
. ) . ) . ) . ) . )
' f g x g x dx F g x c × = +
¦
- Trigonometric Formulae: Used to integrate expressions with trigonometric functions.
NOT in MF15: In MF15:
2 2
sin cos 1 x x + =
2 2
tan 1 sec x x + =
2 2
cot 1 cosec x x + =
sin 2 2sin cos x x x =
2 2
2
2
cos2 cos sin
2cos 1
1 2sin
x x x
x
x
= ÷
= ÷
= ÷
(and others)
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
45
- Partial Fractions: Used to integrate expressions with an algebraic fraction, which has a
denominator that can be factorized (or is already factorized). Splitting the fraction into
partial fractions can help integrate it, since the partial fractions will be of the form
. )
. )
' f x
f x
.
Partial fraction decomposition formulae are given in MF15.
- Integration by Parts: Used to integrate expressions with a product of 2 distinct functions.
dv du
u dx uv v dx
dx dx
= ÷
¦ ¦
Generally, choose u in the following order such that
dv
dx
is easier to integrate:
LIATE: Logarithmic, Inverse trigonometric, Algebraic, Trigonometric, Exponential.
o To find ( ) f x dx
¦
when you can find '( ) xf x dx
¦
, let 1 u = and ( )
dv
f x
dx
= .
o Use algebraic manipulation if ( ) f x dx
¦
becomes the only integral on both sides.
- Integration by Substitution: If a substitution is required, it will be given in the question.
To find ( ) f x dx
¦
with the substitution
. ) x g u = :
. ) . ) . )
dx
f x dx f g u du
du
=
¦ ¦
For definite integrals,
. ) . )
2 2
1 1
( )
x u
x u
dx
f x dx f g u du
du
=
¦ ¦
where
. )
1 1
x g u = and
. )
2 2
x g u = .
o For indefinite integrals, express your final answer in terms of x , not u .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
46
Applications of Integration
If
. ) 0 f x > for all a x b < < , and R is the
area between the curve and the x -axis from
x a = to x b = ,
The area of R can be approximated by the
sum of areas of n rectangles, as shown.
The exact area of R is the limit of this sum
of areas as n ÷ ·, i.e. the definite integral:
. )
b
a
f x dx
¦
If
. ) . ) 0 f x g x > > for all a x b < < ,
Area of R :
. ) . )
b
a
f x g x dx ] ÷
] ¦
Volume of solid formed when R is revolved
completely about the x -axis:
. ) . ) . ) . )
2 2 b
a
f x g x dx 
]
÷
]
]
¦
If
. ) . ) 0 f y g y > > for all a x b < < ,
Area of R :
. ) . )
b
a
f y g y dx ] ÷
] ¦
Volume of solid formed when R is revolved
completely about the y -axis:
. ) . ) . ) . )
2 2 b
a
f y g y dy 
]
÷
]
]
¦
x
( ) y f x =
y
0 a
b
( ) y f x =
( ) y g x =
y
a
0 b x
R
a
0
b
x
y
R
( ) x g y = ( ) x f y =
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
47
If a function is defined in parametric form
as
. ) x f t = ,
. ) y g t = , and
. ) 0 y g t = > for
all a x b < < ,
Area of R :
. )
q
p
dx
g t dt
dt
¦
where
. ) f p a = and
. ) f q b = .
If
. ) y f x = has both positive and negative
components within the limits of integration,
split the area between the curve and the x -
axis into its positive and negative parts.
If
. ) 0 f x < for all a x c < < and
. ) 0 f x > for all c x b < < ,
Area of R :
. ) . )
c b
a c
f x dx f x dx ÷ +
¦ ¦
Or:
. )
b
a
f x dx
¦
If
. ) y f x = and
. ) y g x = intersect each
other within the limits of integration, split the
area between the 2 curves into where
. ) . ) f x g x > and where
. ) . ) g x f x > .
If
. ) . ) 0 g x f x > > for all a x c < < and
. ) . ) 0 f x g x > > for all c x b < < ,
Area of R :
. ) . ) . ) . )
c b
a c
g x f x dx f x g x dx ] ] ÷ + ÷
] ] ¦ ¦
Or:
. ) ( )
b
a
f x g x dx ÷
¦
x
R
. ) y f x =
y
0
a
b c
x
x
. ) x f t = , . ) y g t =
y
0 a
b
R
R
. ) y f x =
y
0 a
b c
. ) y g x =
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
48
If y k = for some k e , 0 k = , and a b < ,
Area of R :
. ) k b a ÷
The solid formed when R is revolved
completely about the x -axis is a cylinder
with radius k , height b a ÷ , and volume:
. )
2
k b a  ÷
- Use the GC to check your answer, or find an integral if an exact answer is not required.
o [MATH] [9], then key in fnInt(<expression>, X , <lower limit>, <upper limit>)
e.g. to find
1
2
0
x dx
¦
, key in fnInt
. )
2
, , 0,1 X X .
- Use the GC to sketch graphs of functions first before finding the areas or volumes
required. For parametric functions, use [MODE] PAR and key in T using [ , ,
Θ ,
X T n].
Example 1 [N08/I/5]
(i) Find the exact value of
1
3
2
0
1
1 9
dx
x +
¦
.
(ii) Find, in terms of n and e ,
1
ln
e
n
x xdx
¦
, where 1 n = ÷ .
Solution Comments
(i)
1 1
3 3
2 2
0 0
2
1 1 1
1 9 9
1
3
dx dx
x
x
=
+
´ }
+
]
' ¹
¦ ¦
. )
1
1
3
0
1
3tan 3
9
x
÷
] =
]
. )
1 1
1
tan 3 tan 0
3
÷ ÷
= ÷
1
3 3
 ´ }
=
]
' ¹
9

= 
Recognize when the integral is of a form given
in MF15, here
2 2
1
x a +
. Either extract
1
9
to get
the
2
x term, as shown, or extract
1
3
(since
. ) 3 3)
d
x
dx
= to get:
. )
1
3
2
2 0
1 3
3
1 3
dx
x +
¦
Then, apply the formula in MF15.
Check your answer using GC.
R
y k =
y
0 a
b
x
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
49
(ii)
Let ln u x = ,
n
dv
x
dx
= .
Then
1 du
dx x
= ,
1
1
n
x
v
n
+
=
+
.
1
1 1
1
ln
ln
1 1
e
n n
e e
n
x x x
x xdx dx
n n
+
]
= ÷
]
+ +
]
¦ ¦
1
1
1
ln 1
1 1
e
n
e
n
x x
x dx
n n
+
]
= ÷
]
+ +
]
¦
. )
1 1
2
1
ln
1
1
e
n n
x x x
n
n
+ +
]
= ÷ ]
+
+ ]
]
. ) . )
1 1
2 2
1
0
1
1 1
n n
e e
n
n n
+ +
] ]
= ÷ ÷ ÷ ] ]
+
+ + ] ]
] ]
. ) . )
1 1
2 2
1
1
1 1
n n
e e
n
n n
+ +
= ÷ +
+
+ +
. )
. )
1 1
2
1 1
1
n n
e n e
n
+ +
+ ÷ +
=
+
. )
1
2
1
1
n
ne
n
+
+
=
+

Here, ln x is Logarithmic and
n
x is Algebraic,
so by LIATE, choose ln u x = . Write out u ,
dv
dx
,
du
dx
, and v .
n is a constant, so treat it as a constant when
integrating.
For definite integrals, remember to keep the
limits on every term (integrals and results of
integrals).
Check your answer using GC by choosing some
n , 1 n = ÷ , e.g. 2 n = .
Example 2 [HCI08/Prelim/I/10]
The diagram below shows the graphs of
2
2
2 1
x
y
x
=
+
and
4
y x = , for 0 x > . R is the region
bounded by the y -axis, both curves and 1 y = .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
50
(i) By substituting
2
1
sin
2
y  = , find the exact value of
1
4
0
1 2
y
dy
y ÷
¦
(ii) Verify that the two curves intersect at the point
1 1
,
4 2
´ }
]
' ¹
. Hence or otherwise find the
exact area of region R.
(iii) Find the volume of the solid generated when R is rotated through 2 radians about
the y -axis, giving your answer correct to 3 decimal places.
Solution Comments
(i)
. ) . )
1
2sin cos
2
sin cos
dy
d
 

 
=
=
When 0 y = , sin 0 0   = ¬ =
When
1
4
y = ,
2
sin
2
 =
or
2
2
÷ (rejected)
4

 ¬ =
To find ( ) f y dx
¦
with the substitution
. ) y g  = , recall the general formula:
. ) . )
2 2
1 1
y
y
dy
f y dy f d
d


 

=
¦ ¦
where . )
1 1
y g  = and . )
2 2
y g  = .
For a definite integral, change 4 things from
the original integral:
1. Limits of integration.
2. The function to be integrated.
3. Multiply by
dy
d
.
4. Integrate with respect to  , not y .
Find 1. and 3. first in your working.
2
1
4 4
2
0 0
1
sin
2
1 2 1 sin
y dy
dy d
y d
 

 
´ }
=
]
÷ ÷
' ¹
¦ ¦
2
4
2
0
1
sin
2
sin cos
cos
d
 
  

=
¦
4
0
1 sin
sin cos
cos 2
d


  

=
¦
2
4
0
1
sin
2
d

  =
¦
Write out the original integral, and make all 4
changes above at the same time. Simplify the
substituted integral.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
51
. )
4
0
1
1 cos 2
2 2
d

  = ÷
¦
4
0
1 1
sin 2
2 2 2

 
]
= ÷
]
]
. )
2 1
0 0
4 4 2
 ]
´ }
= ÷ ÷ ÷
] ]
' ¹
]
2( 2)
16
 ÷
= units
2

To integrate
2
sin x or
2
cos x , use the double-
angle formula for cos 2x to convert them.
2 2
cos 2 1 2sin 2cos 1 x x x = ÷ = ÷
2
1
sin (1 cos 2 )
2
x x = ÷
. )
2
1
cos 1 cos 2
2
x x = +
(ii)
When
1
2
x = ,
4
4
1 1
4 2
x
´ }
= =
]
' ¹
2
2
1
1
2
1 2 1 4
2 1
2
x
x
= =
+ ´ }
+
]
' ¹
Hence, the curves intersect
at the point
1 1
,
4 2
´ }
]
' ¹
.  (verified)
The question only requires you to verify
(NOT show) that at
1 1
,
4 2
´ }
]
' ¹
, the curves
intersect. Hence, just show that the point lies
on both curves.
If the question requires us to find all
intersection points, you must use the equation
2
4
2
2 1
x
x
x
=
+
and solve for x .
1
4
4
y x x y = ¬ =
2
2
2 1
x
y
x
=
+
2 2
2 0 yx y x ¬ + ÷ =
. )
2
2 1 y x y ¬ ÷ =÷
1 2
y
x
y
¬ =
÷
or
1 2
y
y
÷
÷
(rejected 0 x > )
1
A
R
1 1
2
1
4
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
52
Area of
1 1
1
4 4
1
0
4
1 2
y
R dy y dy
y
= +
÷
¦ ¦
Look at the graph above to decide what to
integrate. From 0 y = to
1
4
y = , R is
bounded by the y -axis and
1 2
y
x
y
=
÷
.
From
1
4
y = to 1 y = , R is bounded by the y
-axis and
1
4
x y = . Hence, split the integral
into these 2 regions.
The method shown here integrates with
respect to y , For each curve, express x in
terms of y , then integrate with the limits
from the y -axis.
. )
1
5
4
1
4
2 2
4
16 5
y
 ÷ ]
= +
]
]
. )
5
4 2 2
4 4 1
16 5 5 4
 ÷
´ }
= + ÷
]
' ¹
. )
1
4 2 2
4 4 1 1
16 5 5 4 4
 ÷
´ }´ }
= + ÷
] ]
' ¹' ¹
. )
1
4
2 2
4 1 1
16 5 5
4

´ }
÷
]
= + ÷
]
' ¹
. ) 2 2
4 1
16 5 5 2
 ÷
= + ÷
4 2 2 2
5 16 8 10

= + ÷ ÷
4 2 9 2
5 16 40

´ }
= + ÷
]
]
' ¹
units
2

We know the value of the first integral,
1
4
0
1 2
y
dy
y ÷
¦
, from part (i).
Alternative method
Let A be the region bounded by the axes,
1 x = and 1 y = , as shown by the dotted
rectangle in the graph above. A is a rectangle
with area 1.
Area of region R
= Area of region A
÷ Area of region outside R and inside A
1 2
1
4
2
1
2
0
2
1
2 1
x
dx x dx
x
]
= ÷ +
]
+
]
¦ ¦
Use long division on
2
2
2 1
x
x +
to get an
integral of the form
2 2
1
a x +
, then use MF15.
The alternative method integrates with respect
to x . Each curve is already expressed as y in
terms of x , hence integrate with the limits
from the x -axis.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
53
(iii) Volume required
1 1
1
4 2
1
0
4
1 2
y
dy y dy
y

]
= +
]
÷
]
¦ ¦
1.984 = units
3
(3 d.p.) 
Volume of revolution about y -axis is given
by
2
x dy 
¦
. Again, split the integral into 2
regions, then express
2
x in terms of y for
each region.
The question states “3 decimal places” but not
“exact”, suggesting that you find the volume
directly using GC.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
54
Differential Equations
Below, DE denotes Differential Equation.
Forms of DEs and General Solutions
Convert any given DE into one of these 3 forms.
Form General Solution
. )
dy
f x
dx
=
Integrate both sides with respect to x .
. ) y f x dx =
¦
. )
dy
f y
dx
=
Divide both sides by . ) f y .
. )
1
1
dy
f y dx
´ }
=
]
]
' ¹
Integrate both sides with respect to x .
You can imagine the dx ’s cancelling out:
. ) . )
1 1 dy
dx dy
f y dx f y
´ }
=
]
]
' ¹
¦ ¦
. )
1
dy x
f y
=
¦
. )
2
2
d y
f x
dx
=
Integrate both sides with respect to x .
( )
dy
f x dx
dx
=
¦
Say
. ) ( ) f x dx g x A = +
¦
,
where A is an arbitrary constant.
. )
dy
g x A
dx
= +
Integrate both sides again with respect to x .
. ) . )
( )
y g x A dx
g x dx Ax
= +
= +
¦
¦
Say
. ) ( ) g x dx h x B = +
¦
,
where B is an arbitrary constant.
. ) y h x B Ax = + +
There are 2 arbitrary constants in the
general solution of this form.
. ) y h x Ax B = + +
Particular Solutions
- To find a “particular solution” to a DE, given some initial conditions,
1. Substitute the initial conditions into the general solution.
2. Find the value of the arbitrary constants.
3. Rewrite the general solution with the known values for the arbitrary constants.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
55
- To draw a family of solution curves,
1. Substitute at least 3 distinct values for the arbitrary constant c .
Generally, ensure you include these 3 cases: 0 c = , 0 c < , and 0 c > .
2. Draw the graphs using GC, indicating intercepts, asymptotes, turning points, etc.
Solving DEs by Substitution
If a DE is given as . ) ,
dy
f x y
dx
= and a substitution . ) , v g x y = is given to solve this DE,
where ( , ) f x y and . ) , g x y are functions of x and y, aim to express the DE in terms of v
and x only (without y).
1. Differentiate the given substitution with respect to x to get
dy
dx
in terms of
dv
dx
, v and x .
2. In the original DE, express y and
dy
dx
in terms of
dv
dx
, v and x .
3. Convert the substituted DE into the form required or one of the 3 standard forms above.
4. Solve the DE.
5. Express v in terms of x and y, such that your final answer is in terms of y and x only.
Modelling
- In modelling questions, focus on forming the DE correctly as a basis for your answer.
1. Define all variables that will be used, and are not already defined in the question.
2. Identify the rate of change, say
dx
dt
. Then the L.H.S. of the DE will be
dx
dt
.
3. Rate of change = Rate of increase ÷ Rate of decrease,
where Rate of increase and Rate of decrease are both positive.
- If
dx
dt
is proportional to something, e.g.
2
x t , then
2
dx
x t
dt
· . Write:
2
dx
kx t
dt
= , where k is a positive constant.
- If
dx
dt
is inversely proportional to something, e.g.
2
x t , then
2
1 dx
dt x t
· . Write:
2
dx k
dt x t
= , where k is a positive constant.
- Look out for key phrases in the question such as “water flows out at a rate of …”, which
indicates a rate of decrease, or “is proportional to”, which indicates proportionality.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
56
Example 1 [N08/I/4]
(i) Find the general solution of the differential equation
2
3
1
dy x
dx x
=
+
(ii) Find the particular solution of the differential equation for which 2 y = when 0 x = .
(iii) What can you say about the gradient of every solution curve as x ÷±·?
(iv) Sketch, on a single diagram, the graph of the solution found in part (ii), together with
2 other members of the family of solution curves.
Solution Comments
(i)
2 2
3 3 2
1 2 1
dy x x
dx x x
´ }
= =
]
+ +
' ¹
2
3 2
2 1
x
y dx
x
=
+
¦
2
3
ln( 1) ,
2
x c = + +
The DE is of the form . )
dy
f x
dx
= , so just
integrate both sides. Remember the arbitrary
constant c .
where c is an arbitrary constant.  Write this line for proper presentation.
(ii) When 2 y = , 0 x = ,
2
3
ln( 1)
2
c y x = ÷ +
3
2 ln1
2
= ÷
2 =
1. Substitute the initial conditions into the
general solution.
2. Find the value of the arbitrary constants.
2
3
ln( 1) 2
2
y x = + + 
3. Rewrite the general solution with the
known values for the arbitrary constants.
(iii) As x ÷±·, the gradient of every
solution curve tends to 0 . 
The gradient of every solution curve is
2
3
1
dy x
dx x
=
+
, from the DE. As x ÷±· ,
because the denominator
2
1 x + will “grow
faster” than the numerator 3x , 0
dy
dx
÷ .
OR
Use the GC to plot the graph of
2
3
1
x
y
x
=
+
.
Observe the asymptote 0 y = as x ÷±·.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
57
(iv) Choose 2 other solution curves which are
representative of all values of c . Since
2 c = is a required curve, choose 0 c = and
some 0 c < , here 2 c = ÷ .
Use the GC to draw the graphs of the
solution curves.
When the general solution is of the form
. ) y f x c = + , the solution curves are
translations of each other along the y-axis.
The gradient is 0 at 0 x  , since
2
3
0
1
dy x
dx x
= =
+
. At some zoom levels on
the GC, e.g. ZStandard, the curve looks like
a “V” with sharp tangents at 0 x = .
The graph has no horizontal asymptote,
even though from part (iii), the gradient of
every solution curve tends to 0 as x ÷±·.
Example 2 [NYJC08/Prelim/II/1 (Part)]
Show that the differential equation
2
2 3 0
dy
x xy
dx
÷ + =
may be reduced by means of the substitution
2
y ux = to
4
3 du
dx x

Hence, or otherwise, find the general solution for y in terms of x .
Solution Comments
2
y ux =
. ) . )
2
2
2
2
dy du
u x x
dx dx
du
ux x
dx
= +
= +
1. Differentiate the given substitution with
respect to x to get
dy
dx
in terms of
du
dx
, u
and x .
y
x
0
2 ÷
2
. )
2
3
ln 1 2
2
y x = + + , 2 c =
2
3
ln( 1)
2
y x = + , 0 c =
2
3
ln( 1) 2
2
y x = + ÷ , 2 c = ÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
58
. )
2 2 2
2 2 3 0
du
x ux x x ux
dx
´ }
+ ÷ + =
]
' ¹
4 3 3
2 2 3
du
x ux ux
dx
+ ÷ = ÷
4
3 du
dx x
= ÷  (shown)
2. In the original DE, express y and
dy
dx
in
terms of
du
dx
, u and x .
3. Convert the substituted DE into the form
required.
4
3 du
dx x

4
4
3
3 u dx x dx
x
÷
= ÷ = ÷
¦ ¦
3
3
3
x
c
÷
´ }
= ÷ +
]
÷
' ¹
,
4. Solve the DE.
The DE is of the form . )
dy
f x
dx
= , so just
integrate both sides. Remember the arbitrary
constant c .
where c is an arbitrary constant. Write this line for proper presentation.
3
1
u c
x
= +
2
2
y
y ux u
x
= ¬ =
2 3
1 y
c
x x
= +
2
1
y cx
x
= + 
Express u in terms of x and y, such that your
final answer is in terms of y and x only.
Example 3 [HCI08/Prelim/I/7]
A hospital patient is receiving a certain drug through a drip at a constant rate of 50 mg per
hour. The rate of loss of the drug from the patient’s body is proportional to x , where x (in
mg) is the amount of drug in the patient at time t (in hours). Form a differential equation
connecting x and t and show that
. )
1
50 ,
kt
x Ae
k
÷
= ÷
where A and k are constants.
Given that
1
20
k = ,
(i) Find the time needed for the amount of drug in the patient to reach 200 mg, if initially
there is no trace of this drug in the patient’s body.
(ii) When there is 80 mg of the drug in the patient’s body, the drip is disconnected.
Assuming that the rate of loss remains the same, find the time taken for the amount of
drug in the patient to fall from 80 mg to 20 mg.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
59
Solution Comments
50 ,
dx
kx
dt
= ÷
where k is a positive constant. 
Rate of increase is constant at 50 . Rate of
decrease is proportional to x , so write it as
kx , where k is a positive constant.
Now, Rate of change
= Rate of increase ÷ Rate of decrease
1
1
50
dx
kx dt
´ }
=
]
÷
' ¹
1
1
50
dx dt
kx
=
÷
¦ ¦
1
1
50
k
dx dt
k kx
÷
÷ =
÷
¦ ¦
The DE is of the form . )
dy
f y
dx
= , so divide
both sides by ( ) f y , then integrate both
sides. Remember the arbitrary constant c .
1
ln 50 , kx t c
k
÷ ÷ = +
where c is an arbitrary constant.
ln 50 kx kt kc ÷ =÷ ÷
50
kt kc
kx e
÷ ÷
÷ =
Evaluate both integrals and introduce the
arbitrary constant in 1 step.
. )
50 ,
kc kt kt
kx e e Ae
÷ ÷ ÷
÷ = ± =
where
kc
A e
÷
= ± is an arbitrary constant.
50
kt
kx Ae
÷
= ÷
. )
5
1
0
kt
x Ae
k
÷
= ÷  (shown)
The modulus is eliminated by transferring it
to the arbitrary constant
kc
A e
÷
= ± .
(i)
20
20 50
t
x Ae
÷ ´ }
= ÷
]
' ¹
When 0 t = , 0 x = ,
. )
0
0 20 50 Ae = ÷
Use the information given:
1
20
k = . Interpret
the question: “Initially” means 0 t = , “no
trace” means 0 x = .
1. Substitute the initial conditions into the
general solution.
50 A = 2. Find the value of the arbitrary constant.
20
20 50 50
t
x e
÷ ´ }
= ÷
]
' ¹
3. Rewrite the general solution with the
known value for the arbitrary constant.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
60
When 200 x = ,
20
200 1000 1
t
e
÷ ´ }
= ÷
]
' ¹
20
1
1
5
t
e
÷
÷ =
20
4
5
t
e
÷
=
4
ln
20 5
t ´ }
÷ =
]
' ¹
4
20ln 4.46
5
t
´ }
= ÷ =
]
' ¹
hours (3 s.f.) 
Interpret the question: Find t when 200 x = .
(ii) 1
20
dx
x
dt
= ÷
20
1
dx
x dt
´ }
÷ =
]
' ¹
1
20 1 dx dt
x
÷ =
¦ ¦
“Rate of loss remains the same”, but the drip
is disconnected, hence the rate of increase is
0 .
With the given value of k , rewrite the DE
and solve it.
20ln x t B ÷ = + ,
where B is an arbitrary constant.
Evaluate both integrals and introduce the
arbitrary constant in 1 step.
Remember the arbitrary constant, but since c
is already used, use another letter, say B .
When 0 t = , 80 x = ,
20ln80 B = ÷
20ln 20ln80 x t ÷ = ÷
When 20 x = ,
20ln80 20ln20
20ln4
27.7 hours (3 s.f.)
t = ÷
=
= 
Interpret the question: Initially, there is 80
mg of drug, so 0 t = and 80 x = . Find t
when 20 x = .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
61
Power Series
If a function . ) f x can be expanded as a power series:
2
( )
( ) (0) '(0) ''(0) (0)
2! !
n
n
x x
f x f xf f f
n
        (In MF15)
where
( )
(0)
n
f is the n th derivative of ( ) f x , i.e.
( )
n
n
d f x
dx
, evaluated at 0 x = .
The expression on the RHS is known as the Maclaurin’s series, a type of power series.
- In the formula above, x may be replaced with any function of x .
- The Maclaurin’s series of the following functions are also in MF15:
o   1
n
x  for 1 x < (This is the binomial expansion for rational index)
o
x
e , sin x and cos x for all x
o ln(1 ) x  for 1 1 x ÷ < < (Note the range of valid x values)
- Maclaurin’s series are infinite series.
- The n th termin the Maclaurin’s series is
. )
. ) 0
!
n
n
x
f
n
FAQ
Q:
Find the Maclaurin’s series for   f x , up to the term in
k
x .
A: 1. Differentiate the expression repeatedly. Implicit differentiation may be helpful (or
necessary if the question states “By further differentiation of this result”).
Find equations connecting all derivatives up to the k th derivative
 
k
k
d f x
dx
.
2. Evaluate . ) f x and its derivatives at 0 = x using the equations found. When 0 x = ,
- ( ) (0) f x f =
- '(0)
dy
f
dx
=
- . )
2
2
'' 0
d y
f
dx
= , etc. up to
. )
. )
. ) 0
k
k
k
d f x
f
dx
= .
3. Substitute (0) f , '(0) f , ''(0) f etc. into the formula for Maclaurin’s series.
Write out each term in ascending powers of x , up to
k
x .
4. Check your answer by substitute a small value of x , e.g. 0.01 x = into both the
original expression and the Maclaurin’s series.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
62
Small Angle Approximations
Using the Maclaurin’s series of sin x , cos x and tan x up to the term in
2
x ,
sin kx kx ~
2
( )
cos 1
2
kx
kx ~ ÷
tan kx kx ~
where k is a constant. The approximation is accurate when x is small, i.e. x is close to 0 .
- To approximate e.g. sin
4
x
 ´ }
+
]
' ¹
where x is small, do NOT write sin
4 4
x x
  ´ }
+ ~ +
]
' ¹
,
since
4
x

+ is not small. Instead, use the addition formula for sin . Since x is small, the
approximations for sin x and cos x are valid.
2
2
sin sin cos cos sin
4 4 4
2 2
1
2 2 2
2
1
2 2
x x x
x
x
x
x
   ´ }
+ = +
]
' ¹
´ }
~ + ÷
]
' ¹
´ }
= ÷ +
]
' ¹
Example 1 [HCI08/Promo/12]
Given that
1
tan y x x
÷
= , prove that
. )
2
2
2
1 2 2 2 0
d y dy
x x y
dx dx
+ + ÷ ÷ =
By repeated differentiation, show that the first two non-zero terms of the Maclaurin’s series
for y are
2 4
1
3
x x ÷ .
Hence evaluate
1
3
0
tan
lim
x
x x
x
÷
÷
´ } ÷
]
' ¹
.
Solution Comments
1
tan y x x
÷
=
1
2
tan
1
dy x
x
dx x
÷
= +
+
Differentiate y to get
dy
dx
.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
63
. ) . )
2 2 1
1 1 tan
dy
x x x x
dx
÷
+ = + +
. )
2
2 1
2
1 2 1 2 tan 1
d y dy
x x x x
dx dx
÷
+ + = + +
. )
2
2
2
1 2 0 2 2
d y dy
x x y
dx dx
+ + ÷ ÷ =  (proven)
Since the expression to be proven contains
. )
2
2
2
1
d y
x
dx
+ , manipulate the equation to
contain
. )
2
1
dy
x
dx
+ , then use implicit
differentiation.
. )
3 2 2
2
3 2 2
1 2 2 2 2 0
d y d y d y dy dy
x x x
dx dx dx dx dx
+ + + + ÷ =
. )
3 2
2
3 2
1 4 0
d y d y
x x
dx dx
+ + =
. )
4 3 3 2
2
4 3 3 2
1 2 4 4 0
d y d y d y d y
x x x
dx dx dx dx
+ + + + =
. )
4 3 2
2
4 3 2
1 6 4 0
d y d y d y
x x
dx dx dx
+ + + =
Use the Product Rule and Chain Rule, and
remember:
. )
d dy
y
dx dx
=
2
2
d dy d y
dx dx dx
´ }
=
]
' ¹
2 3
2 3
d d y d y
dx dx dx
´ }
=
]
' ¹
, etc.
Since the question requires the series up to
4
x , find equations connecting all
derivatives up to
4
4
d y
dx
.
When 0 x = ,
2 3 4
2 3 4
0, 0, 2, 0, 8
dy d y d y d y
y
dx dx dx dx
= = = = = ÷
2 4
3
0 0 2 0 8
2! 4!
x x
y x x
´ } ´ }
= + + + ÷ +.
] ]
' ¹ ' ¹
2 4
1
3
y x x = ÷ +. (shown)
Evaluate y and its derivatives at 0 = x ,
using the equations found above.
Substitute the resulting values into the
formula for Maclaurin’s series.
1 2 1
3 4
tan tan x x x x x
x x
÷ ÷
÷ ÷
=
2 2 4
4
1
3
x x x
x
´ }
÷ ÷ +.
]
' ¹
=
4
4
1
3
x
x
+.
=
Convert the expression to contain
1
tan x x
÷
. Substitute the series for
1
tan x x
÷
which has been found into the
expression.
4
1
3 4
0 0
1
tan 1
3
lim lim
3
x x
x
x x
x x
÷
÷ ÷
´ }
+.
]
´ } ÷
= =
] ]
' ¹
]
' ¹

As x tends to 0 , the
5
x ,
6
x etc. terms get
smaller faster than the
4
1
3
x term, thus the
4
1
3
x term dominates the numerator.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
64
Example 2 [NYJC08/Prelim/II/3 (Part)]
A curve C is defined by the equation
2
2 1
dy
y y
dx
= ÷ and (0, 2) is a point on C .
(i) Show that at 0 x = ,
3
3
57
256
d y
dx
= .
Find the Maclaurin’s series of y up to and including the term in
3
x .
(ii) Hence find the series expansion of
y
e , up to and including the term in
2
x .
Solution Comments
(i)
2
2 1
dy
y y
dx
= ÷
2
2
2
2 2 2
d y dy dy
y y
dx dx dx
´ }
+ =
]
' ¹
2
2
2
d y dy dy
y y
dx dx dx
´ }
+ =
]
' ¹
3 2 2
3 2 2
2
d y dy d y dy d y
y
dx dx dx dx dx
´ } ´ }
´ } ´ }
+ +
] ] ] ]
' ¹ ' ¹
' ¹ ' ¹
2
2
2
d y dy
y
dx dx
´ }
= +
]
' ¹
2
3 2 2
3 2 2
3
d y dy d y d y dy
y y
dx dx dx dx dx
´ }
´ } ´ }
+ = +
] ] ]
' ¹ ' ¹
' ¹
When 0 x = ,
3
2,
4
dy
y
dx
= =
2
2 2
2 2
3 3 15
4 2 4
4 4 32
d y d y
dx dx
´ } ´ }
+ = ¬ =
] ]
' ¹ ' ¹
2
3
3
3 15 15 3
2 3 2
4 32 32 4
d y
dx
´ }´ } ´ } ´ }
+ = +
] ] ] ]
' ¹' ¹ ' ¹ ' ¹
3
3
57
2 6 5
d y
dx
=  (shown)
Differentiate the equation of C
repeatedly. Since the question requires
the series up to
3
x , find equations
connecting all derivatives up to
3
3
d y
dx
.
Since differentiation here is with respect
to x , by the Product Rule and Chain
Rule:
2
2
2
d dy d y dy
y y
dx dx dx dx
´ }
´ } ´ }
= +
] ] ]
' ¹ ' ¹
' ¹
Evaluate y and its derivatives at 0 = x ,
using the equations found above.
2 3
3 15 57
2
4 32 2! 256 3!
x x
y x
´ } ´ }
= + + + +.
] ]
' ¹ ' ¹
2 3
3 15 19
2
4 64 512
y x x x = + + + +. 
Substitute the resulting values into the
formula for Maclaurin’s series.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
65
(ii)
2 2
3 15 3 15
2
2
4 64 4 64
x x x x
y
e e e e
+ + +. + +. ´ }
= =
]
' ¹
2
2 2 2
3 15 1 3 15
1
4 64 2! 4 64
e x x x x
]
´ }
= + + + + +.
]
]
' ¹
]
]
2
2 2 2
3 15 1 3
1
4 64 2 4
e x x x
]
´ }
= + + + +.
]
]
' ¹
]
]
2 2 2 2
3 15
64 4
e e x e x = + + +. 
Substitute the series for y which has
been found into the expression.
2
e must
be extracted. Otherwise,
2 3
1 1
1
2! 3!
y
e y y y = + + + +.
will have infinitely many constant terms,
since each y contains a 2 . Hence, this
series cannot be simplified.
Example 3 [YJC08/Prelim/I/4iii]
It is given that
. )
tan 1
x
y e = ÷ .
(i) Prove that
. )
2
2
1 2
x
d y dy
e y
dx dx
= + .
(ii) Find Maclaurin’s series for y in ascending powers of x , up to and including the term
in
3
x .
(iii) Deduce the first three terms of the Maclaurin’s series for
. )
2
sec 1
x
x
e
e
÷
÷
.
We will focus on part (iii) only.
From part (i):
 
2
sec 1
x x
dy
e e
dx
 
From part (ii):
2 3
1 1
2 2
y x x x    
Solution Comments
(iii)
. )
2
sec 1
x x
dy
e e
dx
= ÷
2 3
1 1
2 2
d
x x x
dx
´ }
= + + +.
]
' ¹
2
3
1
2
x x = + + +.
dy
dx
looks similar to
2
sec ( 1)
x
x
e
e
÷
÷
. To obtain the series
of
dy
dx
, differentiate both sides of
2 3
1 1
2 2
y x x x = + + +.
. )
. )
2
2
sec 1
sec 1
x
x x
x
e
e e
e
÷
÷
÷ =
. ) . )
2 3
1
2
x x = + ÷ + ÷ +.
2
3
1
2
x x = ÷ + +. 
Since
2
sec ( 1)
x
x
e
e
÷
÷
has an
x
e
÷
term, substitute x with
– x .
Questions on binomial expansion or power series
commonly include differentiation and integration
concepts.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
66
Binomial Expansion
Binomial Expansion for Rational Index
 
     
2
1 1 1
1 1
2! !
n
r
n n n n n r
x nx x x
r
   
      

  (In MF15)
where is either a non-integer rational number or a negative integer, and 1 x < .
- The expansion is an infinite series.
- For the expansion to be valid, 1 x < .
- Use the GC to check your answer by substituting some within the range of validity. The
actual value and the value given by the expansion should be approximately equal.
- Knowing “shortcut” expansions (NOT in MF15) can be convenient. For 1 x < ,
. )
. ) . )
. ) . )
1
2 3
1
2 3
2
2 3
1 1
1 1 1
1 1 2 3 4 1
r
r
r
r
x x x x x
x x x x x
x x x x r x
÷
÷
÷
÷ = + + + + + +
+ = ÷ + ÷ + + ÷ +
÷ = + + + + + + +
. .
. .
. .
FAQ
Q1:
Expand
 
n
c
a bx  in ascending powers of x up to the term in
k
x ,
where a and b are constants and
c
x is a positive power of x .
A1:
1. Extract
n
a to convert the expression such that it contains 1 as a term.
. )
1
n
c
n
c n
bx
a bx a
a
]
+ = +
]
]
2. Use the binomial expansion for rational index, including only terms up to
k
x .
Use the above “shortcut” expansions if possible to save time.
Write out the entire expression before simplifying to reduce careless mistakes.
3. Simplify the coefficients.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
67
Q2:
State the range of values of x for which the expansion of
 
n
c
a bx  is valid.
A2: 1. Look at the expression immediately before the binomial expansion was used.
It should be of the form [1 ( )]
n
f x  .
2. For the expansion of [1 ( )]
n
f x + ,
the range of validity is ( ) 1 f x < , i.e.
1 ( ) 1 f x   
3. Simplify the inequality to give a range of values of x.
Q3:
Expand an expression in ascending powers of x up to the term in
k
x .
A3:
1. Express the expression as a product (or sum) of terms of the form
 
n
c
a bx  ,
using partial fractions if required.
2. For each factor, follow the 3 steps in Answer (1) above.
3. Multiply (or add) the expansions for each term together.
4. Discard any terms which have a power of x greater than k.
Q4: State the range of values of x for which the expansion of an expression is valid.
A4: 1. Using Answer (2) above, find the range of values of x for which the expansion of
each termof the form
 
n
c
a bx  is valid.
2. The set of values of x for which the entire expression is valid is then the
intersection of the sets of values of x for each term.
Q5:
Expand
 
n
c
a bx  in descending powers of x up to the term in
k
x .
A5: Same as Answer (1) above, but for Step 1:
Extract
 
n
c
bx instead of
n
a to convert each factor such that it contains 1 as a term.
. ) . )
1
n
n n
c c
c
a
a bx bx
bx
]
+ = +
]
]
Q6: Find the r th term (or its coefficient) in the expansion of an expression.
A6:
Use the fact that the . ) 1
th
r + term in the expansion is
. ) . ) 1 1
!
r
n n n r
x
r
÷ ÷ + .
Q7: By putting = x a , find an approximation for b ,
where a and b are constants.
A7: 1. Ensure a is within the range of values of x such that the expansion is valid.
2. Substitute x a = into both the original expression and the expansion.
3. Manipulate the equation to obtain an approximation for b .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
68
Example 1 [N2006/I/12 (Modified)]
(i) Express
. ) . )
2
2
1 2
( )
2 1
x x
f x
x x
+ ÷
=
÷ +
in partial fractions.
(ii) Expand ( ) f x in ascending powers of x, up to and including the term in
2
x .
(iii) Find the set of values of x for which the expansion is valid.
(iv) Deduce the equation of the tangent to the curve ( ) y f x = at the point where 0 x = .
Solution Comments
(i) (working omitted)
2
1 1
( )
1 2
x
f x
x x
+
= ÷
+ ÷

Use the partial fractions decomposition
for a non-repeated quadratic factor in the
denominator (in MF15).
(ii)
. ) . ) . )
1
1
2
( ) 1 1 2 f x x x x
÷
÷
= + + ÷ ÷
Express ( ) f x as a sum of terms of the
form
. )
n
c
a bx + .
. ) . )
1
1
2 1
1 1 2 1
2
x
x x
÷
÷
÷
´ }
= + + ÷ ÷
]
' ¹
. ) . )
1
1
2
1
1 1 1
2 2
x
x x
÷
÷
´ }
= + + ÷ ÷
]
' ¹
For each term,
1. Extract
n
a to convert the expression
such that it contains as a term.
. ) . )
2
1 1 x x = + ÷ +.
2
1 1 1
1
2 2 2
x x
]
´ }
÷ + + +
]
]
' ¹
]
]
.
2. Use the binomial expansion for
rational index, including only terms
up to
2
x .
Use the “shortcut” expansions of
. )
1
1 x
÷
÷ and . )
1
1 x
÷
+ to save time.
. ) . )
2
1 1 x x = + ÷ +.
2
1 1 1
...
2 4 8
x x
´ }
+ ÷ ÷ ÷ +
]
' ¹
3. Simplify the coefficients.
2 2
1 1 1
1
2 4 8
x x x x = ÷ + ÷ ÷ ÷ +.
Ignore
. )
2 3
x x x ÷ = ÷ when simplifying,
since the expansion should be up to
2
x .
2
1 3 9
2 4 8
x x = + ÷ +. 
Write “” after the first few terms of an
expansion (after any expansion in any
step).
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
69
(iii)
1
1
2
x ÷ < and
2
1 x <
2 x < and 1 x <
2 2 x ÷ < < and 1 1 x ÷ < <
For the expansion to be valid,
Set of values of { : 1 1} x x x = e ÷ < < ¯ 
Immediately before the binomial
expansion was used, the expressions were
1
1
1
2
x
÷
´ }
÷
]
' ¹
and
. )
1
2
1 x
÷
+ .
For the entire expansion to be valid, both
binomial expansions involved must be
valid, i.e.
1
1
2
x ÷ < and
2
1 x < .
Find the intersection of the 2 sets of
values, and give the final answer in set
notation, since the question specified “set
of values”.
Use set-builder notation, NOT interval
notation i.e. . ) 1,1 ÷ .
(iv)
2
1 3 9
2 4 8
y x x = + ÷ +.
3 9
4 4
dy
x
dx
= ÷ +.
When 0 x = ,
1
2
y = and
3
4
dy
dx
= .
Equation of required tangent:
. )
1 3
0
2 4
y x ÷ = ÷
3 1
4 2
y x = + 
To find the gradient of the tangent, find
dy
dx
when 0 x = . To find the point of
tangency, find y when 0 x = .
A straight line with gradient m and
passing through a point with coordinates
. )
1 1
, x y has equation:
1 1
( ) y y m x x ÷ = ÷
Questions on binomial expansion or
power series commonly include
differentiation and integration concepts.
Example 2 [HCI08/Prelim/I/6]
Expand
1
1
n
x
x
÷ ´ }
]
+
' ¹
in ascending powers of x up to and including the term in
2
x .
State the set of values of x for which the series expansion is valid.
Hence find an approximation to the fourth root of
19
21
, in the form
p
q
, where p and q are
positive integers.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
70
Solution Comments
. ) . )
1
1 1
1
n
n n x
x x
x
÷ ÷ ´ }
= ÷ +
]
+
' ¹
Express . ) f x as a product of terms of
the form
. )
n
c
a bx + .
. )
. )
. )
2
1
1
2!
n n
n x x
] ÷
= + ÷ + ÷ +
]
]
.
. )
. ) . )
2
1
1
2!
n n
n x x
] ÷ ÷ ÷
× + ÷ + +
]
]
.
. )
2
1
1
2
n n
nx x
] ÷
= ÷ + +.
]
]
. )
2
1
1
2
n n
nx x
] +
× ÷ + +.
]
]
Use the binomial expansion for rational
index, including only terms up to
2
x .
Write out the entire expression before
simplifying to reduce careless mistakes.
. ) . )
2 2 2 2
1 1
1
2 2
n n n n
nx x nx n x x
+ ÷
= ÷ + ÷ + + +.
2 2
1 2 2 nx n x = ÷ + +. 
Choose 1 from the first bracket and
multiply it with 1, nx ÷ , and
. )
2
1
2
n n
x
+
in the second bracket in turn. Do the
same with the other 2 terms in the first
bracket.
While doing so, discard any terms which
have a power more than 2, since the
expansion should be up to
2
x .
For the expansion to be valid, 1 x < .
Set of values of { : 1 1} x x x = e ÷ < < ¯ 
Immediately before the binomial
expansion was used, the expressions
were . ) 1
n
x ÷ and . ) 1
n
x
÷
+ .
For the entire expansion to be valid, both
binomial expansions involved must be
valid, i.e. – 1 x < and 1 x < , but both
inequalities are equivalent to 1 x < .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
71
Let
1
20
x = and
1
4
n = .
Since
1
1
20
x = < , the expansion is valid.
1
4
1
1
1
20
1
1
1
20
n
x
x
´ }
÷
]
÷ ´ }
=
]
]
+
' ¹
]
+
' ¹
1
4
20 1
20 1
÷ ´ }
=
]
+
' ¹
4
19
21
=
1
4
n = , since
19
21
raised to the power
1
4
is the fourth root of
19
21
. It remains to
make
19 1
,
21 1
x
x
÷
=
+
with 1 x < for the expansion to be valid.
Solve for x directly to get
1
20
x = .
OR
Try to write
19
21
in a similar form as
1
1
x
x
÷
+
. You may come up with
20 1
20 1
÷
+
,
which suggests dividing both top and
bottom by 20 to get
1
1
20
1
1
20
÷
+
Hence,
1
20
x = .
2 2
1 1 1 1
1 2 2
4 20 4 20
´ }´ } ´ } ´ }
~ ÷ +
] ] ] ]
' ¹' ¹ ' ¹ ' ¹
3121
3200
= , where 3121 p = and 3200 q = . 
Substitute
1
20
x = back into the
expansion. Since only the first few terms
are taken, write “ ~ ” instead of “ = ”
where the approximation occurs.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
72
Vectors
When representing vectors as single letters, write a squiggle underneath: a
¯
.
Basic Formulae
Distance between
2 points
If 2 points have position vectors
1
2
3
a
a
a
´ }
]
=
]
]
' ¹
a and
1
2
3
b
b
b
´ }
]
=
]
]
' ¹
b , then:
. ) . ) . )
2 2 2
1 1 2 2 3 3
AB a b a b a b = ÷ + ÷ + ÷

Unit vector A unit vector is a vector with modulus 1, usually used to specify
direction. The unit vector ˆ a of a vector a is given by:
ˆ =
a
a
a
Ratio theorem If the point P divides AB such that : : AP PB   = , then:
OA OB
OP
 
 
+
=
+

In particular, if P is the midpoint of AB , then
2
OA OB
OP
+
=

.
Angle, , between:
2 vectors a and b :
1
cos 
÷
´ }

=
]
]
' ¹
a b
a b
2 lines
1
: l  = +
1 1
r a b ,
2
: l  = +
2 2
r a b :
1
cos 
÷
´ }
=
]
]
' ¹
1 2
1 2
b b
b b
2 planes
1 1
: p  =
1
r n ,
2 2
: p  =
2
r n :
1
cos 
÷
´ }
=
]
]
' ¹
1 2
1 2
n n
n n
a line : l  = + r a b and
a plane : p  = r n :
1
sin 
÷
´ }
=
]
]
' ¹
b n
b n
Between 2 lines, 2 planes, or a line and a plane, there are generally 2
angles, 1 acute and 1 obtuse. The modulus around the dot product in
finds the acute angle.
Between 2 vectors, there is only 1 unique angle, thus there is no
modulus around the dot product.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
73
Equations of Lines & Planes
Line Vector Cartesian
 = + r a b, e¯
Make  the subject for each
of the 3 equations, then
combine them to get the
Cartesian form.
3 1 2
1 2 3
z a x a y a
b b b
÷ ÷ ÷
= =
Use
1
2
3
a
a
a
´ }
]
=
]
]
' ¹
a and
1
2
3
b
b
b
´ }
]
=
]
]
' ¹
b to get vector form.
Plane Vector Cartesian Scalar Product
  = + + r a b c , ,  e¯
Use × = b c n to find a normal
to the plane, then p = a n to
get the scalar product form.
1 2 3
n x n y n z p + + =
Set special conditions, e.g.
0 x y = = to find 3 points
A , P , Q on the plane.
The vector form is then
OA AP AQ   = + + r

.
p = = r n a n
Expand the dot product
to get the Cartesian
equation.
Dot Product VS Cross Product
Dot Product ( a b) Cross Product ( × a b)
Output Scalar Vector perpendicular to both a and
b
Formula
cos = a b a b , where  is the
angle between a and b .
Magnitude: sin a b
Direction: Right-Hand Grip Rule
Calculation
1 1
2 2 1 1 2 2 3 3
3 3
a b
a b a b a b a b
a b
´ } ´ }
] ]
= + +
] ]
] ]
' ¹ ' ¹
2
= a a a
1 1 2 3 3 2
2 2 3 1 1 3
3 3 1 2 2 1
a b a b a b
a b a b a b
a b a b a b
÷ ´ } ´ } ´ }
] ] ]
× = ÷
] ] ]
] ] ]
÷
' ¹ ' ¹ ' ¹
Lengths & Areas
Say the vectors a , b , c form a right-angled triangle, as shown, and
you know the vectors a and b . You can then find the lengths AC
and BC.
Let A  . = .
Length AC
Length of projection =
of AB onto b
cos AB  =
ˆ
= a b
Length BC
Perpendicular distance =
from B to b
sin AB  =
ˆ
= × a b
A
C B
a

b
c
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
74
Let the points A and B have position vectors a
and b respectively. Let C be the point such that
OACB is a parallelogram.
Area of triangle
1
2
OAB = × a b
Area of parallelogram OACB = × a b
Foot of Perpendicular & Perpendicular Distance
Say a point P has position vector
p , and a line l has equation
 = + r a b, as shown. Say point A
lies on l .
From P to l :
Foot of perpendicular, F , with position vector f :
- Since F lies on l ,  = + f a b for some e¯.
- Since PF

and l are perpendicular, . ) 0 ÷ = f p b .
- . ) 0  + ÷ = a b p b
- Solve for  to find f .
Perpendicular distance, PF PF =

:
Use any of these 3 methods.
1. Find f as above.
PF = ÷ f p
2. Find AP and AF .
By Pythagoras’ Theorem,
2 2
PF AP AF = ÷ .
3.
´
PF AP = ×b

Say a point P has position vector p ,
and a plane  has equation p = r n ,
as shown. Say point A lies on  .
From P to :
Foot of perpendicular, F , with position vector f :
- Since PF

is perpendicular to  , PF

is parallel
to n. Also, since P lies on PF

:
:
PF
l  = + r p n , e¯.
- Since F lies on
PF
l ,  = + f p n for some e¯.
- Since F lies on  , p = f n
- . ) p  + = p n n
- Solve for  to find f .
Perpendicular distance, PF PF =

:
-
´
PF AP = n

O
A
C
B b
a
: p  = r n
P
F A
n
P
F A
: l  = + r a b
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
75
SLE denotes system of linear equations. Use PlySmlt2 on the GC to solve SLEs.
- 2 Vectors, a and b
Parallel: The SLE given by  = a b has a non-zero real solution for  .
Perpendicular: 0 = a b
- 2 Lines in 3 dimensions,
1
: l  = +
1 1
r a b and
2
: l  = +
2 2
r a b
Parallel:
1
b and
2
b are parallel (see above).
Intersecting at a point:
-
1
b and
2
b are NOT parallel (see above), AND
- The SLE given by   + = +
1 1 2 2
a b a b has a unique real
solution for  and  .
Point of intersection, P : If the solutions are
0
 and
0
 ,
0
OP  = +
1 1
a b

(or
0
 +
2 2
a b ).
Skew:
-
1
b and
2
b are NOT parallel (see above), AND
- The SLE given by   + = +
1 1 2 2
a b a b has no real solution
for  and  .
- 3 Points, A , B , C
Collinear
(on the same line):
AB

, AC

, BC

are all parallel.
You may pick any 2 and show that they are parallel (see above).
- Point, Q , with position vector
1
2
3
q
q
q
´ }
]
=
]
]
' ¹
q
Lies on a line : l  = + r a b: The SLE given by  = + q a b has a real solution
for  .
Lies on a plane : p  = r n :
p = q n
Lies on a plane :    = + + r a b c : The SLE given by   = + + q a b c has a real
solution for  and  .
Lies on a plane
1 2 3
: n x n y n z p  + + = :
1 1 2 2 3 3
n q n q n q p + + =
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
76
- Line & Plane, : l  = + r a b and : p  = r n
l lies on  : 0 = b n AND p = a n
l is parallel to  ,
but does not lie on  :
0 = b n AND p = a n
l intersects  at one point: 0 = b n
l is perpendicular to  if b and n are parallel (see above).
Point of intersection, P : Solve . ) p  + = a b n for  . If
the solution is
0
 ,
0
OP  = + a b

- 2 Planes,
1
: p  = r n and
2
: q  = r m , where
1
2
3
n
n
n
´ }
]
=
]
]
' ¹
n and
1
2
3
m
m
m
´ }
]
=
]
]
' ¹
m
Parallel: n and m are parallel (see above).
Intersecting along a line: n and m are NOT parallel (see above)
Line of intersection, l : Expand the dot products into
Cartesian equations to obtain the SLE:
1 2 3
1 2 3
n x n y n z p
m x m y m z q
+ + = ¹
´
+ + =
¹
If PlySmlt2 on the GC solves the SLE to give:
1 1 1 3
2 2 2 3
3 3
x a b x
x a b x
x x
= +
= +
=
where
1
x ,
2
x ,
3
x represent x , y , z respectively and
1
a ,
2
a ,
1
b ,
2
b are constants,
1 1
2 2
0 1
a b
l a b 
´ } ´ }
] ]
= +
] ]
] ]
' ¹ ' ¹
, e¯
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
77
- 3 Planes
Use their Cartesian equations to obtain a SLE with 3 equations in 3 variables: x , y , z .
Do not intersect
at any common point or line:
No real solution for x , y , z .
Intersect at one common point: Unique solution for x , y , z .
Point of intersection, P : If the solution to the SLE is
0
x ,
0
y ,
0
z ,
0
0
0
x
OP y
z
´ }
]
=
]
]
' ¹

Intersect along one common line: x and y can be solved in terms of z .
Line of intersection, l : See above, where 2 planes
intersect along a line.
Example 1 [TJC06/Prelim/I/14]
The lines
1
l and
2
l have equations
8 5
1
2 2
x y
z
+ ÷
= = + and
6
4
1

´ }
]
= ÷
]
]
' ¹
r respectively, where
e¯.
Find
(i) the acute angle between
1
l and the z -axis.
(ii) the position vector of the foot of the perpendicular, N , from the origin to
1
l .
(iii) the point of intersection of
1
l and
2
l .
Hence, find a vector equation of the reflection of line
2
l in the line
1
l .
Solution Comments
(i)
Let
8 5
1
2 2
x y
z 
+ ÷
= = = + .
Convert Cartesian equations of lines to
vector equations, which are usually much
more useful.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
78
Then:
8 2
5 2
1
x
y
z



= ÷ + ¹
¦
= ÷
´
¦
= ÷ +
¹
1
8 2
: 5 2
1 1
l 
÷ ´ } ´ }
] ]
= + ÷
] ]
] ]
÷
' ¹ ' ¹
r , e¯
Express x , y , z in terms of  .
z -axis:
0
0
1

´ }
]
=
]
]
' ¹
r ,  e¯
z varies along the z -axis, while x and y
do not, and the z -axis passes through the
origin, explaining its vector equation.
Let  be the acute angle between
1
l and
the z -axis.
1
2 0
2 0
1 1
cos 70.5
2 0
2 0
1 1

÷
]
´ } ´ }
]
] ]
÷
]
] ]
] ]
]
' ¹ ' ¹
] = = °
´ } ´ } ]
] ]
]
÷
] ]
]
] ]
]
' ¹ ' ¹
]
(1 d.p.) 
For the acute angle between 2 lines, use:
1
cos 
÷
´ }
=
]
]
' ¹
1 2
1 2
b b
b b
Since we are finding the acute angle, place
a modulus around the dot product.
(ii)
8 2
5 2
1
ON



÷ + ´ }
]
= ÷
]
]
÷ +
' ¹

, for some e¯.
(Since N lies on
1
l )
2 8 2 2
2 5 2 2 0
1 1 1
ON



÷ + ´ } ´ } ´ }
] ] ]
÷ = ÷ ÷ =
] ] ]
] ] ]
÷ +
' ¹ ' ¹ ' ¹

(Since ON

is perpendicular to
1
l )
16 10 1 4 4 0    ÷ ÷ ÷ + + + =
3  =
Expand the dot product, then solve for  .
. )
. )
8 2 3 2
5 2 3 1
1 3 2
ON
÷ + ÷ ´ } ´ }
] ]
= ÷ = ÷
] ]
] ]
÷ +
' ¹ ' ¹


Substitute  into the original equation for
ON

.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
79
(iii) Let P be the point of intersection
between
1
l and
2
l .
Then
8 2 6
5 2 4
1
OP
 
 
 
÷ + ´ } ´ }
] ]
= ÷ = ÷
] ]
] ]
÷ +
' ¹ ' ¹

,
for some ,  e¯.
2 6 8
2 4 5
1
 
 
 
÷ = ¹
¦
÷ + = ÷
´
¦
÷ =
¹
From the GC, 0.5  =÷ , 1.5 =÷ .
Write the system of linear equations such
that LHS has  and  in that order, and
RHS has a constant.
Then, solve using PlySmlt2 on the GC.
. )
. )
6 1.5 9
4 1.5 6
1.5 1.5
OP
÷ ÷ ´ } ´ }
] ]
= ÷ ÷ =
] ]
] ]
÷ ÷
' ¹ ' ¹

Hence, P is . ) 9, 6, 1.5 ÷ ÷ . 
The question asks for “the point”, so give
the coordinates of the required point
instead of its position vector.
Draw a diagram first to illustrate the
question, and label the points clearly.
In a “Hence” question, use information
from previous parts. With reference to the
diagram, part (ii) gives ON

and part (iii)
gives OP

.
Let
3
l be the reflection of
2
l in
1
l .
4
2 2
4
OA ON
÷ ´ }
]
= = ÷
]
]
' ¹

4 9 5
2 6 8
4 1.5 5.5
PA
÷ ÷ ´ } ´ } ´ }
] ] ]
= ÷ ÷ = ÷
] ] ]
] ] ]
÷
' ¹ ' ¹ ' ¹

To obtain the vector equation of
3
l , we
can find 2 points which lie on
3
l . 1 point is
P , and part (ii) gives OP

.
From the diagram, we can use ON

from
part (ii) to find the position vector of A ,
another point on
3
l , since ON AN = .
Now PA

is a direction vector of
3
l .
N P
O
A
1
l
2
l
3
l
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
80
10
1
16
2
11
PA
´ }
]
= ÷
]
]
' ¹

3
4 10
: 2 16
4 11
l 
÷ ´ } ´ }
] ]
= ÷ + ÷
] ]
] ]
' ¹ ' ¹
r , e¯ 
When writing the equation of a line
: l  = + r a b , choose the simplest vector
for a (here, choose OA

over OP

) and
leave b in the simplest integer form. This
makes your working easier in any later
parts.
Example 2 [N08/I/11]
The equations of three planes
1
p ,
2
p ,
3
p are
2 5 3 3
3 2 5 5
5 17
x y z
x y z
x y z  
÷ + =
+ ÷ = ÷
+ + =
respectively, where  and  are constants. When 20.9  =÷ and 16.6  = , find the
coordinates of the point at which these planes meet.
The planes
1
p and
2
p intersect in a line l .
(i) Find a vector equation of l .
(ii) Given that all three planes meet in the line l , find  and  .
(iii) Given instead that the three planes have no point in common, what can be said about
the values of  and  ?
(iv) Find the Cartesian equation of the plane which contains l and the point . ) 1, 1, 3 ÷ .
Solution Comments
When 20.9  = ÷ , 16.6  = ,
2 5 3 3
3 2 5 5
5 20.9 17 16.6
x y z
x y z
x y z
÷ + = ¹
¦
+ ÷ = ÷
´
¦
÷ + =
¹
From the GC,
4
11
x = ÷ ,
4
11
y = ÷ ,
7
11
z =
Coordinates of the point at which
1
p ,
2
p
and
3
p meet:
4 4 7
, ,
11 11 11
´ }
÷ ÷
]
' ¹

Substitute the given values for  and .
Use the Cartesian equations of the 3 planes
to obtain a SLE with 3 equations in 3
variables: x , y , z .
Solve it using GC, then the unique solution
for x , y , z
gives the coordinates of the
point of intersection.
Give the coordinates of the point, not its
position vector.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
81
(i)
2 5 3 3
3 2 5 5
x y z
x y z
÷ + = ¹
´
+ ÷ = ÷
¹
From the GC,
1
1
x
y
z



= ÷ + ¹
¦
= ÷ +
´
¦
=
¹
, where e¯.
1 1
: 1 1
0 1
l 
÷ ´ } ´ }
] ]
= ÷ +
] ]
] ]
' ¹ ' ¹
r , e¯ 
Use the Cartesian equations of the 2 planes
to obtain a SLE with 2 equations in 3
variables: x , y , z .
Solve it using GC. If the GC gives the
solutions for x and y in terms of z , the 2
planes intersect along a line.
Write z as a parameter, say , to get the
vector equation of the line.
(ii)
3
5
:
17
p  
´ }
]
=
]
]
' ¹
r
Since l lies on
3
p , every point which
lies on l also lies on
3
p .
1 1 5
1 1
0 1 17
  
÷ ] ´ } ´ } ´ }
] ] ] ]
÷ + =
] ] ] ]
] ] ]
]
' ¹ ' ¹ ' ¹ ]
, for all e¯.
5 5 0 17      ÷ + ÷ + + + =
. ) 22 5     + ÷ ÷ =
1
p and
2
p already intersect in the line l ,
with equation found in part (i). If all 3
planes intersect in l , l must lie on
3
p .
The method shown here uses the fact that
for each value of e¯,
1 1
1 1
0 1

÷ ´ } ´ }
] ]
= ÷ +
] ]
] ]
' ¹ ' ¹
s
gives the position vector of a point on l .
Since the above equation is true for all
e¯, comparing coefficients,
22 0
5

 
+ = ¹
´
÷ ÷ =
¹
22  = ÷
5 17   = + = ÷
22  = ÷ , 17  = ÷ 
Consider . ) 22 5     + ÷ ÷ = to have 
as variable, with  and  constant. Since
every point on l lies on
3
p , every value of
e¯ fulfils the equation.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
82
(iii)
1
p ,
2
p and
3
p have no point in common
if and only if l is parallel to
3
p and l
does not intersect
3
p .
1
p and
2
p already intersect in the line l .
Case 1: l is not parallel to
3
p , then l and
3
p will intersect. This intersection point
will be a point in common to
1
p ,
2
p and
3
p .
Case 2: l lies on
3
p , then any point on l
will be a point in common to
1
p ,
2
p and
3
p .
Case 3: l is parallel to
3
p and l does not
intersect
3
p , then since
1
p and
2
p only
intersect in l ,
1
p ,
2
p and
3
p have no point
in common.
l is parallel to
3
p if and only if:
1 5
1 0
1 17

´ } ´ }
] ]
=
] ]
] ]
' ¹ ' ¹
5 17 0  + + =
22  = ÷
l is parallel to
3
p if and only if the
direction vector of l is perpendicular to the
normal to
3
p , i.e. their dot product is 0.
When l is parallel to
3
p ,
l intersects
3
p if and only if:
1 5
1
0 17
 
÷ ´ } ´ }
] ]
÷ =
] ]
] ]
' ¹ ' ¹
5   = ÷ ÷
When l is parallel to
3
p , either every point
on l or no point on l lies on
3
p . Hence, it
is enough to check if a point on l lies on
3
p .
Hence, both 22  = ÷ and 5   = ÷ ÷
must be true.
22  = ÷ and 17  = ÷ 
This corresponds to the situation when l is
parallel to
3
p and l does not intersect
3
p .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
83
(iv) Let  be the required plane.
1 1 2
1 1 0
0 3 3
÷ ÷ ´ } ´ } ´ }
] ] ]
÷ ÷ ÷ =
] ] ]
] ] ]
÷
' ¹ ' ¹ ' ¹
1 2 3
1 0 1
1 3 2
÷ ÷ ´ } ´ } ´ }
] ] ]
× =
] ] ]
] ] ]
÷
' ¹ ' ¹ ' ¹
To get the scalar product equation of the
plane  in the form p = r n ,
1. Find the normal n by taking the cross
product of 2 vectors which lie on the
plane.
Since l lies on  , the direction vector
of l also lies on  . Use a point which
lies on l and the given point . ) 1, 1, 3 ÷
to find another vector which lies on  .
4
3 1 3
: 1 1 1 2
2 3 2
p
÷ ÷ ´ } ´ } ´ }
] ] ]
= ÷ =
] ] ]
] ] ]
' ¹ ' ¹ ' ¹
r
2. Once n is known, find the constant p
using
p = = r n a n
where a is the position vector of any
point on  .
3
1 2
2
x
y
z
÷ ´ } ´ }
] ]
=
] ]
] ]
' ¹ ' ¹
Cartesian equation of  :
3 2 2 x y z ÷ + + = 
Expand the dot product to get the Cartesian
equation of  , as required.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
84
Complex Numbers
A complex number in Cartesian form:
z a bi = +
consists of a real part a and an imaginary
part bi , where , a be¯ and 1 i = ÷ .
Any complex number z a bi = + can be
represented geometrically as a point on the
complex plane.
On an Argand diagram, the horizontal axis is
the real axis, and the vertical axis is the
imaginary axis.
Let z a bi = + , , a be¯ be a complex number in Cartesian form.
On an Argand diagram, let Z and
*
Z be the points representing z and
*
z respectively.
- Real part,   Re z and imaginary part,   Im z
o . ) Re z a =
o . ) Im z b = (not bi )
o . ) Re z is the x -coordinate of Z
o . ) Im z is the y -coordinate of Z
- Modulus, z
o
2 2
z a b = +
o Distance of Z from the origin (length of OZ )
- Argument,   arg z
o Angle that OZ

makes with the positive
real axis, in an anticlockwise sense
o . ) arg z   ÷ < <
o . ) arg 0 is undefined
o Calculate the basic angle,
1
tan
b
A
a
÷
=
o Draw the Argand diagram: . ) arg z
depends on both A and the quadrant
which the point Z is in.
0
Re
Im
  , Z a b
a
b
z
. ) arg z
0
Re
Im
A
A A
A
. ) arg z A = . ) arg z A  = ÷
. ) arg z A = ÷
. ) arg z A  = ÷ +
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
85
- Conjugate,
*
z a bi  
o
. ) . )
*
Re Re z z =
o
. ) . )
*
Im Im z z =
o
*
z z =
o
. ) . )
*
arg arg z z = ÷
o
2
*
zz z =
o . )
*
2Re z z z + =
o . )
*
2 Im z z i z ÷ =
o The conjugate distributes over addition, subtraction, multiplication, and division:
 . )
*
* *
zw z w = ,
*
*
*
z z
w w
´ }
=
]
' ¹
, . )
*
* *
z w z w + = + , . )
*
* *
z w z w ÷ = ÷
o To simplify a fraction with a complex number in its denominator, multiply both
numerator and denominator by the complex conjugate of the denominator.
o Z is
*
Z reflected about the x -axis.
Forms of Complex Numbers
- Cartesian form
o z a bi  
o Useful for addition and subtraction of complex numbers
- Polar form(also Trigonometric form or Modulus-Argument form)
o   cos sin z r i    
o Convert from polar to Cartesian or exponential forms, which are more useful.
- Exponential form
o
i
z re


o Useful for multiplication, division and exponentiation of complex numbers
- If . ) cos sin
i
z re r i

  = = + , i.e. z is in Polar or Exponential form,
o z r =
o . ) arg z  =
o . )
*
cos sin
i
z re r i

 
÷
= = ÷
- Conversion between forms
o Cartesian to Polar or Exponential: r z = , . ) arg z  =
o Polar or Exponential to Cartesian: cos a r  = , sin b r  =
o Between Polar and Exponential: Use the same r and 
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
86
Operations on Complex Numbers
Let
1
1 1 1 1
i
z a bi re

= + = and
2
2 2 2 2
i
z a b i r e

= + = .
On an Argand diagram, let
1
Z and
2
Z be the points representing
1
z and
2
z respectively.
- Addition & Subtraction
o . ) . )
1 2 1 2 1 2
z z a a b b i + = + + +
o . ) . )
1 2 1 2 1 2
z z a a b b i ÷ = ÷ + ÷
o Complex numbers can be represented as vectors on an Argand diagram. Hence,
addition and subtraction of complex numbers can be represented as addition and
subtraction of vectors.
o
1 2
OZ OZ OA + =

, the position vector of the point representing
1 2
z z + .
o
1 2
OZ OZ OB ÷ =

, the position vector of the point representing
1 2
z z ÷ .
- Multiplication & Division
o
1 2 1 2 1 2
z z z z r r = =
o
1
1 1
2 2 2
z
z r
z z r
= =
o . ) . ) . )
1 2 1 2 1 2
arg arg arg z z z z   = + = +
o . ) . )
1
1 2 1 2
2
arg arg arg
z
z z
z
 
´ }
= ÷ = ÷
]
' ¹
o Starting from
1
OZ

,
 An anticlockwise pivot by
2
 radians about the origin and scaling by a
factor of
2
r gives the position vector of the point representing
1 2
z z .
 A clockwise pivot by
2
 radians about the origin and scaling by a factor of
2
r gives the position vector of the point representing
1
2
z
z
.
 An anticlockwise pivot by  radians about the origin gives the position
vector of the point representing
1
i
z e

. In particular, when
2

 = , the
position vector of the point representing
1
iz is given.
1
Z
0
2
Z
A
B
Im
Re
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
87
Complex Roots of Polynomial Equations
Non-real complex roots of a polynomial equation with real coefficients occur in conjugate
pairs. If  is a root of a polynomial equation with real coefficients, then
*
 is also a root of
the equation.
Example: Since
3 2
5 17 13 0 z z z + + + = is a polynomial equation with real
coefficients and 2 3 z i  = =÷ + is a root,
*
2 3 z i  = = ÷ ÷ is also a root. Equivalently,
since 2 3 z i ÷ ÷ is a factor of the polynomial, 2 3 z i ÷ + is also a factor.
De Moivre’s Theorem
If . ) cos sin
i
z re r i

  = = + ,
. )
. ) . ) cos sin
i n n n n
z r e r n i n

  = = + ]
]
Loci of Complex Numbers
A locus (plural loci) is a set of points which fulfill a given condition.
On an Argand diagram, let Z , A , B be the points representing z , a , b respectively.
Z is a variable point which can move, while A and B are fixed points.
- z a ÷ is the distance between Z and A .
- . ) arg z a ÷ is the angle made by the vector AZ

with the positive real axis.
z a r  
What it means:
The distance between Z and A is a constant, r
Locus of Z :
A circle with centre A and radius r
z a r  
The distance between Z and A is less than r .
Locus of Z :
The area inside the circle with centre A and radius r .
z a r  
The distance between Z and A is more than r .
Locus of Z :
The area outside the circle with centre A and radius r .
r
z a r ÷ =
Im
Re 0
A
z a r ÷ <
r
Im
Re 0
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
88
z a z b   
What it means:
Z is an equal distance from each of A and B
Locus of Z :
The perpendicular bisector of
the line joining A and B
z a z b   
Z is closer to A than it is to B .
z a z b   
Z is closer to B than it is to A .
Locus of Z in each case:
z a z b ÷ = ÷ separates the plane into 2 areas:
- z a z b ÷ < ÷ is the area containing A .
- z a z b ÷ > ÷ is the area containing B .
  arg z a   
What it means:
The angle made by AZ

with the positive real axis is a
constant, 
Locus of Z :
The half-line starting at (but not including) A and
making an angle  with the positive real axis
 
1 2
arg z a     
The angle made by AZ

with the positive real axis is
between
1
 and
2
 .
Locus of Z :
The area bounded from the locus . )
1
arg z a  ÷ = ,
anticlockwise to the locus . )
2
arg z a  ÷ = .
z a z b ÷ = ÷
Im
Re 0
A
B
Im
Re 0
. ) arg z a  ÷ =

A
1

2

. )
1 2
arg z a   < ÷ <
A
Im
Re 0
A
B
z a z b ÷ < ÷ Im
Re 0
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
89
FAQ
Q:
Find the roots of the equation  
n
z w    ,
where z and w are complex numbers,  is a constant, and n is a positive integer.
A:
1. Convert w to exponential form. Say
i
w re

= .
. )
n
i
z re

 ÷ =
2. Introduce 2k to  .
. )
. ) 2
n
k i
z re
 

+
÷ = , where k e¯ .
3. Raise both sides to the power of
1
n
.
. )
1 1
2k i
n n
z r e
 

+
]
÷ =
]
4. By De Moivre’s Theorem,
2
1
k
i
n
n
z r e
 

+ ´ }
]
' ¹
÷ =
5. Solve for z .
2
1
k
i
n
n
z r e
 

+ ´ }
]
' ¹
= +
6. Write the possible values of k . In general, 0,1, 2, , 1 k n = ÷ . .
However, if the roots must be given in the form
i
re

with    ÷ < < , substitute
values of k to check if  is in the required range. There will be n consecutive
integer values of k .
Reject those values of k which make any expression undefined (denominator 0 ).
Q: Solve for z and w in 2 simultaneous linear equations.
Example [PJC07/Prelim/I/1]
Solve the simultaneous equations
1 iw z i + = ÷ ÷ ----- (A)
. )
40
2 1
6 2
z i w
i
÷ + =
÷
----- (B)
giving each answer in the form a bi + , where a and b are real.
A: Method 1
1. Multiply (A) by 2 such that the coefficient of z in (A) and (B) are both 2.
2. Use (A) ÷ (B) to eliminate the 2z term from both equations.
3. Solve for w, then solve for z .
Method 2
1. Using (A), express z in terms of w: 1 z i iw = ÷ ÷ ÷ .
2. Substitute z in (B) with the expression in terms of w.
3. Solve for w, then solve for z .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
90
Q: Find the complex (real and non-real) roots of a polynomial equation.
A:
- If the polynomial equation in z has degree n (the highest power of z is n ), and
the coefficient of
n
z is k , the equation can be expressed as a product of k and n
linear factors of the form z  ÷ .
- If  is a root of the equation, z   is a factor of the polynomial.
1. If the polynomial equation has real coefficients,
Given that  is a non-real complex root of the equation,
*
 is also a root.
2. Using any given roots or factors, long division or comparing coefficients, express
the polynomial as a product of its factors.
3. Use the general quadratic formula to factorize any quadratic factors.
4. When the equation is in this form with n linear factors of the form z  ÷ :
. ) . ) . )
1 2
0
n
k z z z    ÷ ÷ ÷ = .
then
1 2
, , ,
n
   . are the n (possibly repeated) roots of the equation.
Q: Sketch loci on an Argand diagram.
A: - Draw all circles, lengths and angles accurately and to scale using a compass, ruler
and protractor. Ensure the real and imaginary axes are to the same scale.
- For loci of the form . ) arg z a  ÷ = , draw a small circle at the point representing a
to show that it is excluded, since . ) . ) arg arg 0 a a ÷ = is undefined.
- For strict inequalities, draw the loci of the equality cases as dashed to show that
they are excluded, e.g. for z a r ÷ < , draw the locus z a r ÷ = as a dashed circle.
- Label all loci with their equations, and label required regions/sets of points.
Q:
Find the minimum/maximum value of z a  ,   arg z a  or z w  ,
OR Find the points of intersection of the loci of points representing z and w,
where z and w are variable while a is fixed.
A: - Interpret what expressions mean geometrically on the Argand diagram, rather than
interpret them algebraically. e.g. minimum value of z w ÷ is the minimum distance
between any point on the locus of points representing z and any point on the locus
of points representing w.
- As a last resort for finding points of intersection, find the Cartesian equations of the
lines or circles and solve the equations simultaneously.
- Find lengths using the formula for the distance between 2 points.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
91
- Find right-angled triangles to use trigonometric ratios.
- If the locus of z is a circle, z b r ÷ = and the fixed complex number a is
represented by the point A outside the circle,
o Use tangents to the circle to find the min./max. values of   arg z a  .
o Draw a line from point A through the centre of the circle. The line and
circle will have 2 points of intersection. Use the distance from A to these 2
points to find the min./max. values of z a  .
Example 1 [SRJC08/Prelim/II/3]
(i) Show that
1
1
i
i
e
e


+
÷
can be reduced to cot
2
i
 ´ }
]
' ¹
.
(ii) Find all the roots of the equation . )
8
1
1
2
z i = ÷ , giving each root in the form
i
re

where 0 r > and    ÷ < < .
(iii) Hence deduce all the roots of . )
8
1 1
1
1 2
w
i
w
÷ ´ }
= ÷
]
+
' ¹
.
Solution Comments
(i)
2 2
2 2
1
1
i i
i
i
i i
e e e
e
e e
 

  
´ } ´ }
÷
] ]
' ¹ ' ¹
´ } ´ }
÷
] ]
' ¹ ' ¹
+ +
=
÷
÷
2
2
2Re
2 Im
i
i
e
i e


´ }
÷
]
' ¹
´ }
÷
]
' ¹
]
]
]
]
=
]
]
]
]
2cos
2
2 sin
2
i


´ }
÷
]
' ¹
=
´ }
÷
]
' ¹
“Factoring out
2

” is a common theme in
Complex Numbers proving questions.
Here, it takes the form of factoring out
2
i
e
 ´ }
]
' ¹
from both the numerator and denominator.
The aim is to create expressions of the form
*
z z + or
*
z z ÷ , which can be written as . ) Re z
or . ) 2 Im i z respectively.
cos
2
sin
2
i


´ }
]
' ¹
=
´ }
÷
]
' ¹
. ) cos cos   ÷ =
. ) sin sin   ÷ = ÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
92
1
cot
2 i
 ´ } ´ }
= ÷
] ]
' ¹ ' ¹
cot
2
i
 ´ }
=
]
' ¹
2
1
1
i i
i
i i
÷ = ÷ = ÷ =
÷
(ii)
. )
8 4
1
1
2
i
z i e
 ´ }
÷
]
' ¹
= ÷ =
8 1
4
k
i
e

÷ ´ }
]
' ¹
=
1. Convert . )
1
1
2
i ÷ to exponential form.
2
4
k i
e


´ }
÷
]
' ¹
= , where k e .
2. Introduce 2k to
4

÷ .
By De Moivre’s Theorem,
8 1
32
k
i
z e

÷ ´ }
]
' ¹
= ,
3. Raise both sides to the power of
1
8
, and
4. Use De Moivre’s Theorem.
where 3, 2, 1, 0,1, 2,3, 4 k = ÷ ÷ ÷  5. Write out the possible values of k . Since
the question requires    ÷ < < , there will
be exactly 8 consecutive integer values of k
which will give  in the required range.
Substitute values of k to check which
values are valid.
(iii)
. )
8
1 1
1
1 2
w
i
w
÷ ´ }
= ÷
]
+
' ¹
Let
8 1
32
k
 
÷ ´ }
=
]
' ¹
.
1
1
i
w
z e
w

÷
= =
+
1 zw z w + = ÷
. ) 1 1 z w z + = ÷
1 1
1 1
i
i
z e
w
z e


+ +
= =
÷ ÷
Since
1
1
w
w
÷
+
corresponds with z in part (ii),
equate the two. To deduce the roots of the
equation in w, express w in terms of z .
8 1
cot cot
2 64
k
w i i


] ÷ ´ } ´ }
= =
] ] ]
' ¹ ' ¹
]
,
where 3, 2, 1, 0,1, 2,3, 4 k = ÷ ÷ ÷ 
k takes the same values as in part (ii), since the
values of k are for the roots of the equation in
z from part (ii).
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
93
Example 2 [AJC08/Prelim/I/6]
If z i = is a root of the equation . ) . )
3 2
1 3 2 3 2 0 z i z i z + ÷ ÷ + ÷ = , determine the other roots.
Hence find the roots of the equation . ) . )
3 2
1 3 3 2 2 0 w i w i w + + + ÷ ÷ = .
Solution Comments
Let . ) . )
3 2
1 3 2 3 2 z i z i z + ÷ ÷ + ÷
. ) . )
2
z i z az b = ÷ + + ,
for some complex numbers a and b .
. ) . )
3 2
1 3 2 3 2 z i z i z + ÷ ÷ + ÷
3 2 2
z az bz iz aiz bi = + + ÷ ÷ ÷
. ) . )
3 2
z a i z b ai z bi = + ÷ + ÷ ÷
Comparing coefficients,
1 3 1 2 i a i a i ÷ = ÷ ¬ = ÷
2
2 2 bi b i
i
÷ = ÷ ¬ = = ÷
. )
2 2
1 2 2 z az b z i z i + + = + ÷ ÷
Let . ) . ) . )
3 2
1 3 2 3 2 z i z P z i z = + ÷ ÷ + ÷ .
Not all of the coefficients of . ) P z are real.
Hence, though i is a root, . )
*
i i = ÷ may not be
a root.
After . ) P z (degree 3) is divided by z i ÷
(degree 2), it should leave a quadratic factor
(degree 2). Expand
. ) . )
2
z i z az b ÷ + + and
compare the coefficients.
Do NOT write . ) P z in terms of linear factors
first: . ) . ) . ) z i z a z b ÷ ÷ ÷ , since it will be
difficult to compare coefficients.
Alternative method
Use long division to find the remainder when
. ) P z is divided by z i ÷ .
When . )
2
1 2 2 0 z i z i + ÷ ÷ = ,
. ) . )
2
(1 2 ) 1 2 4 1 ( 2 )
2(1)
2 1 3 4
2
2 or 1
i i i
z
i i
i
÷ ÷ ± ÷ ÷ ÷
=
÷ ± ÷ +
=
= ÷
Hence, the other roots are 2i and 1 ÷ . 
Use the general quadratic formula to factorize
the quadratic factor.
Use the GC to perform calculations with
complex numbers, including square roots, to
avoid careless mistakes.
Taking conjugates on both sides,
. ) . )
*
3 2 *
1 3 2 3 2 0 z i z i z ] + ÷ ÷ + ÷ =
]
Let . ) . ) . )
3 2
1 3 3 2 2 w i w w Q w i + + + ÷ ÷ = .
Observe that the coefficients of w in . ) Q w are
the conjugates of the coefficients of z in . ) P z .
Hence, convert the coefficients of z in . ) P z to
their conjugates.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
94
. ) . ) . ) . )
* *
* *
3 2 *
1 3 2 3 2 0 z i z i z + ÷ ÷ + ÷ =
. ) . ) . ) . )
3 2
* * *
1 3 3 2 2 0 z i z i z + + + ÷ ÷ =
Hence, the roots of
. ) . )
3 2
1 3 3 2 2 0 w i w i w + + + ÷ ÷ =
are the conjugates of the roots of
. ) . )
3 2
1 3 2 3 2 0, z i z i z + ÷ ÷ + ÷ =
i.e. i , 2i ÷ , and 1 ÷ . 
Simplify the LHS using . )
*
* *
z w z w ± = ± , and
. )
*
* *
zw z w = , which extends to:
. ) . ) . )
* 3
*
3 * * * *
z z z z z z z z = = =
Since the equation in
*
z is now similar to the
equation in w, each root of . ) Q w corresponds
to the conjugate of a root of . ) P z .
Example 3 [AJC07/Prelim/II/3]
In an Argand diagram, the point A represents the fixed complex number a , where
. ) 0 arg
2
a

< < . The complex numbers z and w are such that
2 z ia a ÷ = and w w ia = +
Sketch, in a single diagram, the loci of the points representing z and w.
Find:
(i) the minimum value of z w ÷ in terms of a ,
(ii) the range of values of
1
arg
z
´ }
]
' ¹
in terms of . ) arg a .
Solution Comments
Let B be the point representing 2ia .
Let C be the point representing –ia .
If . ) 0 arg
2
a

< < , A is in the first
quadrant of the complex plane. Plot any
such point A .
The locus of points representing z is a
circle centred at B with radius a .
The locus of points representing w is the
perpendicular bisector of the line
segment joining the origin and C.
w w ia = +
Im
Re 0
A
2 z ia a ÷ =
a
B
C
D
a
a
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
95
From OA

, an anticlockwise pivot about
the origin by
2

radians and a stretching
by factor 2 gives OB

.
From OA

, a clockwise pivot about the
origin by
2

radians gives OC

.
(i)
Minimum value of z w ÷
1
2
OD OC = +
1
2
a a = +
3
2
a = 
Geometrically, the minimum value of
z w ÷ is the minimum distance between
a point on the locus of points
representing Z and a point on the locus
of points representing W .
Since the locus of points representing w
is a bisector of OC , it divides OC into
2 segments of equal length,
1
2
a .
(ii)
. ) . ) . )
1
arg arg 1 arg arg z z
z
´ }
= ÷ = ÷
]
' ¹
Express the condition in terms of a
known variable, . ) arg z .
Do not redraw the diagram, a relevant
part of it is copied here for easier
reference.
Let M and N be the points at which the
tangent to the locus of z passes through the
origin, as shown.
Let M and N represent the complex
numbers m and n respectively.
Then . ) . ) arg( ) arg arg m z n < < .
Using tangents to the circle is a
common technique to find the min./max.
value of . ) arg z a ÷ when the locus of
points representing z is a circle.
If Z is the point representing z , . ) arg z
is the angle made by the vector OZ

with the positive real axis.
0
A
2 z ia a ÷ =
B
D
N
M
a
a
a
a
Im
Re
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
96
1
sin
2 6
a
BOM BON
a

÷
´ }
. = . = =
]
]
' ¹
. ) . )
. )
. )
arg arg
2
arg
2 6
arg
3
m a BOM
a
a

 

= + ÷.
= + ÷
= +
. ) . )
. )
. )
arg arg
2
arg
2 6
2
arg
3
n a BON
a
a

 

= + +.
= + +
= +
Use . ) arg a since the question requires
the final answer to be in terms of
. ) arg a .
The line OB passes through the centre
of the circle. Hence, use the angle OB
makes with the positive real axis as a
reference, and add/subtract from it.
OA

makes an angle of . ) arg a with the
positive real axis, and
2
BOA

. = .
Hence, OB

makes an angle of
. ) arg
2
a

+ with the positive real axis.
. ) . ) . )
2
arg arg arg
3 3
a z a
 
+ < < +
Convert the range of values of . ) arg z
to the range of values of
1
arg
z
´ }
]
' ¹
.
. ) . ) . )
2
arg arg arg
3 3
a z a
 
÷ ÷ > ÷ > ÷ ÷
. ) . )
2 1
arg arg arg
3 3
a a
z
  ´ }
÷ ÷ < < ÷ ÷
]
' ¹

Multiply each term in the inequality by
1 ÷ , and flip the inequality signs.
Example 4 [NJC08/Prelim/I/6]
(i) The set of points P in an Argand diagram represents the complex number z that
satisfies 2 2 z i ÷ ÷ = . Sketch the locus of P .
(ii) The set of points Q represents another complex number w given by
. ) arg 2 3 w i  ÷ + = where    ÷ < < . Give a geometrical description of the locus of
Q .
(a) Find the range of values of  such that the locus of Q meets the locus of P
more than once.
(b) In the case where
1
3
tan
4

÷
´ }
=
]
' ¹
, find the least value of z w ÷ in exact form.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
97
Solution Comments
(i)
. ) 2 2 z i ÷ + =
The locus of P is a circle centred at . ) 2,1
with radius 2 .
Show on the diagram that the circle is
tangent to the imaginary axis. Examiners
will look out for the relative position of
lines and circles to the axes.
(ii) The locus of Q is a half-line starting at
(but not including) . ) 2, 3 ÷ and making
an angle  with the positive real axis.
. ) arg 2 3 w i  ] ÷ ÷ =
]
Remember to write “(but not including)”.
(ii)(a) Using tangents to the circle is the same
technique as in Example 3, part (ii).
1
1
2
sin
2 4 2 6 3
   

÷
´ }
= ÷ = ÷ =
]
' ¹
1
2
2 2
sin
2 4 2 6 3
   

÷
´ }
= + = + =
]
' ¹
1 2
   < <
2
3 3
 
 < < 
The line passing through the centre of the
circle makes an angle of
2

with the
positive real axis. Hence, use this angle as
a reference, and add/subtract from it.
2
2 2 z i ÷ ÷ =
0
(2,1 )
Im
Re
2 2 z i ÷ ÷ = Im
2
0
(2,1 )
(2, 3) ÷
2
4
1

2

Re
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
98
(ii)(b) The locus of Q (which represents w)
should have gradient
1
3 3
tan tan
4 4
÷
] ´ }
=
] ]
' ¹ ]
Geometrically, the least value of
z w ÷ is the minimum distance
between a point on the locus of P
and a point on the locus of Q .
From the diagram, this minimum
distance is DC , where DCB . and
ACB . are right angles.
4 AB =
1
3
tan
2 4
ABC

÷
´ }
. = ÷
]
' ¹
sin
AC
ABC
AB
. =
1
sin
3
4sin tan
2 4
AC AB ABC

÷
= .
] ´ }
= ÷
] ]
' ¹ ]
The only useful right-angled triangle
is ABC . Since 2 DC AC = ÷ , aim to
find AC first.
Knowing AB and ABC . , AC can
be found using trigonometric ratios
on the right-angled triangle ABC .
4 16
4
5 5
AC
´ }
= =
]
' ¹
Least value of
2
6
5
z w DC
AC
÷ =
= ÷
= 
Let
1
3
tan
4

÷
´ }
=
]
' ¹
.
To calculate sin
2


´ }
÷
]
' ¹
in exact
form, draw a right-angled triangle
involving  . Since
3
tan
4
 = , the
side opposite  and the side adjacent
to  have lengths in the ratio 3: 4.
2 2 z i ÷ ÷ = Im
0
(2,1 ) A
(2, 3) B ÷
2
4
1
3
tan
4
÷
´ }
]
' ¹
Re
6
3
4
. ) arg 2 3 w i  ÷ + =
C
D
4
3
5

2

 ÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
99
Permutations & Combinations
Basic Counting Principles
Addition Principle: If there are a ways to do something and b ways to do another thing,
and the two things cannot be done simultaneously, then there are a b + total ways to do
exactly one of the things.
Multiplication Principle: If there are a ways to do something and b ways to do another
thing, then there are a b × ways to do both things in succession.
Choosing & Permuting
Is the order of objects important?
- Look for keywords in the question, e.g.
o Important: arrange, order, in a row, around a table
o Not Important: choose, select, form groups/committees/teams
Can each object be repeated any number of times?
Are all the objects distinct?
No. of ways to choose r
objects from n distinct
objects:
. )
!
! !
n
r
n
C
r n r
=
÷
e.g. Choose 2 monitors
from a class of 24
students: 24 n = , 2 r =
No. of ways
24
2
C =
No. of arrangements of r objects (possibly
repeated) from n distinct objects:
r
n
e.g. Form a string of 4 letters using the English
alphabet: 26 n = , 4 r = , No. of ways
4
26 =
No. of arrangements of r objects (all
distinct) from n distinct objects:
. )
!
!
n
r
n
P
n r
=
÷
e.g. Choose 1 monitor, 1 treasurer, 1
secretary (all distinct) from a class of
24 students: 24, n = 3 r = ,
No. of ways
24
3
P =
No. of arrangements of n objects, where:
1
n are of type 1,
2
n are of type 2, …
k
n are of type k :
1 2
!
! ! !
k
n
n n n .
e.g. Arrange 2 red balls, 3 green balls, 5
blue balls in a row. Since 2 3 5 10 + + = ,
No. of ways
. ) . )
10!
2! 3! (5!)
=
Yes No
Yes
No
Yes No
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
100
Useful Techniques
- Circular permutations: No. of permutations of distinct objects in a circle = ( − 1)! ,
assuming the positions around the circle are identical.
- Grouping adjacent objects: If some objects must be adjacent, group them and treat them
as a single object. When arranging all the objects/groups, remember to also arrange the
grouped objects within each group.
- Slots: e.g. If 3 women and 5 men are to be seated in a row such that no 2 women are
adjacent, arrange the 5 men first. There are 6 ‘slots’ (denoted | ) between and beside the 5
men (denoted M):
| M | M | M | M | M |
Then, arrange the 3 women in 3 distinct ‘slots’. Hence, the number of ways is
6
3
5! P × .
- Complement method: Use it when direct computation is tedious, especially if key
phrases like ‘at least one’ appear in the question.
If A is a condition, the number of objects with A equals the total number of objects,
minus the number of objects without A :
. ) . ) total ( ') n A n n A = ÷
- Case-by-case method: Use it if all else fails. Check that each possibility is counted in
one and exactly one case, such that no possibilities are double-counted or missed out.
Example 1 [RJC08/Tutorial 12/Section D/11]
Find the number of permutations of all eight letters of the word BELIEVED in which
(i) there is no restriction;
(ii) the three letter ‘E’s are next to each other;
(iii) the three letter ‘E’s are always separated.
Solution Comments
(i) 8!
6
!
720
3
= 
The letters in BELIEVED are 3 ‘E’s and 1 each of ‘B’, ‘L’,
‘I’, ‘V’, ‘D’. Since only ‘E’ is repeated, use
1 2
!
! ! !
k
n
n n n .
with 8 n = and
1
3 n = (
2 8
, , 1 n n . = ).
(ii) 6! 720 =  Group the 3 ‘E’s as 1 object. Arrange the 6 resulting objects.
(iii)
6
3
5! 0 24 0 C × = 
Arrange B, L, I, V, D first, for 5! permutations. There are 6
‘slots’ between and beside these 5 letters: | B | L | I | V | D |.
Place the 3 ‘E’s in 3 of these 6 slots. Since the order of the
‘E’s does not matter, use
6
3
C instead of
6
3
P .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
101
Example 2 [RJC08/Tutorial 12/Section D/12]
3 boys, 2 girls, and a dog sit at a round table. How many ways can they be arranged if
(i) there is no restriction?
(ii) the dog is to be seated between the 2 girls?
(iii) the dog is to be seated between any 2 boys?
Solution Comments
(i)
. ) 6 1 ! 120 ÷ =  Use . ) 1 ! n ÷ directly with 6 n = .
(ii) (4 1)! 2 12 ÷ × =  Group the 2 girls and the dog as 1 object. Arrange the 4
resulting objects, then multiply by 2 as the objects within the
group can be arranged in 2 ways.
Denoting the girls by
1
G and
2
G and the dog by D, the group
can be
1 2
G DG or
2 1
G DG in clockwise order.
(iii)
. ) 4 1 ! 3! 36 ÷ × = 
Group some 2 boys and the dog as 1 object. Arrange the 4
resulting objects to get . ) 4 1 ! ÷ . Since the 3 boys are distinct,
multiply by 3! to arrange them.
Example 3 [HCJC01/I/7]
Thirteen cards each bear a single letter from the first thirteen letters A, B, …, L, M of the
alphabet. Seven cards are selected at random from the thirteen cards, and the order of
selection is not relevant. How many selections can be made if the seven cards
(i) contain at least two but not more than four of the letters A, B, C, D and E?
(ii) contain four but not five consecutive letters, e.g. A, B, C, D, G, H, I?
Solution Comments
(i)
5 8 5 8 5 8
2 5 3 4 4 3
C C C C C C × + × + ×
1580 = 
Use a case-by-case method.
The 7 cards chosen contain:
Case (1): exactly 2 of the letters A, B, C, D, E.
Case (2): exactly 3 of those letters.
Case (3): exactly 4 of those letters.
e.g. for Case (1) when there are 2 of those 5 letters,
choose those 2 first to get
5
2
C . We are left with 5
letters to choose and 8 letters to choose from, to get
8
5
C .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
102
(ii)
8 7
3 3
2 8 392 C C × + × = 
Since there are only 7 cards, they can contain at
most 1 string of 4 consecutive letters. Use a case-by-
case method.
The 7 cards chosen have no 5 consecutive letters,
and:
Case (1):
Contain A, B, C, D or J, K, L, M (2 sub-cases)
Case (2):
Contain any other 4 consecutive letters (8 sub-cases)
For Case (1), after the 4 consecutive letters are
fixed, 1 other letter cannot be chosen, since it would
form 5 consecutive letters, i.e.:
If A, B, C, D are fixed, E cannot be chosen.
If J, K, L, M are fixed, I cannot be chosen.
We are left with 3 letters to choose and 8 letters to
choose from, to get
8
3
C for each sub-case.
Case (2) is similar, except after the 4 consecutive
letters are fixed, 2 (not 1) other letters cannot be
chosen, e.g.:
If B, C, D, E are fixed, A and F cannot be chosen.
Hence, we get
7
3
C for each sub-case.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
103
Probability
Definitions
Let A and B be random events.
. ) P A B · is the probability that both A and B occur.
. ) P A B is the probability that only A occurs, only B occurs, or both A and B occur.
- Conditional probability: . ) | P A B is the conditional probability of A given B .
Equivalently, . ) | P A B is the probability of A occurring, given that B has occurred.
o . )
. )
. )
|
P A B
P A B
P B
·
= , where . ) 0 P B = .
- Mutually exclusive: A and B are mutually exclusive if A and B cannot both occur.
o . ) 0 P A B · =
o . ) . ) . ) P A B P A P B = +
- Independent: A and B are independent if the occurrence or non-occurrence of one
event has no influence on the occurrence or non-occurrence of the other.
o . ) . ) | P A B P A =
o . ) . ) | P B A P B =
o . ) . ) . ) P A B P A P B · = ×
Useful Techniques
- Use set diagrams to illustrate the problem.
- Complement method: Use it when direct computation is tedious, especially if key
phrases like ‘at least one’ appear in the question.
. ) . ) 1 ' P A P A = ÷
- Case-by-case method: Use it if all else fails. Check that each possibility is counted in
one and exactly one case, such that no possibilities are double-counted or missed out.
- For any 2 random events A and B ,
. ) . ) . ) . ) P A B P A P B P A B = + ÷ ·
In particular, if A and B are mutually exclusive, then since . ) 0 P A B · = ,
. ) . ) . ) P A B P A P B = +
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
104
- To find the probability of A occurring given that B has occurred,
o Calculate . )
. )
. )
|
P A B
P A B
P B
·
= , or
o Consider the reduced sample space given that B has occurred, then calculate
. )
. )
. )
|
n A B
P A B
n B
·
= , where . ) n X is the number of outcomes for X to occur.
- Use concepts from Permutations & Combinations (P&C) to count cases.
- Use concepts from other topics for alternative approaches to questions.
o P&C, when objects are chosen without replacement.
o The Binomial distribution, when objects are chosen with replacement.
Probability Tree (Tree Diagram)
Use a probability tree when the fixed probabilities or conditional probabilities of 2 or 3
events are given.
Example: . ) P A  = , . )
1
| P B A  = , and . )
2
| ' P B A  = .
Probability tree:
-  is the probability of A , while
1
 is the conditional probability of B given A . In
general, the probabilities on the first level of the diagram are for only 1 event, while those
on the second level are conditional probabilities.
- To find . ) P A B · , . ) ' P A B · , . ) ' P A B · , . ) ' ' P A B · , multiply the 2 relevant
probabilities.
e.g. . ) . ) . ) . )
1
' ' | ' 1 P A B P A P B A   · = × = ÷
- To find . ) P B or . ) ' P B , add the probabilities of the branches which lead to B or ' B .
e.g. . ) . ) . ) . ) . ) . )
1 2
| ' | ' 1 P B P A P B A P A P B A    = × + × = + ÷

B
1
1  ÷
1

B
2

2
1  ÷
' B
' B
A
' A
1  ÷
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
105
Example 1 [D90/II/6]
A bag contains 4 red counters and 6 green counters. 4 counters are drawn at random from the
bag, without replacement. Calculate the probability that:
(i) all the counters drawn are green,
(ii) at least 1 counter of each colour is drawn,
(iii) at least 2 green counters are drawn,
(iv) at least 2 green counters are drawn, given that at least 1 counter of each colour is
drawn.
State with a reason whether or not the events ‘at least 2 green counters are drawn’ and ‘at
least 1 counter of each colour is drawn’ are independent.
Probability approach P&C approach
(i) 6 5 4 3 1
10 9 8 7 14
× × × = 
6
4
10
4
14
1 C
C
= 
Let G be the number of green counters out
of the 4 counters drawn. Then, this is
( 4) P G = .
There is no need to multiply by anything
else to account for order, since all 4 counters
are the same type (green).
Assume the counters are distinct, so that
each of the
10
4
C combinations of 4
counters have an equal probability of
being chosen. Out of these,
6
4
C
combinations are all green.
(ii)
1 4 3 2 1 97
1
14 10 9 8 7 105
´ }
÷ + × × × =
]
' ¹

4
4
10
4
1 97
1
14 1 5 0
C
C
´ }
÷ + =
]
' ¹

Using the complement method:
. ) 1, 2, 3 P G =
. ) 1 0 ( 4) P G P G = ÷ = ÷ =
. ) 1 0 (i) P G = ÷ = ÷
Similar to the Probability approach.
(iii) 1 97 6 4 3 2 4! 37
14 105 10 9 8 7 3! 42
+ ÷ × × × × = 
6 4
1 3
10
4
1 97 37
14 105 42
C C
C
×
+ ÷ = 
. ) 2, 3, 4 P G =
. ) . ) . ) 4 1, 2, 3 1 P G P G P G = = + = ÷ =
(i) (ii) ( 1) P G = + ÷ =
Multiply by
4!
3!
when finding ( 1) P G = , to
account for order when there are different
types of items (green and red).
Again, assume the counters are distinct.
There are
6
1
C combinations of 1 green
counter, and
4
3
C of 3 red counters,
giving
6 4
1 3
C C × total combinations of 1
green and 3 red.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
106
Solution Comments
(iv)
37 1 97 85
42 14 105 97
´ }
÷ ÷ =
]
' ¹

In probability questions with many parts, look out
for ways to use your answers to earlier parts, even
though the question may not require it explicitly.
We want to find the conditional probability of
2, 3, 4 G = given that 1, 2, 3 G = . Since the
intersection of these 2 events is 2, 3 G = , using the
formula for conditional probability:
. ) 2, 3, 4 ( 4)
( 2, 3) (iii) (i)
( 1, 2, 3) ( 1, 2, 3) (ii)
P G P G
P G
P G P G
= ÷ =
= ÷
= =
= =
Let A be the event that at least
2 green counters are drawn. Let
B be the event that at least 1
counter of each colour is drawn.
Since
. )
85
|
97
P A B =
and
. )
37
42
P A = ,
. ) . ) | P A B P B = , thus A and B
are not independent. 
Since ( ) P B , ( ) P A , and ( | ) P A B are known from
parts (ii), (iii), and (iv) respectively, choose the
condition . ) ( | ) P A P A B = to check if A and B are
independent.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
107
Example 2 [J88/II/6]
4 girls, Amanda, Beryl, Claire, and Dorothy, and 3 boys, Edward, Frank, and George, stand
in a queue in random order. Find the probability that:
(i) the first 2 in the queue are Amanda and Beryl, in that order,
(ii) either Frank is first or Edward is last (or both),
(iii) no 2 girls stand next to each other,
(iv) all 4 girls stand next to each other,
(v) all 4 girls stand next to each other, given that at least 2 girls stand next to each other.
Probability approach P&C approach
(i) 1 1 1
7 6 42
× = 
5! 1
! 7 42
= 
Amanda is in position 1 with probability
1
7
, then once there are 6 positions
remaining, Beryl is in position 2 with
probability
1
6
.
There are 7! permutations of the 7 people,
each with equal probability. 5! of these
permutations have Amanda in position 1
and Beryl in position 2.
(ii) 1 1 1 1 11
7 7 7 6 42
+ ÷ × = 
6! 6! 5! 11
7! 42
+ ÷
= 
Let F be the event that Frank is first,
and E be the event that Edward is last.
Use
. ) . ) . ) ( ) P F E P F P E P F E = + ÷ · .
Similar to the probability approach.
(iii) 4 3 2 1 1
7 6 5 4 35
× × × = 
4! 3! 1
7! 35
×
= 
The 7 people must be in the order
‘gbgbgbg’, where ‘g’ denotes a girl and
‘b’ a boy.
Position the girls first, then the positions
of the boys are fixed and the condition is
satisfied.
Since we are counting permutations with
this approach, we must permute both boys
and girls in the ‘gbgbgbg’ order, then
divide it by the total of 7! permutations.
(iv) 4 3 2 1 4
4
7 6 5 4 35
× × × × = 
4! 3! 4 4
7! 5 3
× ×
= 
This is similar to part (iii), except there
are 4 possible positions of ‘gggg’ within
or beside ‘bbb’.
Similar to the probability approach.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
108
Solution Comments
(v) 4
2
35
1
17
1
35
=
÷

Again, in probability questions with many parts, look out for
ways to use your answers to earlier parts, even though the
question may not require it explicitly.
Let A be the event that all 4 girls stand next to each other. Let B
be the event that at least 2 girls stand next to each other. Then
A B A · = , since B is a subset of A .
'
( ) P B is given by the answer to (iii), since ‘at least 2 girls stand
next to each other’ and ‘no 2 girls stand next to each other’ are
complementary events. Using the formula for conditional
probability:
. )
( ) ( ) ( ) (iv)
|
( ) ( ) 1 ( ') 1 (iii)
P A B P A P A
P A B
P B P B P B
·
= = = =
÷ ÷
Example 3 [N01/II/11a]
Events A and B are such that
. )
1
3
P A = ,
. )
1
|
3
P B A = , and
. )
1
' '
6
P A B · = .
Find
(i) . ) P A B ,
(ii) . ) P B .
Solution Comments
(i)
. ) . )
5
1 ' '
6
P A B P A B · = ÷ = 
On a Venn diagram, A B is the area within one
or both of A and B , and ' ' A B · is the area not
within either A or B .
Hence, . ) . ) 1 ' ' P A B P A B = ÷ · .
(ii)
. ) P B
. ) . ) . ) P A B P A P A B = ÷ + ·
. ) . ) . ) ( | ) P A B P A P A P B A = ÷ +
5 1 1 1
6 3 3 3
´ }´ }
= ÷ +
] ]
' ¹' ¹
11
18
= 
Rewrite the given equations in terms of ( ) P A ,
( ) P B , and ( ) P A B · for easier manipulation.
Use
. ) . ) . ) . ) P A B P A P B P A B = + ÷ ·
then
. )
( )
|
( )
P A B
P B A
P A
·
=
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
109
Distributions
Expectation
Since the expectation is a summation, you can apply summation rules to it.
Let a and b be constants, and X and Y be random variables.
- . ) E a a =
- . ) . ) E aX aE X =
- . ) . ) E aX b aE X b ± = ±
- . ) . ) . ) E aX bY aE X bE Y ± = ±
Variance
The variance is always positive. When combined, 2 variances are always added together.
- . ) Var 0 a = , since a constant has no variation.
- . ) . )
2
Var Var aX a X =
- . ) . )
2
Var Var aX b a X ± =
- If X and Y are independent,
. ) . ) . )
2 2
Var Var Var aX bY a X b Y ± = +
- Standard deviation . ) Var X  = =
- . ) . ) Var 2 4 Var X X = , but . ) . )
1 2
Var 2 Var X X X + = . (Example 2)
Properties of Distributions
- Binomial: . ) ~ , X B n p
o . ) . ) 1
n x
x
n
P X x p p
x
÷ ´ }
= = ÷
]
' ¹
(In MF15) for 0,1, 2,..., x n =
o Mean, . ) E X np =
o Variance, . ) . ) Var 1 X np p = ÷
o Assumptions:
 n independent trials
 Constant probability of success
 Each outcome is either success or failure
 Occur randomly
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
110
- Poisson: . ) ~ Po X 
o . )
!
x
P X x e
x


÷
= = (In MF15) for 0,1, 2,... x =
o Mean, . ) E X  =
o Variance, . ) Var X  =
o Assumptions:
 Occur independently
 Uniform rate of occurrence
 Occur singly
 Occur randomly
- Normal:
. )
2
~ , X N  
o Mean, . ) E X  =
o Variance, . )
2
Var X  =
GC Syntax
If you forget GC syntax, use CtlgHelp. For example, to find the syntax for normalcdf(,
1. [APPS] 3:CtlgHelp [ENTER]
2. [2ND][VARS], then press the down key until normalcdf( is highlighted.
3. Press the [+] key to display the syntax.
If . ) ~ , X B n p ,
- . ) P X a = is given by . ) , , n p a binompdf
- . ) P X a < is given by . ) , , n p a binomcdf
If . ) ~ Po X  ,
- . ) P X a = is given by . ) , a  poissonpdf
- . ) P X a < is given by . ) , a  poissoncdf
If
. )
2
~ , X N   ,
- . ) P X a < is given by . ) 99, , , E a   ÷ normalcdf
- . ) P X a > is given by . ) , 99, , a E   normalcdf
- . ) P a X b < < is given by . ) , , , a b   normalcdf
- . ) . ) P X a P X a < = <
- . ) . ) P X a P X a > = >
- . ) . ) . ) . ) P a X b P a X b P a X b P a X b < < = < < = < < = < <
- For . ) ~ 0,1 Z N , the standard normal random variable,  and  can be omitted in the
GC syntax. The GC will assume that 0  = and 1  = .
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
111
FAQ
Q: Which distribution does a random variable X follow?
A: - If X can have non-integer values,
X is continuous and follows a normal distribution.
- If X can have any integer value from 0 onwards with no upper limit,
X follows a Poisson distribution.
- If X can have any integer value from 0 onwards to an upper limit n ,
X follows a binomial distribution.
Q: “Using a suitable approximation,…”
or “Find the approximate number of…”
or “Estimate…”
A: Show that the original distribution fulfills the conditions for approximation.
Legend: “ ” denotes “approximates to”
If X follows a binomial distribution, i.e. . ) ~ , X B n p , use the value of p to decide
whether to approximate to a Poisson or normal distribution.
- If p is close to 1, approximate X to a Poisson distribution.
- If p is moderate (close to 0.5), approximate X to a normal distribution.
- If p is close to 0, define a new variable Y which has p close to 1, and approximate
Y to a Poisson distribution.
Binomial
. ) ~ , X B n p
Normal
. ) . )
~ , 1 X N np np p ÷
Poisson
. ) ~ Po X 
n is large ( 50 > ),
p is small ( 0.1 < ),
such that 5 np <
Continuity
correction
Continuity
correction
n is large
such that 5 np >
and . ) 1 5 n p ÷ >
Normal
. ) ~ , X N  
10  >
Poisson
. ) ~ Po X np
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
112
Q: Sum random variables.
A: Let X and Y be independent random variables.
- If . )
1
~ Po X  and . )
2
~ Po Y  ,
. )
1 2
~ Po X Y   + +
Subtraction ( X Y ÷ ) is not allowed.
- If
. )
2
1 1
~ , X N   and
. )
2
2 2
~ , Y N   ,
o
. )
2 2
1 2 1 2
~ , X Y N     ± ± +
o
. )
2 2
1 1
~ , aX b N a b a   ± ±
o
. )
2 2 2 2
1 2 1 2
~ , aX bY N a b a b     ± ± +
Q: Solve for multiple unknowns, e.g.  and  .
A:
- For Binomial or Poisson, use GC or use the formula for . ) P X x = .
- For Normal, if  ,  , or both are unknown, always standardize X to Z .
1. If
. )
2
~ , X N   , then let
X
Z


÷
= .
2. Then . ) ~ 0,1 Z N and . )
X
P X x P Z


÷ ´ }
< = <
]
' ¹
.
3. Since for Z , both mean (0) and standard deviation (1) are known,
use InvNormon the GC to find an equation for
X 

÷
.
4. Solve simultaneous equations if necessary.
Example 1 [N01/II/8]
The random variable X has a normal distribution with . ) 7.460 0.01 P X > = and
. ) 3.120 0.25 P X < ÷ = . Find the standard deviation of X .
200 independent observations of X are taken.
(i) Using a Poisson approximation, find the probability that at least 197 of these
observations are less than 7.460.
(ii) Using a suitable approximation, find the probability that at least 40 of these
observations are less than -3.120.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
113
Solution Comments
Let
2
~ ( , ) X N   .
Let
x
Z


÷
= , then ~ (0,1) Z N .
. )
7.460
7.460 0.01 P X P Z


÷ ´ }
> = > =
]
' ¹
Since  and  are unknown for X , we
cannot immediately use InvNorm, and
standardization is necessary.
7.460
0.99 P Z


÷ ´ }
< =
]
' ¹
Once you get an equation in this form,
InvNorm can be used as  and  are
known for Z .
From the GC,
7.460
2.3263


÷
= (5 s.f.)
Key in (0.99) InvNorm . If you do not
specify  and  , the GC will assume
0  = and 1  = , as with the standard
normal distribution.
(working omitted)
From the GC,
3.120
0.67449


÷ ÷
= ÷ (5 s.f.)
Similarly, standardize and use InvNorm
for the other equation from the question,
. ) 3.120 0.25 P X < ÷ = .
2.3263 7.460
0.67449 3.120
 
 
+ = ¹
´
÷ = ÷
¹
From the GC, 3.53  = (3 s.f.) 
Write the system of linear equations and
use PlySmlt2 on the GC to solve for 
and  .
(i) Let Y be the number of observations not
less than 7.460 out of 200 observations.
Then ~ (200, 0.01) Y B .
Since 200 n = is large (>50) and 0.01 p =
is small such that 2 5 np = < ,
~ (2) Y Po approximately.
We define Y with “not less than” rather
than “less than”, because p must be
small for a Poisson approximation.
Remember to write “out of 200
observations”.
. ) 200 197 P Y ÷ >
. ) 3 0.857 P Y = < = (3 s.f.) 
200 Y ÷ is the number of observations
less than 7.460, as required.
(ii) Let W be the number of observations less
than 3.120 ÷ out of 200 observations.
Then ~ (200, 0.25) W B .
Z is reserved for the variable with the
standard normal distribution, so define the
variable as W instead.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
114
Since 200 n = is large such that
50 5 np = > and . ) 1 150 5 n p ÷ = > ,
. ) ~ 50, 37.5 W N approximately.
To approximate a binomial variable with
moderate p e.g. 0.25 , use a normal
approximation. If p is extreme e.g. 0.01,
use a Poisson approximation.
. ) 40 ( 39.5) P W P W > = >
(by continuity correction)
0.957 = (3 s.f.) 
Remember to key in 37.5  = for
normalcdf on the GC, not
2
37.5  = .
Example 2 [AJC05/II/27]
The weights of large eggs are normally distributed with mean 65 grams and standard
deviation 4 grams. The weights of standard eggs are normally distributed with mean 50
grams and standard deviation 3 grams.
(a) One large egg and one standard egg are chosen at random. Find the probability that
the weight of the standard egg is more than
4
5
the weight of the large egg.
(b) Standard eggs are sold in packs of 12 while large eggs are sold in packs of 5.
(i) Find the probability that the weight of a pack of standard eggs differs from
twice the weight of a pack of large eggs by at most 5 grams.
(ii) A customer bought two packs of large eggs. Find the value of m (to the nearest
0.1 gram) if there is a probability of 0.03 that the average weight of the eggs in
one pack is more than the average weight of the eggs in the other pack by at
least m grams.
Solution Comments
(a)
. ) . )
2 2
Let represent the weight of a large egg, in grams.
Let represent the weight of a standard egg, in grams.
Then ~ 65, 4 and ~ 50, 3
4
Required probability is
5
X
Y
X N Y N
P Y X
´ }
>
]
' ¹
. )
2
2
4 4
50 (65) 2
5 5
4 16
Var 3 4 19.24
5 25
E Y X
Y X
´ }
÷ = ÷ = ÷
]
' ¹
´ }
÷ = + =
]
' ¹
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
115
. )
4
~ 2,19.24
5
4 4
0 0.324 (3 s.f.)
5 5
Y X N
P Y X P Y X
÷ ÷
´ } ´ }
¬ > = ÷ > =
] ]
' ¹ ' ¹

(b)(i)
. ) . )
. ) . )
. )
. ) . )
. )
. ) . )
. ) . )
1 2 12 1 5
1 12 1 5
1 12 1 5
2
1 12 1 5
1 2 12 1 5
1 12
Required probability
... 2 ... 5
... 2 ...
50 12 2 5 65
50
... 2 ...
12 3 4 5 16
428
... 2 ... ~ 50, 428
... 2 ... 5
5 ...
P Y Y Y X X
E Y Y X X
Var Y Y X X
Y Y X X N
P Y Y Y X X
P Y Y
= + + + ÷ + + <
+ + ÷ + +
= × ÷ ×
= ÷
+ + ÷ + +
= × + ×
=
+ + ÷ + + ÷
+ + + ÷ + + <
= ÷ < + + ÷ . ) . )
. )
1 5
2 ... 5
0.0109 3 s.f.
X X + + <
= 
A pack of standard eggs
contains 12 different eggs,
hence each egg is an
independent observation of
Y
.
. )
. ) . )
. )
1 5 6 10
1 5 6 10
... ...
~ 0, 6.4
5 5
... ...
Let
5 5
| | 0.03 0.015
From G.C., 5.5 nearest 0.1g
X X X X
N
X X X X
Q
P Q m P Q m
m g
+ + + +
÷
+ + + +
= ÷
> = ¬ < ÷ =
= 
Even though one pack is
more than the other, use the
modulus sign because either
pack could be the heavier
one.
Example 3 [NJC05/II/30]
A Geography student is studying the distribution of bullfrogs in a large rural field where there
is an average of 500 bullfrogs per 400 km
2
. One part of the field is identified and divided into
50 equal squares, each with sides measuring 2 km. The student wishes to model the
distribution of bullfrogs in each square by using the Poisson distribution with mean  .
(i) State the value of  and one assumption made in using the Poisson distribution.
(ii) Four squares are selected at random. Find the probability that two of these squares
contain no bullfrogs and each of the other two squares contains at least 2 bullfrogs.
(iii) Using a suitable approximation, find the probability that, out of 50 squares, there are at
least 35 squares that contain at most 5 bullfrogs each.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
116
(iv) Determine the approximate probability that the total number of bullfrogs in 10 randomly
selected squares is between 40 and 60 (not inclusive).
(v) The student suggests using the same model on another rural field in another country.
Comment on the suitability of the model in this situation.
Solution Comments
(i)
2
Area of a square is 4km .
400
Total squares in field 100.
4
500
5
100

= =
= = 
Assumption: The event that a bullfrog lies in a
square is independent of other bullfrogs. 
Do not write the “probability” that
a bullfrog lies in a square is
independent. Use the word
“event”.
Alternative assumption
Average rate of occurrence of
bullfrogs in the field is constant.
(ii)
. )
Let be number of bullfrogs in a square.
Then ~ 5
X
X Po
. ) . ) . )
. )
2
2
Required probability
4!
[ 0 ] 1 1
2!2!
0.000251 3 s.f.
P X P X = = ÷ < ×
=
Use P & C:
There are
4!
2!2!
permutations of 4
objects where 2 objects are of type
A and the other 2 are of type B.
(iii)
. ) 5 0.61596 P X < =
Let Y be the number of squares that contain at
most 5 bullfrogs, out of 50.
. )
. )
. )
. )
. )
. )
~ 50, 0.61596
Since 50 is large,
30.8 5, 1 19.2 5,
~ 30.798,11.828 approximately.
35
34.5 (by continuity correction)
0.141 3 s.f.
Y B
n
np n p
Y N
P Y
P Y
=
= > ÷ = >
>
= >
= 
If np > 5, try approximating
binomial to normal. Otherwise, try
binomial to Poisson, since these
are the only two possible cases.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
117
(iv)
. )
Let Q denote number of bullfrogs in 10 squares.
Then ~ 50 Q Po
. )
. )
. )
Since ,
~ 50, 50 approximately.
P 40 60
(40.5 59.5) (by continuity correction)
0.821 3 s.f.
Q N
Q
P Q
 = 50 >10
< <
= < <
= 
(v) The number of bullfrogs may not be randomly
distributed in another country. Therefore, the
model may not be suitable. 
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
118
Sampling & Testing
Sampling Methods
Example: Say we want to select a sample of 50 workers from a company with 200 workers.
40 workers are under 30 years old, 100 are 30 to 40 years old, and 60 are above 40 years old.
Simple random sampling Systematic sampling
Random Random
Method Assign each worker a number from 1 to 200.
Select a sample of 50 workers
using a list of random
numbers.
As
200
4
50
= , randomly select one worker from
the first 4. If say the 3
rd
worker is chosen, include
every 4
th
worker from the 3
rd
worker in the
sample, i.e. the 3
rd
, 7
th
, 11
th
, …, 199
th
workers.
Pros - Even spread over population
Cons - Requires sampling frame
- Cannot replace unavailable
sampling units
- Time-consuming to locate
selected units
- Requires sampling frame
- Cannot replace unavailable sampling units
- Biased if members of the population have a
periodic or cyclic pattern
Stratified sampling Quota sampling
Random Non-random
Method Select random samples from the age
groups with sample size proportional to the
relative size of each age group, i.e.:
Age Sample size
30 < 20
30 – 40 50
40 > 30
Same as stratified, except:
- Sample size need not be (but can
be) proportional
- Non-random samples are drawn
from each age group
Pros - Likely to give more representative
sample
- Lower cost
- Less time-consuming
- Does not require sampling frame
Cons - Requires sampling frame
- Cannot replace unavailable sampling
units
- Strata may not be clearly defined
- Non-random
- Likely to give less representative
sample
- Biased as researcher may just pick
members who are easier to collect
data from
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
119
FAQ (Sampling)
Questions on sampling methods Use the context of the question in your answer.
- Identify type of sampling method used
- Explain why:
o Method is inadequate: list cons
o Method is biased/unbiased
o Sample is likely to be representative or not
- Suggest a better method if required
- Describe your suggested method
Given x
¯
and
2
x
¯
,
or . )
2
x x ÷
¯
,
- Find unbiased estimate x of
population mean 
- Find unbiased estimate
2
s of
population variance
2

1
x x
n
=
¯
. )
. )
2
2
2 2
1 1

1 1
x
s x x
n n
x
n
´ }
]
= ÷ = ÷
]
÷ ÷
' ¹
¯
¯ ¯
Find x and
2
s ,
given ( ) x a +
¯
and . )
2
x b +
¯
.
- Let y x b = + .
- y
¯
. ) x a b a = + + ÷
¯ ¯ ¯
. ) ( ) x a n b a = + + ÷
¯
- . )
2
2
y x b = +
¯ ¯
- Find y and
2
y
s (see above).
- x b y = ÷ ,
2 2
x y
s s =
Find x and
2
s ,
given a list of data.
GC gives
x
s , square the value to get
2
x
s .
- Individual data
o Values in
1
L
o Key in: 1-Var Stats
1
L
- Grouped data
o Values in
1
L , frequencies in
2
L
o Key in: 1-Var Stats
1 2
, L L
Testing
Do this: Only if:
Use t -test instead of z -test - sample is small ( 50 n < ), and
-
2
 is unknown
Use
2
s instead of
2
 -
2
 is unknown
Use Central Limit Theorem
- sample is large ( 50 n > ), and
- population distribution is unknown
Assume population distribution is normal
- sample is small ( 50 n < ), and
- population distribution is unknown
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
120
If the sample is large, population is normal, and
2
 is unknown, Central Limit Theorem is
not required. However, since
2
s must be used as an unbiased estimate of
2
 , write
“approximately” after the distribution of X .
FAQ (Testing)
Explain what is meant by ‘a
%  significance level’ in the
context of the question
A  % significance level means that the probability of
concluding that
0
H is rejected given that
0
H is true is
100

.
Example: A 5% significance level means that the
probability of concluding that the mean mass of a chicken is
greater than 1.75kg when it is actually 1.75kg is 0.05.
State the assumptions made, if
any
- If the sample is large ( 50 n > ):
o Central Limit Theorem holds, hence the
population follows a normal distribution
approximately, and no assumptions about its
distribution are needed.
- If the sample is small ( 50 n < ):
o Assume the distribution of the population is
normal
Compare outcomes of z -test
and t -test, given same data
Since the graph of the t -distribution has “thicker tails” than
that of the z -distribution about
0
 , given the same
significance level, the t -test would be lead to a conclusion
of
0
H being rejected for more values of x .
Compare effect of change in
significance level, sample
mean, or population mean on
outcome
- If
. )
0
/
p value P P Z
x
n
X x


÷ ´ }
÷ = < = <
]
' ¹
,
o If
0
/
x
n


÷
increases, p value ÷ increases
- If
. )
0
/
x
p value P X x P Z
n


÷ ´ }
÷ = > = >
]
' ¹
,
o If
0
/
x
n


÷
increases, p value ÷ decreases
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
121
Find the range of  , x ,
0
 ,
s , or n such that
0
H is
rejected (or not rejected) at a
%  significance level.
- Find an expression for the p -value
o 1-tailed:
. )
P X x < or
. )
P X x >
o 2-tailed:
 If
0
x  < ,
. )
2P X x <
 If
0
x  > ,
. )
2P X x >
- State the condition on the p -value for the outcome (For
e.g. a 5% significance level, 0.05 value  ÷ = )
o If
0
H is rejected, p value value  ÷ < ÷
o If
0
H is not rejected, p value value  ÷ > ÷
- Standardize, only if the mean
0
 or variance
2
s
n
of X
is unknown, i.e. when finding the range of
0
 , s , or n
- Use InvNorm to find the range of  , x , or
0
 that
satisfies the condition on the p -value
- Conclude in the context of the question
Example [TJC07/Prelim/II/9 (Modified)]
The manufacturer of the ‘Genius’ brand of cigarettes claimed that the average tar content of a
‘Genius’ cigarette is not more than 14mg per cigarette (mg/cig). A surveyor from a rival
company took a random sample of 8 ‘Genius’ cigarettes and lab measurements of the tar
content yielded 14.2, 14.5, 13.9, 14.1, 14.0, 14.1, 14.2, 14.4 mg/cig. Will he be able to reject
the manufacturer’s claim at 5% level of significance? State any necessary assumption made
by the surveyor in carrying out the test.
To confirm his findings, the surveyor took another random sample of 50 ‘Genius’ cigarettes
and measured the tar content x in mg/cig. The following results were obtained:
. ) 10 208.2515 x ÷ =
¯
, . )
2
14 5.2455 x ÷ =
¯
.
Based on this sample,
(i) Find unbiased estimates of the population mean and variance of the tar content in
mg/cig.
(ii) Find the range of values for the average tar content per cigarette that the manufacturer
should claim in order that it will not be rejected as an underestimate at the 1% level of
significance if a z -test is carried out.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
122
Solution Comments
Let X be the tar content of a cigarette in
mg/cig.
Always define X if the question does not.
Specify its units, if any (mg/cig).
Since 8 n = is small, the surveyor must
assume that X follows a normal
distribution.
Since the sample is small and the
distribution of the population is unknown, it
must be assumed to be normal.
0
: 14 H  = vs
1
: 14 H  > 1. State the null hypothesis
0
H and the
alternative hypothesis
1
H .
Perform a 1-tailed test at the 5%
significance level.
2. Write down 1-tailed or 2-tailed test, and
the level of significance α %.
Under
0
H ,
0
~ ( 1)
/
X
T t n
s n
 ÷
= ÷ ,
where 14.175 x = ,
0
14  = ,
0.19821 s = , 8 n = .
3. Decide on the test statistic (t or z) to be
used and determine its distribution.
x and
2
s will appear on the GC screen
with the p -value, so we do not need to
calculate them.
0
T
s / n
X ÷
= , not
0
σ / n
X ÷
, since σ is
unknown and we estimated it with s .
X is a random variable representing the
sample mean, which can have many values.
x is the mean of a particular sample,
usually given in the question.
Using a t -test, 0.0206 p value ÷ = (3 s.f.) 4. Use the GC to calculate p -value.
Since 0.0206 0.05 p value ÷ = < , the
surveyor will be able to reject
0
H and
conclude that there is sufficient evidence
at the 5% significance level that 14 X >
and the manufacturer’s claim is false. █
5. Conclude in the context of the question.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
123
(i) Let 14 y x = ÷ .
. ) 14 y x = ÷
¯ ¯
. ) . ) 10 50 4 8.2515 x = ÷ ÷ =
¯
. )
2
2
14 5.2455 y x = ÷ =
¯ ¯
8.2515
14 14
50
x y = + = +
14.165 (5 . .) s f = 14.2 (3 . .) s f = █
2
2
1 8.2515
5.2455
50 1 50
x
s
´ }
= ÷
]
÷
' ¹
0.079260 (5 . .) 0.0793 (3 . .) s f s f = = █
Always let y be the expression which is
squared in the summary statistics, e.g. let
y x 14 = ÷ when . )
2
x 14 ÷
¯
is given.
Give values which will be used in later
parts of the question to 5 s.f. first, then to 3
s.f.
(ii)
0 0
: H   = vs
1 0
: H   > .
Perform a 1-tailed test at the 1%
significance level.
Under
0
H , by the Central Limit Theorem
since 50 n = is large,
2
0
~ ,
s
X N
n

´ }
]
' ¹
approximately, where
14.165 x = , 0.079260 s = , 50 n = .
Translate the question into mathematical
terms: What is the range of
0
such that
0
H is not rejected when a z-test is carried
out at the 1% significance level?
The manufacturer’s claim is
0
H :
0
μ = ,
and if it were “rejected as an underestimate,
1 0
H :
μ μ
> would be true.
Let
0
/
X
Z
s n
 ÷
= . Then ~ (0,1) Z N .
If
0
H is not rejected, 0.01 p value ÷ > .
. )
0
14.165 0.01
/
x
P X P Z
s n
 ÷ ´ }
> = > >
]
' ¹
Since
0
is unknown for X , we cannot
immediately use InvNorm, and
standardization is necessary.
0
0.99
/
x
P Z
s n
 ÷ ´ }
< <
]
' ¹
From the GC,
0
14.165
2.3263
0.079260 / 50
 ÷
< (5 s.f.)
Solving,
0
14.1  > (3 s.f.) █
Once you get an equation in this form,
InvNorm can be used as μ and σ are
known for Z.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
124
Correlation & Regression
Product Moment Correlation Coefficient (PMCC)
r > 0 Positive Linear Correlation
r < 0 Negative linear correlation
r ≈ 0 No linear correlation
(Note that no linear correlation does not
imply no relation)
- The closer the absolute value of r is to 1, the stronger the linear correlation between 2
variables.
Limitations of PMCC
- Does not necessarily prove that there is a linear relationship between two variables
o e.g. there may be an outlier
- Unable to prove any non-linear relationship between variables using the coefficient
o e.g. points on a parabola can be chosen to give a PPMC of close to zero
Therefore, use BOTH the scatter diagram and the r-value together when making
judgements/statements/comparisons.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
124
Correlation & Regression
Product Moment Correlation Coefficient (PMCC)
r > 0 Positive Linear Correlation
r < 0 Negative linear correlation
r ≈ 0 No linear correlation
(Note that no linear correlation does not
imply no relation)
- The closer the absolute value of r is to 1, the stronger the linear correlation between 2
variables.
Limitations of PMCC
- Does not necessarily prove that there is a linear relationship between two variables
o e.g. there may be an outlier
- Unable to prove any non-linear relationship between variables using the coefficient
o e.g. points on a parabola can be chosen to give a PPMC of close to zero
Therefore, use BOTH the scatter diagram and the r-value together when making
judgements/statements/comparisons.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
124
Correlation & Regression
Product Moment Correlation Coefficient (PMCC)
r > 0 Positive Linear Correlation
r < 0 Negative linear correlation
r ≈ 0 No linear correlation
(Note that no linear correlation does not
imply no relation)
- The closer the absolute value of r is to 1, the stronger the linear correlation between 2
variables.
Limitations of PMCC
- Does not necessarily prove that there is a linear relationship between two variables
o e.g. there may be an outlier
- Unable to prove any non-linear relationship between variables using the coefficient
o e.g. points on a parabola can be chosen to give a PPMC of close to zero
Therefore, use BOTH the scatter diagram and the r-value together when making
judgements/statements/comparisons.
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
125
Relevant Formulae (In MF15)
Estimated Product Moment Correlation Coefficient
: :
. ) . )
2 2 2 2
2 2
( )( )
( ) ( )
x y
xy
x x y y
n
r
x x y y
x y
x y
n n
÷
÷ ÷
= =
´ }´ }
÷ ÷
] ]
÷ ÷
] ]
' ¹' ¹
¯ ¯
¯
¯
¯ ¯
¯ ¯
¯ ¯
Estimated regression line of y on x
2
( )
( )( )
( )
y y b x x
x x y y
b
x x
÷ = ÷
÷ ÷
=
÷
¯
¯
Although not in MF15, the formula for the regression line of x on y can be obtained by
swapping x with y in the above formula for the regression line of y on x)
Linear Regression
The least square regression line is the line which minimises the sum of squares of the
residuals. (Note that the a residual is the difference between the observed & predicted values.)
Scatter Diagrams
- Label each axis with the variable and its units.
- Ensure the relative position of the data points is correct.
- Display the origin at the correct relative position. If the data is far away from the origin,
use a jagged line to indicate an omitted range of values.
GC Tips & FAQ
1. How do I key in the data set into the GC?
Key STAT EDIT Type in your data
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
126
2. When a question requires me to plot ln y against x, but it only provides data set for x
and y, what can I do other than manually evaluating each value of ln y?
For example, if the data given is a list of values of y, but the question requires you to plot
ln y against x, you can do so using GC. For example, y values are in L2. Key all your
values into the lists, with x values in L1, y values in L2. Then, key ln(L2) STO L3
3. What is the significance of the intersection point of regression lines of x on y and y
on x?
The intersection point of regression lines of x on y and y on x is ( ̅, ).
4. How do you assess whether an estimate is reliable?
- First, we look at the r-value. If it is close to 1 or – 1, then the estimate would be reliable.
- Next, we look at whether it is an interpolation or an extrapolation. Assuming r is close to
1 or – 1, an interpolation would be very reliable.
- In the case of an extrapolation, we would look the amount of extrapolation. If it is close to
the experimental range of values, then that estimate is reliable. Otherwise, it is not
reliable.
Note that you should use the scatter plot in line with the r-value.
Example 1 [SAJC08/Prelim/II/6]
Each of a random sample of 10 students are asked about the average number of minutes
spent on doing mathematics tutorials in a week (x), and their percentage score for the
mathematics final examination (y). The results are tabulated below:
x 20 35 45 60 70 80 100 110 120 140
y 16 25 35 50 60 65 70 75 80 85
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
127
(i) Find the equation of the regression line of y on x.
The question is indirectly saying that y is the dependent variable and x is dependent.
Key all the values into the lists in GC. Remember to turn on the diagnostic function.
From G.C.,
. ) 0.591 10.0 3 . . y x s f = +
[Whether you use LinReg(ax+b) or LinReg(a+bx), linear regression should be done with L1,
L2 where L1 is the independent variable and L2 is the dependent variable.]
(ii) Find the linear product moment correlation coefficient between y and x, and
comment on the relationship between x and y.
Use the G.C values calculated earlier
. ) 0.972 3 . . r s f ~
(iii) Use the appropriate regression line to estimate the percentage score of a student
who spends 10 minutes doing mathematics tutorial in a week. Comment on the
reliability of the estimate.
. ) 0.591 10.0 3 . . y x s f = +
Using G.C, 15.9 y ~
However, the estimate is not reliable as 10 minutes is well outside the given range of x values
so the approximation would be unreliable.
Example 2 [HCI08/Prelim/II/8 (Modified)]
A teacher decided to investigate the association between students’ performances in
Mathematics and Physics. She selected 5 students at random and recorded their scores,
x and y, in Mathematics and Physics tests respectively. She found that, for this set of
data, the equation of the regression line of y on x was 18.5 0.1 y x = + and the equation of
the regression line of x on y was 16.6 0.4 x y = + .
Show that 125 =
¯
x and 105 =
¯
y .
18.5 0.1
16.6 0.4
25 and 21
5 25 125 and 5 21 105
x y
y x
x y
x y
¹
+ =
¦
´
+ =
¦
¹
= =
= × = = × =
¯ ¯
Remember that the intercept of the
regression lines (y on x and x on y)
is
. )
, x y .
Combination of concepts from
sampling
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
128
(i) The Physics score of a sixth student was mislaid but his Mathematics score was
known to be 26.
Use the appropriate line of regression to estimate this student’s Physics score,
and give a reason for the use of the chosen equation.
18.5 0.1
When 26, 21(nearest integer)
y x
x y
= +
= =
Since x is given, y is the dependent variable and therefore we use the regression line of y on x.
(iii) Comment on the reliability of your estimate.
The estimate is not reliable because | | r is small, i.e. it is close to 0, hence the linear
correlation between x and y is very weak, indicating that the linear model is not suitable.
Example 3 [NYJC08/Prelim/II/11]
A student wishes to determine the relationship between the length of a pendulum, l, and the
corresponding period, T. After conducting the experiment, he obtained the following set of
data:
l/cm 150 135 120 105 90 75 60 45 30 15
T/s 2.45 2.31 2.22 2.07 1.91 1.74 1.56 1.35 1.10 0.779
(i) Obtain the scatter plot of this set of data. [2]
(ii) The student proposes two models.
. )
2
: ln
:
A T a b l
B T a bl
= +
= +
Calculate the product moment correlation coefficients for both models, giving your
answer to 4 decimal places. Determine which model is more appropriate. [3]
(iii) Using the model determined in part (ii), estimate the value of l when T = 3s to 1
decimal place. Comment on the suitability of this method. [3]
(iv) Find a value of l and its corresponding value of T such that the equation of the
regression line for the chosen model will remain the same after the addition of this pair
of values. [3]
Mathematics Revision Notes 2009
RI(JC) Mathematics Society
129
Solution Comments
(i) Note that your scatter plot must be
to scale. This means that your
points should be relatively correct
to each other.
Axis should be labelled correctly.
(ii) Using a G.C.,
r for model A is 0.9871. (4 d.p.)
r for model B is 0.9996. (4 d.p.)
Since the value of r for model B is closer to 1 as
compared to model A, model B is more
appropriate.
Be careful as question asks for 4
decimal places.
For appropriateness of model,
look at r values.
(iii) Since T is given, we should use the regression line
of l on T
2
.
From G.C., the value of l is 223.9 (1d.p.)
Although r is close to 1, the value of T = 3s lie
well outside of the range of the T values used to
plot the regression line. Hence it is not a good
estimate of l.
Even though it is in the Physics
mindset that the period depends
on the length of the pendulum, it
is NOT the regression line of T
2
on l.
Be aware that model B uses T
2
instead of T.
(iv)
The values to be used are l and
2
T .
82.5 l l = = and . )
2
3.3301 5 s.f. T =
. ) 1.82 3 s.f. T T = =
Use G.C. to evaluate the values.
Note that you can plot the
regression lines for both l on T
2
and T
2
on l. The intersection of
the two lines will obviously be the
point you are looking for.

Mathematics Revision Notes 2009 RI(JC) Mathematics Society

Acknowledgements
Members of RI(JC) Mathematics Society, batch of 2009, for their time and effort in creating these notes; o o o o o o o o o o o Bi Ran Hu Yi Jie Kan Wai Keong Desmond (Editor) Koo Cheng Wai (Editor) Kung Ying Xiang Liew Xiao Tian Ong Chin Hwee Samuel Lim Yong Peng Tan Chor Hiang Tan Jun Xian Gabriel Tan Wei Jie

Mr. Kok Chee Kean, teacher-in-charge, for his guidance and support; Members of RJC Mathematics Society, batch of 2008, for pioneering this project; and you, the reader, for supporting our work.

2

Mathematics Revision Notes 2009 RI(JC) Mathematics Society

Functions
A function f is a relation which maps each element in its domain D f to one and only one element in its range R f , e.g. f : x x 2 1 , x , where x2 1 is the rule of f and is the domain of f . Use the Vertical Line Test to determine if a given relation f is a function. Replace D f and R f below with the actual domain and range of f in interval notation. One-one Functions Use the Horizontal Line Test to determine if a given function is one-one. e.g. f : x

x2 1 , x
y

and g : x

x2 1 , x 0 .
f ( x)
any line y k , k Rg [1, )
y

y

y

g ( x)

k

y

2

2 1

x

1

x

0

0

Since the horizontal line y 2 , where Since any horizontal line y k , where 2 R f [1, ) , does not cut the graph of f k Rg 1, , cuts the graph of g at one at one and only one point, f is not one-one. and only one point, g is one-one. OR Since f
1
Df

f 1

2,

where 1,1

, f is not one-one.

g is a restriction of f if f and g have the same rule and D g

D f , i.e. g is f with its

domain restricted. Even if f is not one-one, a restriction of f which is one-one can be defined. Use turning points or points where f is undefined to show that f is not one-one or restrict its domain.

3

given the rule of f : 2. with its domain restricted. D ff 1 D f 1 . since the graphs of f and f 1 must intersect on x . but D f 1 f D f . Q: What is the domain/range of f ? A: Observe from the rule of f by substituting values. 1 x or f x x instead. A: Substitute the rule of f . such that x 1 g y . Df Rg . The rules of ff 1 and f 1 f are always x . ) . Let y 3. Composite Functions gf is the composite function of f followed by g . Usually. is a function such that for all x y x if and only if f x y. x . Domain of f 1 R f Range of f 1 Df 1 To find the rule of f 1. Then f f x . with the domain of g restricted to R f . the inverse function of f . then sketch the graph using GC. Q: Solve f x A: Solve f x the line y f 1 Dg . x g x . gf exists if and only if R f Domain of gf Range of gf FAQ Q: Sketch the graph of f . Q: What is the minimum value of k such that f . Express x in terms of y . f 1 exists if and only if f is one-one. or from the graph using GC. passes the Horizontal Line Test.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Inverse Functions f f 1 1 . x k will be a turning point or point where f is undefined on the graph of f . is one-one? A: Find the minimum value of k such that f . where g y is a function of y . (Example 2) 4 . with its domain restricted to ( k . but they usually have different domains. D f and y Rf .

such that the in a similar form. Df 1 R f . x 1. where x Df . Solution (a) Largest possible D f ( . since both ff 1 and f 1 f have rule x . at x 1 by on the GC or based on the looking for Df 1 Rf 1.1 h:x x 2 2 x 1. Restrict D f ( x 1) 2 discontinuity instead of turning points. where x Q: Sketch the graphs of f ( x ) and f 1 ( x ) (on a single diagram).Mathematics Revision Notes 2009 RI(JC) Mathematics Society Q: Solve ff 1 x f 1f x . A: The solution set is the intersection of the domains of f and f 1 . A: Show their geometrical relationship: the graphs of f and f 1 are reflections of each other about the line y x .1) Comments f 1 exists if and only if Observe a discontinuity sketching the graph of f 1 from . Q: Sketch the graph of ff 1 (or f 1 f ). Hence define f 1 . Give the rule and domain of the composite function hg and find its range. State the largest possible domain of f in the form inverse function of f exists. Similarly. f is one-one. the f is the section of the line y x . x . Example 1 [NJC06/CT/Q9 (Modified)] A mapping f is given by f : x 1 ( x 1)2 1. x Determine whether the composite function gh exists. 5 . where b . b . Use the GC to find R f . A: The graph of ff 1 is the section of the line y graph of f 1 x . The functions g and h are defined as follows: g:x ln( x 2).

x . Dg .1) Rg Rhg g 1 . (b)(ii) Dhg Dg 1.1 hg x hg : x h g x ln x 2 h ln x 2 2 1 You do not need to show hg exists. 2.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Let y f x 1 x 1 2 1 x 1 2 1.1 . and x 1 or 1 1 y 1 0 for all y .1 . Since h 0 h(ln 3) . h( x) decreases from 0 to 1 . Rh with Dh restricted to R g Rhg h 1 . y 1 x 1 x 1 x 1 2 1 y 1 1 y 1 1 y 1 1 y 1 1 (rejected) y 1 Use D f to decide which of the 2 expressions for x to accept and which to reject. hence gh exists if and only if Rh the composite function gh does not exist.ln 3) g x is strictly increasing on ( 1. D f . x x 1 2x 1 2. .1 ) Rh Dg . ln 3) . 2 Express f 1 in a similar form. 2 ln x 2 x ( 1.g 1 (0. f (b)(i) 1 :x x2 1 1 . ( 1. x 1 1. h x Rh Since 2 h( x) has a minimum point at x 1 . then increases from 1 to ln 3 . 1 In the restricted Dh R g (0.h 0 2.1) . h 1. h(0) is the larger endpoint of Rhg . 6 . Here. since the question already assumes it. and write down both its rule and domain. so x 1 . . and its domain includes 1 . reaches its minimum at h 1 2 .

1 2 (i) D ff 1 Df 1 The rules of ff 1 and f 1 f are always x . f Solution 1 and f 1 1 f . 2 in the domain of both ff 1 and f x D ff 1 D f 1 f R f D f f .e. key in DrawInv( Y1 ) to draw the inverse of the graph of Y1 . 1 0. 2 2 y f (x ) 1 2 y ff 1 (x ) x GC syntax: Set Y1 1 / (1 e ^ ( X )) . solve ff 1 x f 1 f x . x and ff x on a single sketch.x 1 ex 0 1 By considering the domains of ff Show the graphs of f ( x) .: y f 1 (x) y x Sketch the graphs on the GC first.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 2 [HCI06/Promo/Q11 (Modified)] The function f is defined by f :x 1 . Comments Rf 0. ff 1 Df 1 f Df 1 [0. then: [2ND][PRGM] 8:DrawInv [VARS] Y-VARS Function… [ENTER] Y1 7 . When the rule of f 1 ( x) is unknown. 2 1 x 2 y 0. but they may have different domains. i. 1 2 0 1 1 . Hence. f 1 ) f ( x) . x f 1 For ff x x f x if and only if x is 1 0 (ii) 1 0.

k ) Hyperbola ( x h )2 a2 ( y k )2 b2 1 ( y k )2 b2 ( x h)2 a2 1 8 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Graphing Techniques Graphs of Conic Sections The dotted lines represent lines of symmetry or asymptotes. k ) x h ( x h )2 ( y k )2 r2 y k r Circle (h. k ) x h x h x h b( y k ) 2 y k b( x h) 2 Parabola y k (h. b Ellipse ( x h) 2 a2 ( y k )2 b2 y k 1 a (h. k ) (h. not the axes.

then the a2 b2 graph looks like a handwritten x (with left/right parts) and always intersects the x axis. When the equation of a conic is given in general form. k Horizontal line of symmetry: x h Vertical line of symmetry: y k b x h a b x h Oblique asymptote (sloping downwards): y k a Oblique asymptote (sloping upwards): y k ( x h) 2 ( y k ) 2 1 . If the x term comes first in the equation of a hyperbola. k 0 ). Centre: h. to identify the conic: Is there both x 2 term and y 2 term? Yes Are the coefficients of x 2 term and y 2 term of same sign? Yes Are the coefficients the same value? Yes Circle No Ellipse Hyperbola Parabola No No Is there either x 2 term or y 2 term? Yes 0. i. 9 . If the equation of a conic is given in the general form Ax 2 By 2 Cx Dy E complete the square to convert it into the standard forms given in the table above.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Properties of Hyperbolas For both forms of hyperbolas above. Sketching Graphs of Conic Sections There is a special case for each conic where its centre or vertex is at the origin (i.e. k ) .e. To draw a conic with centre/vertex (h. translate this special case by replacing x with x h and y with y k . h 0 .

y y a negative y -direction positive x -direction negative x -direction Replace y by y a Replace x by x a Replace x by x a i. express y as px q Vertical asymptote: x o When dx e 0 . hence y o As x . hence y dx e Let the initial graph be of y Translating Graphs Translate by a units in the: positive y -direction In the equation Replace y by y Equation after transformation f ( x) . p. q . dx e Oblique asymptote: y px q r o As x .e. where p . px q . d c 0. and let a be a positive constant. y f ( x) f x a y y f ( x a) f ( x a) 10 . and r are constants. cx d q y ax 2 bx c . hence y is undefined. where c cx d 0 . y a a f ( x) f x a i.e.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Graphs of Rational Functions y ax b . 0 . numerator and denominator have no common factor. numerator and denominator have no common factor. where a. d dx e 0 . Using long division. e d r r . cx d a Horizontal asymptote: y p c q 0 . hence y is undefined. where p and q are constants. express y as p Vertical asymptote: x o When cx d q cx d . Using long division. dx e is undefined. is undefined.

which is drawn with a dashed line below. Reflect the part of the graph of = ( ) where > 0 in the -axis. with x -axis invariant y f ( x) a i. y af (x) y y y Replace y by a Stretch parallel to the x -axis by x Replace x by factor a . y f ( x) x y f ( x) Transformations Involving Modulus y f (x) y y f ( x) y y f x y 0 x 0 x x 0 x Reflect the part of the graph Delete the part of the graph of = ( ) where ( ) < 0 of = ( ) where < 0. y y f '(x ) 1 f ( x) y2 f (x ) o Related to y f ( x) or y f ( x) 11 . in the -axis.e.e. Other Transformations Assume the same graph of y Other transformations include: f x as above. with y -axis invariant a Reflect in the x -axis ( y Reflect in the y -axis ( x f y x a f x 0) 0) Replace y by Replace x by i.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Stretching & Reflecting Graphs Stretch parallel to the y -axis by factor a .

x Approaches 0 Maximum point Minimum point 1. 1 distinct real root: vertical tangent 2 equal real roots: sharp tangent 2 equal real roots: horizontal tangent 0 x 3. f x by reflecting the f x in the x -axis. Combine the f x .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Graph Graph of y y f x Transformed graph y f '( x ) Stationary point Increasing Decreasing x -intercept Positive Negative Increasing Decreasing Maximum point Minimum point Positive Negative Decreasing Increasing 0 x Concave upwards Concave downwards Maximum gradient Minimum gradient y 1 f ( x) Positive y Negative Increasing Decreasing 0 Approaches or Approaches 0 Approaches Minimum point Maximum point f x where f x 0. Look at x -intercepts of y f x to decide the kind of tangent on the transformed graph. This is the graph of y 4. and y f x to get the graph of y 2 12 . Discard the part of y or y 2 f (x ) y 2. graphs of y f x f x . Find the graph of y graph of y 5.

k 2x 3 . y f ( x) Use GC to graph normally or use its Conics application.g. For points. replace x by x last. e. 3. Work from the left of the graph ( x ) to the right ( x ). Use substitutions. to get a graph symmetric about the y -axis. perform transformations from the innermost to outermost brackets. Find the value of k for which the x -axis is tangent to C . find the range of values of x that satisfy the inequality x 2 3x 1 4 x 6 . Check your answer. Identify critical features on the original graph. Hence. e. hence replace x by x first. then transform x to 2 x 3 . equations of asymptotes. which are given on the original graph. Obtain the equations of the asymptotes of C . 2. when f (x) 10 .g. 4. stationary points etc. 1 0. write ' .g. write “translate by 3 units…”. 13 . not “translate for 3 units…”. Write the exact phrasing of the transformations when the question asks you to “describe” them. General Tips Generally. using a graphical method. Example [YJC/I/Q8] The curve C has equation y (i) (ii) (iii) x2 kx 1 . Sequences of Transformations Involving Modulus Example: Transform the graph of y Solution: f ( x) f ( x) to that of y f 2 3x f (2 x) f (2 x) f (2 3x ) f 2 3x When x is involved. e.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Sketching a Transformed Graph 1. label all corresponding coordinates of points. Example: Transform the graph of y Solution: f ( x) f ( x) to that of y f 2 3x f x f 2 x f 2 x f 2 3x Here 2 3x is inside the modulus. On the new graph.1. use A ' to denote point A after transformation. Sketch the case for k 3 .

x 2 2 2 4 Graphs of y asymptotes: ax 2 bx c dx e usually have 2 1. D 0 : 2 distinct real roots. D 0 : No real roots. k2 4 k 2 or 2 (rejected k 2 k 2 to check 14 . Use the GC to plot the graph for k your answer. take y 13 6k x k 3 0 and y such that . Furthermore. 0. 4 x 2 k 2 3 4 y x2 kx 1 2x 3 x k 3 13 6k 2 2 4 4(2 x 3) Equations of asymptotes: 3 x k 3 and y . set the x term in the denominator 0 . D 0 : 1 (repeated) real root. 2 x 3 0 . y 0 x2 kx 1 0 Since there can only be 1 value of x where y 0 . differentiation cannot be used. 4(2 x 3) 2 2 4 (ii) Where x -axis is tangent to C . To find the vertical asymptote.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution (i) Comments Use long division to get a quotient of and a remainder of 13 6k . D k2 4 1 1 0 For a quadratic equation ax 2 bx c discriminant D b 2 4ac . i. To find the oblique asymptote. if the question states “algebraically”. ) Always try using the discriminant with quadratic equations in Graphing Techniques questions. 2. Differentiation is tedious and will result in a complicated expression. a 0.e.

Manipulate given inequality until 1 side. x or x 3 2 5.f. x above. and sketch it using the GC. The solution usually involves points of intersection of the 2 graphs. y 3x 1 3x 1 x2 3x 1 . say LHS.21 or x 2. 3 Exclude the values of x for which the function is undefined.79 3 2 x x 3 2 1.79 or 1. Sketch graph of RHS on GC and your diagram. 2.g.79 Substitute k 3 into the equations of the graph and its asymptotes. resembles equation of graph. 15 .) and asymptotes.333 0 0.62 0.382 2 x2 3x 1 2x 3 y 2 x2 3x 1 2x 3 2 x y 2 y 2 or y 2 From the graph. Label all axis intercepts (to 3 s. “Graphical method” suggests: 1. the blue marks on the x -axis in the diagram above. 2 We can divide both sides of the inequality by 2 x 3 because it is always positive. 2x 3 4x 6 2 2x 3 y 2.g.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iii) y x2 3x 1 2x 3 y y x 2 3 4 When k x2 x2 3. e. e. 3.

To write the set of values. 6. o e. 2] Check for any values (especially endpoints) to be excluded 1 o e. Eliminate factors which are either always positive or always negative. e. 5. NOT (1. Write the range of values of x (e. Model a problem with a system of linear equations and solve the system. ln x . Factorize the terms as far as possible. e x . etc. 2. {x :1 x 2} .) in x the original inequality. x 2 when the inequality contains x 2 o e. Zero the inequality (bring all the terms to one side. substitute x with the same function f ( x) in the original solution set to obtain the new x. Using the GC Find intersection between 2 graphs: [2ND][TRACE] 5:intersect o Find roots of a graph: [2ND][TRACE] 2:zero Find stationary points of a graph: [2ND][TRACE] 3:minimum (or 4:maximum) Solve system of linear equations: [APPS] 5:PlySmlt2 FAQ Solve a given inequality. such that the other side is 0). such that it becomes the new inequality. 4. multiplying it on both sides preserves the inequality sign. 3. f ( x) (usually 1 . x2 4 x 5 x 2 2 1 1 0.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Equations & Inequalities General Approach 1. solution set. o Substitute x with some function of x . Draw number line and graph and shade the regions where the inequality holds.g. use set-builder notation. Use g x Since g x 2 is always positive. Complete the square. 1 x 2 ).g. using algebraic or graphical approaches.g.g. Using the solution to an original inequality. NOT interval notation. whether the inequality is strict ( ) or not ( ). for all x . Then. f x 0 f x g x 0 to convert the inequality to a product of factors.g. solve a new inequality. Combine the terms to form a single fraction. 16 .

g.g. if they are either both positive or both negative 1 1 e. a b . do not cross-multiply inequalities! Cross-multiplying only works if the denominators on both sides of the inequality are either both positive or both negative. Use the GC if the question does not specify “exact” values. Remember to flip the sign of the inequality when you: o Multiply or divide both sides by a negative number 2a 2b e. e. x2 9 Solution x 9 1 x2 9 x 9 1 0 x2 9 Comments Use the General Approach outlined above. a or f ( x) a o OR Square both sides to get f x 2 a 2 . where a.Mathematics Revision Notes 2009 RI(JC) Mathematics Society General Tips In general. where f x f x f x a a a f ( x) f x a where a is a positive constant.g. 17 . f x o Divide it into cases. But be careful when using the GC directly: some solution intervals are not visible at certain zoom levels! (Example 1) Given an inequality involving modulus. a b o Take the reciprocal of both sides. 0 and f x a 0. b 0 ) a b Apply your numerical answer to the context of the question. b 0 (or a. Zero the inequality. 1. if appropriate. Example [N2006/II/Q1] Solve the inequality x 9 1. removing the modulus.

0 . If f ( x) is a polynomial with the coefficient of the highest power of x being positive. exclude x 3 and x 3 where the inequality is undefined. Draw number line and graph and shade the regions where inequality holds. Since the original inequality has denominator x2 9 x 3 x 3 . Do NOT use commas: e. positive. 1. x x 3 . Write out “or” between every 2 adjacent inequalities. the graph of f ( x) will start as positive from the right. 3. The graph has a “wavy” shape. Factorize the terms as far as possible. x 3 or 0 x 1 or x 3 Write the range of values of x . The graph has x -intercepts at the roots. 18 . here x 4 .g. so that the graph in Step 5 will start as positive from the right. since the terms are already linear. checking for any values (especially endpoints) to be excluded. Skip Step 4. 5. 3 0 1 3 x Mark the roots 3 . Combine the terms to form a single fraction. x2 9 x x2 0 x2 9 x x2 x2 9 0 0 3. Remember to flip the sign of the inequality when multiplying both sides by 1 . 3 of the equation y x x 1 x 3 x 3 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society x 9 x2 9 2. 0 x 1. x1 x x 3 x 3 x x 1 x 3 x 3 0 0 Make the coefficient of the highest power of x .

x 3 . or x x 9 1) x2 9 Alternative method Since the question doesn’t ask for exact values. 3 19 . Always zoom in to check places where the graph is near the x -axis. ZStandard and ZDecimal.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Alternative method from this step (Sketch the graph of y From the GC.g. the solution interval 0 x 1 is not clearly visible at some zoom levels.g. zero the inequality. then use the GC directly. 0 x 1 . where the inequality is near equality. e. e.

b. with first term a ar n 1 u1 and common ratio r : n th term. Use the various AP & GP formulae to solve.e.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Arithmetic & Geometric Progressions Below. 2. u k uk 1 is constant for all k . uk uk Sk uk 1 S k 1 . any k th term u k . b. i. Q: Show that a sequence is a GP. u3 . c form a GP. u 2 .e. given other properties of it. u 2 . u 2 . k is constant for all k . c form an AP. d (or r ). then b a c b Sum of first n terms. FAQ Q: Find various properties of an AP or GP. then For any sequence u1 . u n a n n 2a n 1 d a un 2 2 2b a c If a. AP denotes arithmetic progression and GP denotes geometric progression. where m k 0. S n Sum to infinity. A: 1. S n For a GP u1 . A: Show that the difference is common. uk 1 20 . Write down a . S a 1 r a (1 r n ) 1 r a rn 1 r 1 1) ( S exists only if r b a c b b2 ac If a. n . A: u Show that the ratio is common. Q: Show that a sequence is an AP. u3 . S k 1. with first term a n 1 d u1 and common difference d : n th term. u3 . For an AP u1 . un Sum of first n terms. i. any sum of first k terms S k which is given in the question. where k um Sm 0.

into the second equation. 3. a a 15 d d a 5 4 a2 400 2 a a 3a Now substitute the first equation. Solve manually OR use GC’s table of values/graph. Solution Comments Let the arithmetic progression have first Define all variables that will be used. 2. first n terms S n . any k th term u k . 3 9 18 2a 8d (2a 17 d ) 2 2 27a 108d 18a 153d a 5d 20 T1 T16 20 T1T16 400 400 Simplify the other statement in the question to another equation in terms of a and d . Example 1 [JJC/II/Q1] An arithmetic progression of positive terms is such that twice the sum of the first nine terms is equal to the sum of the next nine terms. Given that T1 . T1 6 forms a geometric progression. which we want to find. Simultaneous equations are a common feature in AP & GP questions. 20 . Use the various AP & GP formulae to solve. d (or r ). any sum of first k terms S k which is given in the question. 2S9 3S 9 S18 S18 S9 Always write sums of consecutive terms in terms of differences of sums of first k terms: T10 T11 T18 S18 S 9 Simplify the first sentence in the question to an equation in terms of a and d . n . a 5d . A: 1. 2. A: 1. Identify if the sequence given is an AP or GP. common difference d . Interpret term a .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Q: Model and solve a problem using an AP or GP. Form an inequality in terms of n . Q: Find the least/most n such that some condition holds. Let Tn denote the n th term of the arithmetic progression. and sum of the question in the context of the defined variables: find a and d . Write down a . find the first term and the common difference of the arithmetic progression. a 100 21 .

After how many complete years will he have. i.05 1.05 1 21 1.05 x u2 1.05 x u3 1.05 x 1. Interest is added to the account at the end of each year at a fixed rate of 5% of the amount in the account at the beginning of that year.052 1. He also decides that he will not draw any money out of the account. 1. at the end of n years. “replace” 3 with n . u1 . but just leave it. u 3 .052 1. and observe a pattern for u n . Here..053 1.05 1. when the interest for the last year has been added. or the pattern will be difficult to observe.05 2 1. he will have a total of $21 1. (i) (ii) Show that.05 x un 1. at least $12x in his account? Solution (i) Comments Let u n be the amount of money in $ in the Define u n as this to avoid writing the same phrase repeatedly.05n 1 x in his account. 1.052 . “At least” means use 1 x 1 12 x 12 ( instead of .053 etc.05 1.05n 1.05n 1 1 x (ii) un 12 x 1 x (shown) Apply the formula for the sum of the first n terms of a GP to simplify u n . Remember to man’s account at the end of n years. to build up. d 2 Example 2 [N97/I/Q15 (Part)] A bank has an account for investors. 1. and any interest.05 n 1 x 0) Manipulate the inequality to solve for n.05n 21 1.05 x 1.05n 1 1 x Evaluate successive values of u k . arithmetic progression is of positive terms) a 10 .05n 1 1. for the first time. u 2 . u1 1. Do not simplify 1. specify the units (in $). A man decides to invest $x at the beginning of one year and then a further $x at the beginning of the second and each subsequent year.e. 21 1.05n 1 12 11 1 21 7 22 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society a 10 or 10 (rejected a 0 as the Look for conditions given in the question whenever you must choose a correct value.05 x x 1.

13 etc.05 0 .1 1 12000 1.1n 2 1 1.000.05 ln 11 7 11 7 n 9.12 1.1n 100000 n 12000 1 1.e.) ln1. Xin Fu will again withdraw $12. u1 1.1n 2 1) is a sum of powers of 1.26 (3 s. where 0 x 1 . and observe a pattern for u n .1n 100000 12000 1. u 3 . Solution Comments Let u n be Xin Fu’s bank account balance in Define u n clearly. Here.Mathematics Revision Notes 2009 RI(JC) Mathematics Society n 1 ln1. Likewise. Here.. $ after his n th withdrawal.1n 1 1.12 100000 u3 1. 12000 1.1n 23 .1 1 Evaluate successive values of u k . Express Xin Fu’s bank account balance in $ after his n th withdrawal in terms of n .f.1n 1 1. etc.1 1 Apply the formula for the sum of the first n terms of a GP to simplify u n . the sign of the inequality of preserved when dividing both sides by ln1. immediately after the bank’s interest has been credited. Xin Fu deposited $100.1n 20000 6 1.12 . Switch the sign when dividing both sides by ln x . or the pattern will be difficult to observe. u1 .05 least integer value of n 10 ln When n is in the exponent. and specify the units ($).000 at the end of each subsequent year. he immediately withdraws $12.1 n 1) 120000 20000 1.1 100000 12000 u2 1.1u2 12000 1.000 with a bank that pays 10% interest per annum at the end of each year. take ln on both sides.1u1 12000 1. Here.1n 1 1. 120000(1. Example 3 [VJC06/J1CT/Q9b (Modified)] At the beginning of the year.13 100000 un 1. so it has n terms (not n 1 ). i.1 100000 1. Do not simplify 1. u 2 . 2 with n 1 . After the interest is credited.1 ranging from 0 to n 1 . “replace” 3 with n . 1. Count the number of terms carefully. (1.

especially if the result to be proven: un 20000 6 1. OR Use induction. Use repeated substitution.1u n 1 12000 Now. 24 .1n is already given. Recurrence & Induction. Refer to Example 2 in Summation.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Alternative methods Write the recurrence relation: u n 1.

H. u 2 . . … R.H. where m n . Aim: Show that P1 is true. where the common difference d is a constant. 25 .S. For constants a and b . n n o o af r r m n a r m f r n n af r r m bg ( r ) a r m f r b g r r m o Sum of AP: general term has the form a br . 1 ru n 1 . n 1 o r n n 1 2 r 1 n 1 o r2 n n 1 2n 1 6 r 1 n 1 2 1 2 o r n n 1 n n 1 r 4 2 r 1 r 1 Convert sums with r 1 to use those formulas above showing r starting from 1. AP can be expressed as u n u n 1 d . P is true. un 1 . Each term u n is defined as a function of the preceding terms u1 . functions of r f (r ) and g (r ) and m n . 3 n 2 2 n n m 1 o r m f r r 1 f r r 1 f r ur when consecutive terms cancel out. o Sum of GP: general term has the form abr . (Example 1) Use method of difference to find Recurrence For a sequence u1 . GP can be expressed as un Mathematical Induction Step 1: Base case.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Summation. u 2 .a recurrence relation defines it recursively.S. L. . where the common difference r is a constant. R. … L.S.H.H. Let Pn be the statement “…” When n 1 . Recurrence & Induction Summation n Definition: r m f r f m f m 1 f ( n) .S.

Know where you’re going when performing the inductive step: 1.S. Aim: Show that Pk Assume Pk is true for some k To prove Pk 1 1 is true whenever Pk is true.e.H. C . of Pk . Make the necessary changes if the question specifies a base value of n other than 1. Substitute appropriate values of r to find A.H. find the sum of the first n terms of the series 2 2 3 4 4 3 4 5 6 4 5 6 Find the sum to infinity of this series. B. Solution Let 2r A B r 1 r 2 r 3 r 1 r 2 for some constants A.H.S. of Pk with R. 3. Hence. Since P1 is true. 2r A r 2 r 3 B r 1 r Comments Recall the partial fractions decomposition for linear factors in the denominator (in MF15). The assumption that Pk is true for some k is the induction hypothesis.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Step 2: Inductive step.H. Convert L. Convert the expression to R.e. Example 1 [RJC08/Assignment 7/Q4 (Modified)] 2r r 1 r 2 ( r 3) Express in partial fractions.S.S. Pn is true for all n . i. i.S. of Pk 1 (proven).H.S. C r 3 3 C r 1 ( r 2) 26 . i. C . … Now. 2. by Mathematical Induction. B. .e.S. … R. of Pk 1 to involve L.H. substitute L. … is true. o The induction hypothesis must be used somewhere in the induction.H. Since we assume Pk is true. L. Check that the solution allows for cancelling out terms using method of difference in the later part: 1 4 3 0 . Pk 1 is true whenever Pk is true. of Pk .

n 1 4 1 r 2 3 4 3 5 3 6 3 r 3 r 1 r 1 4 2 3 1 4 3 4 1 4 4 5 1 n 1 n 1 1 4 3 n n 1 n 2 1 4 3 n 1 n 2 n 3 1 2 1 2 4 1 3 3 3 n 2 1 3 n 2 n 3 4 n 2 3 n 3 4 n 3 Show the cancelling out of terms clearly. 3at the top. n at the bottom. Sum to infinity r 1 1 4 3 r 3 r 1 r 2 n lim n r 1 1 1 n 2 4 3 n 3 3 r 1 r 2 r 3 lim n 1 2 The sum to infinity is the limit as n of 1 the sum to n terms. 2. think of the method of difference.Mathematics Revision Notes 2009 RI(JC) Mathematics Society When r When r When r 1. Write out the remaining terms and simplify. and r n 2. As n . If a sum cannot be found directly or you have decomposed the term into partial fractions. 4 B 3 . where r 1 r 2 (r 3) r 1. Write out the lines where r 1. leaving the fractions in their original form (do not simplify them). 2. the sum to infinity is n 3 1 . 2 2A 2. not r . 0 n 2 3 and 0 . The terms should be in terms of n . . 2 1 2 27 . n 1. Always do this for the first 3 values and last 3 values of r . Hence. 6 2C A B C 1 4 3 2r r 1 r 2 r 3 1 4 3 r 1 r 2 r 3 n Required sum r 1 2r r 1 r 2 (r 3) Observe that the terms of the given series are 2r of the form .

Solution (i) (ii) un 1. substitute k n 1 into the general form.1n (iii) Let Pn be the statement “ un 20000 6 1. the general form for un is: 1.1k un k 1. 28 .1un 1.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 2 (Refer to first paragraph of Example 3 in Arithmetic & Geometric Progressions.1 n 1 98000 98000 12000 1 1. Prove your result in part (ii) for all n 1 by mathematical induction.1n ”.1n 1.1n 1 u1 12000 1.1 1 1 1. find un in terms of n .12 1.1u n 1 12000 Comments u1 1.1n 2 1.1 ranging from 0 to n 2 .1n 1 1 1. The recurrence relation expresses un in terms of u n 1 . 1. Hence. Observe a pattern and express un in terms of u1 (which can be evaluated). To express un in terms of u1 .1un 1.1n 3 1) is a sum of powers of 1.12 1.1 n 1 1 120000 22000 1.1un 1 2 12000 12000 12000 12000 1.1n 1 120000 1. so it has n 1 terms (not n 2 ).1 1 3 1. (i) (ii) (iii) State an expression for un in terms of un 1 .1 1.1 1 2 12000(1.12 un 3 2 Use repeated substitution. Use the relation repeatedly to express un in terms of un 2 .1n 3 1 u n k u1 n k 1 k n 1.1n 20000 6 1.1 un 3 12000 12000 1.) Let un be Xin Fu’s bank account balance in $ after his n th withdrawal. (1. Here.1 100000 12000 98000 un 1.1 1 12000 1. un 3 etc.1n 120000 20000 1.1k 1 1) 1.

1 20000 6 1. u k 20000 6 1.H.S.e. u k 1 Now. of Pk .S.S. 1 20000 6 1.H.1k . L. Pk Since 1 is true whenever Pk is true.1k 120000 20000 1.1u k 12000 . of Pk 1 .S. of Pk .S.S. i. Replace n by k 1 in the statement of Pn . P is true.1k 1 1 20000 6 1.e.H. 1 . u1 98000 R.H. Assume the induction hypothesis.1k 12000 (by induction hypothesis) 132000 12000 20000 1.H.H. (proven) i. Replace n by k in the statement of Pn .1k . Use the induction hypothesis that Pk is true.1k uk 1 . 1. L.S. Pn is true for all n 29 . L.Mathematics Revision Notes 2009 RI(JC) Mathematics Society When n 1 . by Mathematical .H.S.H. of Pk 1 .11 98000 Show how the value 98000 is calculated. P1 is true. Manipulate the expression to be more and more similar to R. to substitute L. Induction. 20000 6 1.S. 1 Assume Pk is true for some k i. To prove Pk i.H. of Pk 1 to involve L. Use the given recurrence relation. R. uk 1 is true.S.1k 20000 6 1.e. of Pk with R.H.H.1u k 12000 1. to convert L.H. u k 1 1.S. Convert the expression to the R.e. R.S.

S.S. 1 1 1! L. To prove Pk k 1 1 i.H.H. Let Pn be the statement n “ r 1 r2 r 1 r! n 1 2 n! 1 ”. L.S. of the given equation as a sum. When n 1 .S. Now.S. 1 3 Assume Pk is true for some k k .S.S. L.S.H. R.S. r 1 r2 r 1 r! is true.H. k 1 k! 1 k2 3k 3 k 1! Use the induction hypothesis that Pk is true to substitute L. for easier manipulation. of Pk Pk . i.H. r 1 k r2 r 1 r ! Convert L.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 3 [N01/I/13b] Use induction to prove that 3 1! Solution 3 1! n 7 2! 13 3! n2 n 1 n! Comments n 1 2 n! 1 7 2! 13 3! n2 n 1 n! Rewrite the L. of Pk 1 to obtain the sum of k terms in L. P is true. Check that it works for the first 3 terms given.S.S.H. 1 to involve L. extract the ( k 1 )th term from the sum of ( k 1 ) terms in L. k 1 k! 1.H. n2 n 1 n! r2 r 1 r ! r 1 shows the general form.H. of Pk .H.S. of Pk with R.e.H. of Pn involves a sum of n terms.S. 12 1 1 1! R.e. 1 2 3 Show how the value 3 is calculated. Replace n by k in the statement of Pn .H. of Pk . r 1 r2 r 1 r! k 1 k 2 2 k 1! 1. 30 . 2 Assume the induction hypothesis.H. of r2 r 1 r ! r 1 k 1 k 2 k 1 1 k 1! r2 r 1 r ! r 1 2 k 2 3k 3 k 1! When L. Replace n by k 1 in the statement of Pn .H.

and k ! is the product up to k .S. 1 k 1 ! is the product of all positive integers 4 k 1! k 2 k 1! 1 R.e.Mathematics Revision Notes 2009 RI(JC) Mathematics Society k 1 k 1! 1 k 2 3 k 3 [ k 1 !] k2 4k 2 Convert the expression to the R. Hence: k 1 ! k 1 k! k 1! Since P1 is true. of Pk 1 . Manipulate the expression to be more and more similar to R. (proven) i. by Mathematical Induction.S. of Pk 1 . m m 1 m 2 k 1 ( k !) m! m m 1 m 2 k! where m k 0 .S.H.H. Pk 1 is true whenever Pk is true.H. up to ( k 1 ). Pn is true for all n k! In general: m! k 1 . k 1 31 .

dy dx f' x nx n ex a x ln a 1 Range of validity/Comments Memorize d 1 dx x d 1 x and 2 dx x 1 2 x . ln x sin x 1 x x 0 cos x sin x sec 2 x cos x tan x cosecx cot x cosec x cot x cosec 2 x 32 . Let y be a function of u . Unless otherwise stated. the derivatives are valid for all x NOT in MF15: y f x xn ex ax .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Differentiation Basic Formulae Let u and v be functions of x . d u v dx d uv dx v du dx du dx u dv dx dv dx (Product Rule) d u dx v dy dx v du dv u dx dx 2 v (Quotient Rule) dy du du dx (Chain Rule) Derivatives of Standard Functions Let n be an integer and a be a constant.

we can find by differentiating the equation term-bydx term. However. while cot x .Mathematics Revision Notes 2009 RI(JC) Mathematics Society In MF15: y f x dy dx f' x Range of validity/Comments secx sin 1 sec x tan x 1 x cos 1 x tan 1 x 1 x2 1 1 x2 1 1 x2 x x 1 1 sin 1 x 1 1 . while cosec x . with respect to x . d g y dx d dy g (y ) dy dx 33 . (Example 2) Use the Chain Rule to differentiate a function of y . say g y . if the equation cannot y f x . differentiating explicitly gives dx dx dy be easily expressed as y f ( x) . sin x sin x 1 1 . use the Chain Rule with u (Example 1) dy dx d f g x dx dy du du dx f' g x g' x g x and y f g x f ( u) : Implicit differentiation: If an equation involving x and y can be expressed as dy d f x . cos 1 x cos x cos x 1 1 . tan 1 x tan x tan x Techniques of Differentiation Let f x be any function and f ' x be its derivative given in the table above. with respect to x . To find the derivative of f g x . while sec x .

each value of the parameter t gives a point with coordinates ( x . f x is: Strictly increasing Strictly decreasing Concave upwards dy ) is: dx Only if for all a dy dx dy dx 0 0 x b: i. then First Derivative Test: To determine the nature of a stationary point at x . . g (t ) on the curve. y ) f (t ). then f x is concave Upwards. after calculating and .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Parametric differentiation: When a curve involving and is described by equations of the form x f ( t ) and y g ( t ) . with t in some given range of values.e. (Example 3) o Use the Chain Rule to find dy dx dy . and at these dx values of x . Gradient ( Positive Negative Increasing d2y dx2 d2y dx2 0 0 Concave downwards Decreasing o If for a range of values a x b . dx dt dt dy dx dy dt dt dx dy dt dx dt Applications of Differentiation Properties of f x from its Derivatives For a range of values a x b . o Use it when is difficult to find ( is difficult to differentiate) or dx2 dx2 dx dy o In the following table. ) or a part of it. evaluate dy at a x value slightly less than and a x value slightly more than . the graph of f x looks like a U ( ) or a part of it. 34 . and observe dx dy whether these values of are positive or negative. If it looks like an n ( f x is concave downwards. write the actual values of . dx d2y d2y dy 0.

Given that the line is the normal to a curve y m 1 f x at x . d2 y dx2 at x Nature of stationary point 0 Maximum turning point 0 Minimum turning point 0 No conclusion possible Tangents & Normals Q: Find the equation of a straight line. A: Given 2 points on the line with coordinates x1 . where x dx 1 . since m1 m 2 dy . y2 . evaluate . write the actual values of and at x dx2 . m y2 x2 y1 x1 f x at x Given that the line is the tangent to a curve y m dy . 35 . y1 and x2 . y1 . where x dx . m . y y1 m( x x1 ) Q: Find the gradient.Mathematics Revision Notes 2009 RI(JC) Mathematics Society x dy dx 0 0 0 0 0 0 0 0 0 0 0 0 Graph Nature of stationary point Maximum turning point Minimum turning point Stationary point of inflexion Second Derivative Test: To determine the nature of a stationary point at x d2y at x and observe whether it is positive or negative. where m1 and m2 are the gradients of 2 perpendicular lines. A: Given the gradient m and a point on the line with coordinates x1 . of a straight line. dx2 d2y o In the following table.

o Use the Chain Rule to connect rates of change. Example 1 Differentiate a x with respect to x . Minimization & Rates of Change o Always show that a stationary point you have found is indeed a maximum point (or minimum point). y -intercepts: Let x 0 in the equation of the curve and solve for y . 1 2 2 A: d x2 x1 y2 y1 Maximization. Solution ax eln a e x ln a x Comments (since k eln k for all k (since log a b r r log a b ) ) 36 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Q: Find the axis intercepts of a curve. dx Q: Find the value of x at which the tangent is vertical (parallel to the y -axis). y2 . Q: Find the value of x at which the tangent is horizontal (parallel to the x -axis). A: dy Let 0 and solve for x . o Make sure that there is only 1 variable in the expression before differentiating the expression with respect to that same variable. d . y and x 2 . Let dy dx Q: Find the distance. between 2 points with coordinates x1 . using either Derivative Test to determine its nature. A: dx dy 1 0 and solve for x . A: x -intercepts: Let y 0 in the equation of the curve and solve for x . o Remember to include units in your final answer. o Similar triangles and trigonometric ratios are useful in forming expressions.

3y2 dy dx x2 dy dx 2 xy 6 x 2 0 3y2 3y2 dy dx dy 2 xy 6 x 2 0 dx dy x2 2 xy 6 x 2 dx 2 x y 3x 2 xy 6 x 2 2 2 3y x 3 y 2 x2 x2 dy to one side and factorize dx 37 . Only write: d x a dx d xln a e dx e x ln a ln a a x ln a By the Chain Rule.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Let u d x a dx x ln a . Using the Chain Rule. dy dx The details are unnecessary for actual working. dy dy du dx du dx d u d e x ln a du dx eu ln a e x ln a ln a a x ln a Substitute u back with x ln a . Comments Whenever an equation involving x and y cannot be easily expressed as y f x . d d d yx 2 x 2 y y x2 dx dx dx dy x2 2 xy dx Group the terms containing it out. dy by factoring dx Differentiating with respect to x . d 3 d 3 dy dy y y 32 y dx dy dx dx Using the Product Rule. Example 2 [NJC06/Midyear] The equation of a curve C is given by y 3 dy Find in terms of x and y . then shift further. dx Solution y3 yx 2 2 x 3 k yx 2 2 x 3 k . where k is a constant. use implicit differentiation. y d xln a e dx eu .

dt dx 1 dx dt d et 2 dt e2t 1 1 dx dt 2 e 2t 2 e2t 1 et et 2e2t e 2t 1 4 e 2t 2 1 1 3 e 2 t 1 et 4et e 2t 1 e 2t 1 3 38 . 4 d2y dy 0 for 0 t 1 . dt dt dy dx dy dt t 2e 2t e 1 dx dt Use the Chain Rule. Find dy dx and first. d2y dx 2 d 2e t dx e 2t 1 dt dx d2y dx 2 d dy dx dx d 2 et dt e 2t 1 Use the Chain Rule to differentiate the expression involving t with respect to t .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 3 [SRJC08/Prelim/I/9i] The parametric equations of a curve are given by 1 2t x e 2 t and y et . use parametric differentiation. Find and show that dx2 dx Solution dy dt dx dt et 1 2t e 2 4 1 2 1 2t ( e 1) 2 Comments Whenever equations for x and y are given in terms of a parameter t . where 0 t 1 .

and any relations between the variables. of (i) (ii) the radius. ex 1 . dA 8 r 24 dr dA . 3 e 2t 1 hence the whole fraction must be positive. dr dt 1 4 r dr dt . 1 2 0 and 1 0. t e 1 and e2t 1 . x Since et e2t e2t d2y dx 2 1 0. then substitute r 3. 4 0 (shown) Show that the factors in the numerator and et e 2 t 1 denominator of are all positive. 4 3 r . its surface area is increasing at the rate of 2 m2 s-1. at the instant when its radius is 3 m. Substitute r 3 only in Alternatively. and not in dx A 4 r 2 . Try to find an inequality satisfied by terms in d2y d2y 0 for 0 t 1 . 39 . Let r be the radius of the balloon in m. Example 4 [N1998/I/7] A spherical balloon is being inflated and. Write down all given rates of change. the volume. with their units. For all x . Find the rate of increase. find the expression for first.] 3 [The formulae for the surface area and volume of a sphere are A 4 r 2 and V Solution (i) Let A be the surface area of the balloon in m2 . since r is a variable which takes a specific value of 3 in the question.Mathematics Revision Notes 2009 RI(JC) Mathematics Society For all t . 0 t 1. A 4 r2 When r 3 . Use to prove that dx 2 dx2 the given range of values of t . et e 2 t 1 e 2t 1 3 0 . not a constant. at the same instant. dA 2 dt Comments Define all variables that will be used.

the maximum value of A is obtained when the height of the cylinder is equal to its diameter. Hence. since r dt and t are already in terms of m and s respectively.Mathematics Revision Notes 2009 RI(JC) Mathematics Society dr dt dr dA dA dt dA 1 dr 1 2 24 1 12 dA dt Identify what you need to find and what is dr dA dA given: given and . where A is the curved surface area of the cylinder. Show that A2 16 2 x2 R2 x 2 . as x varies. 40 . the volume is increasing at 72 m3 s-1. 4 3 V r 3 When r 3 . dV 4 r 2 36 dr dV dV dr 36 2 72 dt dr dt Hence. Example 5 [SRJC08/Prelim/I/6] The diagram below shows the cross-section of a cylinder of radius x that is inscribed in a sphere of fixed internal radius R . Connect dt dt dr them using the Chain Rule. You can think of the dA as being “cancelled out”. the radius is 1 increasing at m s-1. Prove that. 12 Remember to include units in your final dr answer. (ii) Let V be the volume of the balloon in m3. Do not write units in .

Introducing rightangled triangles is useful because it allows the use of Pythagoras’ Theorem and trigonometric ratios. x 2 dA dx 4 4 R2 R2 x2 x2 x 2x 2 R2 x2 R 2 x2 d 2 R dx 0. so find dA the appropriate derivative . R 2R R h 2x Let h be the height of the cylinder. and are not already defined in the question. 41 . The curved surface area of a cylinder is its circumference height 2 rh 2 xh .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution Comments Draw a line across the rectangular crosssection of the cylinder. dx Since R is a constant. Since the expression to be proven is in terms of R and x only. 2 2 2R h2 2 x h 2 R2 x2 A 2 xh 2 x 2 R2 x2 A2 4 x R2 x2 16 2 x 2 R 2 x 2 4 d x R2 dx x2 (shown) Use the Product Rule. express h in terms of R and x first. Define all variables that will be used. The question states A is to be maximized as x varies. By Pythagoras’ Theorem.

such as constants Factorize and f x 2 . then solve for x . choose some R e. dx so use the First Derivative Test instead. x R here ( so R 2 x 2 0 . This allows cancelling out of terms when dA is equated to 0 . Here. To carry out the test. x 0) x dA dx R 2 R 2 0 R 2 0 0 Always verify the nature of the stationary dA point. e. Since there is only 1 value of x which gives a stationary point. dx 2x .g. f x . R 1 . R 2 x2 2 2x . When x h 2 R2 R . it must be the “maximum value” to be proven. x R h 0 ). 2 0. Hence.Mathematics Revision Notes 2009 RI(JC) Mathematics Society 4 R2 x2 dA When 0. 2 2 x2 2x x2 x 0 and h 0 ) (proven) 2x or 2x (rejected h 2x diameter of cylinder 42 . dx x x x 2 R2 2 R or 2 R 2 R (rejected 2 To prove A is maximum when h dA first let 0 . is difficult to differentiate. where f x 0 for R 2 2x 2 0 all x . x R maximizes A . then substitute for R in the equation h 2 R2 x2 . dx R2 2x2 dA by removing terms which dx you know are non-zero. Express R in terms of x .g.

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Integration Basic Formulae Let u and v be functions of x . Unless otherwise stated. If f x is a function which is defined and continuous for all a f x dx F x c x b and then the definite integral of f x from a to b with respect to x is: b a f x dx F x b a F b F a Integration of Standard Functions Let n be an integer and a be a constant. Differentiation reverses integration. Remember to include the arbitrary constant c for indefinite integrals. u v dx udx vdx d f x dx dx d f x dx dx f x f x c Integration reverses differentiation. the derivatives below hold for all x . NOT in MF15: f x f x dx Range of validity f x f x dx xn 1 x xn 1 n 1 ln x n 1 cosx sin x sin x cos x tan x ex a x ex ax ln a sec2 x 43 .

Mathematics Revision Notes 2009 RI(JC) Mathematics Society In MF15:
f x

f x dx

Range of validity

f x

f x dx
ln sec x ln sin x ln cosec x ln sec x cotx tan x

1 x
2

a

2

1 tan a sin
1

1

x a x a
x a

tan x cot x

Range of validity 1 x 2

1 a 2 x2
1 x a
2

0 0
x

x x
1 2

a 1 x

2

1 x a ln 2a x a 1 a x ln 2a a x

x a
x a

cosecx
tan x

2

2

If

f ( x ) dx involves a logarithm ( ln g ( x ) , where g ( x ) is a function of x ), place a

modulus around the expression within the logarithm, i.e. write ln g ( x) . This removes the need to specify the range of validity. e.g. cot x
ln sin x c

Techniques of Integration Let f x be any function and F x be its integral given in the table above. To find the integral of f g x
g ' x , where g x is a function of x , use the fact that by the

Chain Rule from differentiation, Hence,

d F g x dx

f g x

g' x .

f g x

g ' x dx

F g x

c

Trigonometric Formulae: Used to integrate expressions with trigonometric functions. NOT in MF15: In MF15:

sin2 x cos2 x 1 tan2 x 1 sec2 x cot 2 x 1 cosec2 x

sin 2 x 2sin x cos x cos 2x cos 2 x sin 2 x

2cos2 x 1 1 2sin 2 x
(and others) 44

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Partial Fractions: Used to integrate expressions with an algebraic fraction, which has a denominator that can be factorized (or is already factorized). Splitting the fraction into f' x partial fractions can help integrate it, since the partial fractions will be of the form . f x Partial fraction decomposition formulae are given in MF15. Integration by Parts: Used to integrate expressions with a product of 2 distinct functions.
u dv dx uv dx v du dx dx

dv is easier to integrate: dx LIATE: Logarithmic, Inverse trigonometric, Algebraic, Trigonometric, Exponential. dv f ( x) . o To find f ( x ) dx when you can find xf '( x ) dx , let u 1 and dx o Use algebraic manipulation if f ( x ) dx becomes the only integral on both sides.

Generally, choose u in the following order such that

Integration by Substitution: If a substitution is required, it will be given in the question. To find f ( x ) dx with the substitution x g u :

f x dx
For definite integrals,
x2 x1

f g u

dx du du

f ( x ) dx

u2 u1

f g u

dx du du

where x 1

g u 1 and x 2

g u2 .

o For indefinite integrals, express your final answer in terms of x , not u .

45

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Applications of Integration
y

y

f (x )

If f x 0 for all a x b , and R is the area between the curve and the x -axis from x a to x b , The area of R can be approximated by the sum of areas of n rectangles, as shown.

a

0

b

x

The exact area of R is the limit of this sum of areas as n , i.e. the definite integral:
b a

f x dx

y

R

y

f (x )

If f x

g x

0 for all a

x b,

Area of R :

y g (x)

b a

f x

g x dx

Volume of solid formed when R is revolved completely about the x -axis:

a

0

b

x

b a

f x

2

g x

2

dx

y

If f y

g y

0 for all a

x b,

a
Area of R :

0
R

x

b a

f y

g y

dx

b

Volume of solid formed when R is revolved completely about the y -axis:
b a

f y

2

g y

2

dy

x

g ( y)

x

f ( y)

46

Mathematics Revision Notes 2009 RI(JC) Mathematics Society y R x f t . Area of R : c a f x dx b c f x dx Or: b a f x dx y y R f x If y f x and y g x intersect each other within the limits of integration. If f x f x 0 for all a 0 for all c x c and x b. split the area between the 2 curves into where f x g x and where g x f x. and y g t 0 for all a x b. split the area between the curve and the x axis into its positive and negative parts. c b x Area of R : c a g x f x dx b c f x g x dx Or: b a f x g ( x) dx 47 . a 0 b x where f p a and f q y R y f x a 0 x c b If y f x has both positive and negative components within the limits of integration. y g t If a function is defined in parametric form as x f t . y g t . Area of R : q p g t dx dt dt b. If g x f x f g x x 0 y g x a 0 for all a x c and 0 for all c x b.

required. k 0 . and a b . as shown. here 2 2 . X . Either extract to get x a 9 1 the x 2 term. and volume: k2 b a Use the GC to check your answer. For parametric functions. 9 48 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society y R If y k for some k . then key in fnInt(<expression>. <lower limit>. apply the formula in MF15. 1 9 x2 e 1 (ii) Find. height b a .1. or extract (since 3 d 3x 3) to get: dx 1 13 3 dx 3 0 12 3 x 2 Then. X . where n 1. in terms of n and e . Example 1 [N08/I/5] (i) Find the exact value of 1 3 0 1 dx . use [MODE] PAR and key in T using [ X . <upper limit>) e. Check your answer using GC. T . to find 1 0 x 2 dx .g. 0. Solution x n ln xdx . key in fnInt X 2 . Use the GC to sketch graphs of functions first before finding the areas or volumes Θ. or find an integral if an exact answer is not required. o [MATH] [9]. y k Area of R : k b a a 0 b x The solid formed when R is revolved completely about the x -axis is a cylinder with radius k . n ]. Comments (i) 0 1 3 1 dx 1 9 x2 1 9 1 3 0 1 x2 1 3 2 dx 1 1 1 3tan 3 x 3 0 9 1 tan 1 3 tan 1 0 3 1 3 3 Recognize when the integral is of a form given 1 1 in MF15.

xn 1 . Check your answer using GC by choosing some n. so treat it as a constant when 1 n 1 integrating.v x Here. for x 0 . and v . ln x is Logarithmic and x n is Algebraic.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (ii) Let u lnx . n 1 . e. dx dx xn dx n is a constant. both curves and y 1 . x ln xdx x n 1 n x n 1 ln x n 1 e 1 ln x n 1 1 n 1 e 1 x n dx 1 e x n 1 ln x n 1 en 1 n 1 en xn 1 2 1 For definite integrals. Write out u . n 1 e e du Then dx e 1 1 . n 2 . dv du . . remember to keep the limits on every term (integrals and results of integrals). so by LIATE. x2 and y 2x2 1 x 4 . R is the region 49 .g. 2 n 1 1 2 0 1 n 1 n 1 en 1 n 1 en 1 en n 1 n 1 1 2 1 n 1 1 n 1 2 2 en 1 ne n 1 1 2 n 1 Example 2 [HCI08/Prelim/I/10] The diagram below shows the graphs of y bounded by the y -axis. dv dx xn . choose u ln x .

. 1 2 sin 2 4 sin cos d 0 cos 2 1 4 sin sin cos d 2 0 cos 1 4 2 sin d 2 0 50 . find the exact value of 2 1 y 4 dy 0 1 2y 1 1 . sin 4 where y1 g 1 and y2 dy d d g 2 . change 4 things from the original integral: 1. The function to be integrated. Find 1. 1 4 0 y dy 1 2y 4 0 1 2 sin 2 1 sin 2 dy d d Write out the original integral. and 3. not y . Integrate with respect to . d 4. and make all 4 changes above at the same time. giving your answer correct to 3 decimal places.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (i) By substituting y 1 2 sin . dy 3. 2 (rejected) 2 4 For a definite integral. Solution Comments radians about (i) dy d 1 2sin 2 sin cos cos To find f ( y )dx with the substitution y g 0 2 2 . 2. Simplify the substituted integral. sin 0 f y dy 1 2 f 1 . Multiply by . recall the general formula: y2 y1 When y When y or 0 . Hence or otherwise find the 2 4 (ii) Verify that the two curves intersect at the point exact area of region R . (iii) Find the volume of the solid generated when R is rotated through 2 the y -axis. Limits of integration. first in your working.

(NOT show) that at . Hence. 2 4 1 4 1 x2 1 2 2 1 2x 1 4 2 1 2 Hence. just show that the point lies on both curves. at the point (verified) 2 4 x 4 1 2 The question only requires you to verify 1 1 . use the doubleangle formula for cos 2x to convert them. 1 2 2 1 sin 2 2 2 4 4 1 2 0 0 2( 2) units2 16 cos 2 x 1 2sin 2 x 2cos 2 x 1 1 sin 2 x (1 cos 2 x) 2 1 cos2 x 1 cos 2x 2 (ii) When x 1 . the curves 2 4 intersect.Mathematics Revision Notes 2009 RI(JC) Mathematics Society 1 2 2 4 0 1 cos 2 d 4 0 To integrate sin 2 x or cos 2 x . the curves intersect 1 1 . you must use the equation x2 x4 and solve for x . 2 x2 1 y x 4 x 2 y 1 4 y x 2 x2 1 2 yx 2 y R x2 0 2y 1 x 2 x or y 1 A y 1 2y y (rejected 1 2y 1 4 x 0) 1 2 1 51 . . If the question requires us to find all intersection points.

For each curve. then use MF15. A is a rectangle with area 1 . 0 1 2y Alternative method 1 4 1 4 2 4 1 5 4 4 1 5 4 2 2 16 2 16 2 4 1 5 5 4 5 1 1 1 Let A be the region bounded by the axes. R is 4 y bounded by the y -axis and x . x 1 and y 1 . a x The alternative method integrates with respect to x . hence integrate with the limits from the x -axis. as shown by the dotted rectangle in the graph above. from part (i).Mathematics Revision Notes 2009 RI(JC) Mathematics Society Area of R 1 4 0 y dy 1 2y 1 1 4 1 y 4 dy Look at the graph above to decide what to 1 integrate. 52 . Area of region R Area of region A Area of region outside R and inside A 1 1 x2 4 2 1 dx 1 x dx 2 0 2x 1 2 44 2 4 2 2 5 16 8 4 2 9 2 units2 5 16 40 5 2 2 10 x2 Use long division on to get an 2 x2 1 1 integral of the form 2 2 . then integrate with the limits from the y -axis. R is bounded by the y 4 1 4 -axis and x y . 1 y 4 dy . The method shown here integrates with respect to y . 1 2y From y 1 to y 1 . split the integral into these 2 regions. Hence. From y 0 to y . Each curve is already expressed as y in terms of x . express x in terms of y . 2 16 2 16 2 16 2 4 5 y4 5 4 5 4 5 1 1 4 5 4 We know the value of the first integral.

then express x 2 in terms of y for each region. 53 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iii) Volume required 1 1 y dy 1 y 2 dy 1 2y 4 3 1.p. suggesting that you find the volume directly using GC. Again.984 units (3 d. The question states “3 decimal places” but not “exact”. split the integral into 2 regions.) 1 4 0 Volume of revolution about y -axis is given by x 2 dy .

There are 2 arbitrary constants in the general solution of this form. DE denotes Differential Equation. 2. 1. Rewrite the general solution with the known values for the arbitrary constants. 54 . Substitute the initial conditions into the general solution. You can imagine the dx ’s cancelling out: 1 f y dy dx dx 1 dy f y 1 f y dy x d2y dx2 f x Integrate both sides with respect to x . dy dx dy dx y f ( x)dx Say f ( x ) dx g x A. Divide both sides by f y . given some initial conditions. g x A where A is an arbitrary constant. y h x Ax B Particular Solutions To find a “particular solution” to a DE. Forms of DEs and General Solutions Convert any given DE into one of these 3 forms. Form dy f x dx General Solution Integrate both sides with respect to x . g x g ( x ) dx A dx Ax Say g ( x ) dx h x B. y 1 f y f x dx dy 1 dx dy dx f y Integrate both sides with respect to x . Find the value of the arbitrary constants.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Differential Equations Below. Integrate both sides again with respect to x . y h x B Ax where B is an arbitrary constant. 3.

4. e. Write: If dx dx is inversely proportional to something.g. dt dt 3. Rate of change Rate of increase Rate of decrease. etc. v and x . ensure you include these 3 cases: c 0 . which indicates a rate of decrease. aim to express the DE in terms of v and x only (without y ). Identify the rate of change. and are not already defined in the question. c 0 . Then the L. express y and dy dv in terms of . focus on forming the DE correctly as a basis for your answer. y and a substitution v g x . dx dx 3. then dt dt dx k . which indicates proportionality. 2. dx where f ( x. indicating intercepts. Modelling In modelling questions. Convert the substituted DE into the form required or one of the 3 standard forms above. Write: x 2t Look out for key phrases in the question such as “water flows out at a rate of …”.Mathematics Revision Notes 2009 RI(JC) Mathematics Society To draw a family of solution curves. dt x 2t 1 . then dt dt dx kx2t . y is given to solve this DE. where k is a positive constant. dx dx 2. If a DE is given as 1. x2t . Substitute at least 3 distinct values for the arbitrary constant c . dt x2t . 55 . Draw the graphs using GC. where Rate of increase and Rate of decrease are both positive.g. x2t . asymptotes. such that your final answer is in terms of y and x only. y ) and g x. turning points. Solve the DE. and c 0 . Express v in terms of x and y . say . where k is a positive constant. y are functions of x and y . Define all variables that will be used.H. In the original DE. of the DE will be . Solving DEs by Substitution dy f x. or “is proportional to”. 1. v and x . Differentiate the given substitution with respect to x to get 2. If dx dx is proportional to something. dx dx dy dv in terms of . Generally. 1. 5. e.S.

3 c y ln(x 2 1) 2 3 2 ln1 2 2 Write this line for proper presentation. The DE is of the form where c is an arbitrary constant. on a single diagram. 1. together with 2 other members of the family of solution curves. The gradient of every solution curve is dy 3x . 2 1 2x dy f x . Rewrite the general solution with the known values for the arbitrary constants. 2 dx x 1 because the denominator x2 1 will “grow dy 0. . Solution Comments 3 2x 2 x2 1 (i) dy dx 3x x 2 y 3 dx 2 x2 1 3 ln( x2 1) c. x 0 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 1 [N08/I/4] (i) Find the general solution of the differential equation dy 3x 2 dx x 1 Find the particular solution of the differential equation for which y What can you say about the gradient of every solution curve as x (ii) (iii) (iv) 2 when x 0 . faster” than the numerator 3 x . dx OR Use the GC to plot the graph of y Observe the asymptote y (iii) As x . Remember the arbitrary constant c . 0 as x 56 . y 3 ln(x 2 1) 2 2 3. As x . ? Sketch. from the DE. 3x x 2 1 . the graph of the solution found in part (ii). Find the value of the arbitrary constants. Substitute the initial conditions into the general solution. the gradient of every solution curve tends to 0 . (ii) When y 2 . 2. so just dx integrate both sides.

The gradient is 0 at x 0 . the curve looks like a “V” with sharp tangents at x 0 . Differentiate the given substitution with dy du respect to x to get in terms of .g. Solution y ux 2 dy u 2x dx Comments 1. x 2 0 y 2 3 ln( x 2 1) . c y 2 Choose 2 other solution curves which are representative of all values of c . Since c 2 is a required curve. Example 2 [NYJC08/Prelim/II/1 (Part)] Show that the differential equation x2 dy 2xy 3 0 dx may be reduced by means of the substitution y ux 2 to du 3 dx x4 Hence. since 3x 2 dy 0 . e. The graph has no horizontal asymptote. du 2 x dx du 2ux x 2 dx 57 . u dx dx and x . choose c 0 and 2. ZStandard. find the general solution for y in terms of x . or otherwise. some c 0 . c 2 When the general solution is of the form y f x c . At some zoom levels on 2 dx x 1 the GC.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iv) y 3 ln x 2 1 2 2. even though from part (iii). here c Use the GC to draw the graphs of the solution curves. the gradient of every solution curve tends to 0 as x . c 0 2 3 y ln( x 2 1) 2 . the solution curves are translations of each other along the y -axis.

4. Remember the arbitrary constant c . so just dx integrate both sides. express y and terms of dy in dx x4 du 2ux3 2ux3 3 dx du 3 (shown) dx x4 du dx du . 58 . Form a differential equation connecting x and t and show that x where A and k are constants. find the time taken for the amount of drug in the patient to fall from 80 mg to 20 mg. u 1 c x3 u c y x2 y ux 2 y x2 y 1 x3 1 cx 2 x Example 3 [HCI08/Prelim/I/7] A hospital patient is receiving a certain drug through a drip at a constant rate of 50 mg per hour. The DE is of the form Write this line for proper presentation. such that your final answer is in terms of y and x only. where x (in mg) is the amount of drug in the patient at time t (in hours). Assuming that the rate of loss remains the same. if initially there is no trace of this drug in the patient’s body. Express u in terms of x and y .Mathematics Revision Notes 2009 RI(JC) Mathematics Society x 2 2ux x 2 du dx 2 x ux 2 3 0 2. u and x . dy f x . Solve the DE. 3 where c is an arbitrary constant. When there is 80 mg of the drug in the patient’s body. The rate of loss of the drug from the patient’s body is proportional to x . In the original DE. dx 3. the drip is disconnected. Convert the substituted DE into the form required. Given that k (i) (ii) 1 50 Ae k kt . 20 Find the time needed for the amount of drug in the patient to reach 200 mg. 3 x4 3 u dx 3 x 4dx 4 x x3 3 c. 1 .

so divide dx both sides by f ( y) . “no trace” means x 0 . e kc is an arbitrary constant. dt where k is a positive constant. where k is a positive constant. Substitute the initial conditions into the general solution. Find the value of the arbitrary constant. dy f y . to the arbitrary constant A x 20 50 Ae t 20 When t 0 . Rate of change Rate of increase Rate of decrease dx 1 50 kx dt 1 dx 1dt 50 kx 1 k dx 1dt k 50 kx 1 dx 50 kx. 0 20 50 Ae 0 1 . A 50 t 20 2. Rewrite the general solution with the known value for the arbitrary constant. ln 50 kx kt kc Evaluate both integrals and introduce the arbitrary constant in 1 step.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution Comments Rate of increase is constant at 50 . Remember the arbitrary constant c . Interpret 20 the question: “Initially” means t 0 . The DE is of the form 1 ln 50 kx t c. Use the information given: k 1. k where c is an arbitrary constant. x 20 50 50e 59 . Now. x 0 . so write it as kx . 50 kx 50 kx e e kt kc kc e kt Ae kt . where A kx 50 Ae kt 1 x 50 Ae kt (shown) k (i) The modulus is eliminated by transferring it e kc . Rate of decrease is proportional to x . then integrate both sides. 3.

there is 80 mg of drug. where B is an arbitrary constant. so t 0 and x 80 . rewrite the DE and solve it. x B 20ln 80 80 . Remember the arbitrary constant.) (ii) dx 1 x dt 20 20 dx 1 x dt 1 20 dx 1dt x “Rate of loss remains the same”. but since c is already used.7 hours (3 s. use another letter. hence the rate of increase is 0.46 hours (3 s. t 20ln80 20ln 20 20ln 4 27. t 20 200 1000 1 e 1 e e t 20 t 20 1 5 4 5 t 20 t 4 5 4 20 ln 5 ln 4. Find t when x 20 . 20ln x t 20ln80 Interpret the question: Initially. When t 0 . Interpret the question: Find t when x 200 . but the drip is disconnected. 20ln x t B .f.f. With the given value of k . say B .Mathematics Revision Notes 2009 RI(JC) Mathematics Society When x 200 . Evaluate both integrals and introduce the arbitrary constant in 1 step.) 60 . When x 20 .

When x 0 . 4. 2 k dx dx 3. i. up to the term in x k . A: 1.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Power Series If a function f x can be expanded as a power series: f ( x) where f (n) f (0) xf '(0) x2 f ''(0) 2! xn f n! ( n) (0) (In MF15) d n f (x ) . sin x and cos x for all x ln(1 x) for 1 x 1 (Note the range of valid x values) Maclaurin’s series are infinite series.01 into both the original expression and the Maclaurin’s series. evaluated at x 0 . f '(0) . up to f k 0 . n 0 f ( x) f (0) dy f '(0) dx dk f x d2y f '' 0 . e. into the formula for Maclaurin’s series. Implicit differentiation may be helpful (or necessary if the question states “By further differentiation of this result”). dk f x Find equations connecting all derivatives up to the k th derivative . up to x k . f ''(0) etc. Write out each term in ascending powers of x . The Maclaurin’s series of the following functions are also in MF15: n o 1 x for x 1 (This is the binomial expansion for rational index) o o e x . x 0. (0) is the n th derivative of f ( x) . Differentiate the expression repeatedly. Substitute f (0) . dx k 2.e. etc. x may be replaced with any function of x . The n th term in the Maclaurin’s series is xn f n! FAQ Q: Find the Maclaurin’s series for f x . Check your answer by substitute a small value of x . Evaluate f x and its derivatives at x 0 using the equations found. In the formula above. a type of power series. 61 .g. dxn The expression on the RHS is known as the Maclaurin’s series.

i.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Small Angle Approximations Using the Maclaurin’s series of sin x . Instead. sin x 4 sin x cos x 2 2 4 1 cos x sin x 2 2 4 2 2 2 1 x 2 x2 2 Example 1 [HCI08/Promo/12] Given that y x tan 1 x . prove that 1 x2 d2y dy 2x 2y 2 0 2 dx dx By repeated differentiation. To approximate e. use the addition formula for sin . is not small. do NOT write sin x x sin kx 4 4 4 . x3 Comments Differentiate y to get Solution y x tan 1 x dy x tan 1 x dx 1 x2 dy . The approximation is accurate when x is small.e. Since x is small. sin x since x where x is small. the 4 approximations for sin x and cos x are valid. cos x and tan x up to the term in x 2 . show that the first two non-zero terms of the Maclaurin’s series 1 4 x . for y are x2 3 Hence evaluate lim x 0 x tan 1 x .g. kx (kx) 2 cos kx 1 2 tan kx kx where k is a constant. x is close to 0 . dx 62 .

4 dx dx dx dx 2 x x4 y 0 0x 2 0x 3 8 2! 4! 0. y x2 1 4 x 3 (shown) x tan 1 x x3 x2 1 4 x 3 x4 x2 x 2 x tan 1 x x4 1 4 x 3 x4 Convert the expression to contain x tan 1 x . etc. and remember: d dy y dx dx d dy d2 y dx dx dx2 1 x2 1 x2 d d2y dx dx 2 d3y .Mathematics Revision Notes 2009 RI(JC) Mathematics Society dy 1 x2 tan 1 x x dx 2 dy 2 d y 1 x 2x 1 2 x tan 1 x 1 2 dx dx 2 d y dy 1 x2 2x 2y 2 0 (proven) 2 dx dx 1 x2 1 x2 1 x2 d3y d2y d2 y dy dy 2x 2 2x 2 2 2 0 3 dx dx dx dx dx d3y d2y 4x 2 0 dx3 dx d4y d3y d3 y d2 y 2x 3 4x 3 4 2 0 dx 4 dx dx dx 4 3 2 d y d y d y 6x 3 4 2 0 4 dx dx dx Since the expression to be proven contains d2y 1 x2 . Use the Product Rule and Chain Rule. 0. dy d2y d3y d4y y 0. manipulate the equation to dx 2 dy contain 1 x 2 . lim x 0 x tan x x3 1 1 4 x lim 3 4 x 0 x 1 3 As x tends to 0 . dx4 Evaluate y and its derivatives at x using the equations found above. terms get 1 smaller faster than the x 4 term. x6 etc. dx3 Since the question requires the series up to x 4 . 3 63 . the x5 . thus the 3 1 4 x term dominates the numerator. then use implicit dx differentiation. 2 2. find equations connecting all d4y derivatives up to . Substitute the series for x tan 1 x which has been found into the expression. When x 0 . 3 0. 8 Substitute the resulting values into the formula for Maclaurin’s series.

y 2 3 15 2 19 3 x x x 4 64 512 64 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 2 [NYJC08/Prelim/II/3 (Part)] A curve C is defined by the equation 2 y dy dx y 2 1 and (0. Solution Comments y2 1 2 (ii) (i) 2y dy dx d2y dy 2y 2 2 dx dx d2y y 2 dx y d y dx 3 3 2y y 2 dy dx dy dx dy dx 2 dy dx 2 2 Differentiate the equation of C repeatedly. 0. dy 3 y 2. Show that at x 0 . (i) d 3 y 57 . 2) is a point on C . find equations d3y connecting all derivatives up to . dx 4 4 d2y dx 2 2 3 4 2 4 15 32 3 4 d2y dx 2 3 4 15 32 2 d3y 3 2 3 3 dx 4 15 2 32 d3y dx3 y 57 256 2 (shown) 57 x 3 256 3! 3 15 x 2 x 4 32 2! Substitute the resulting values into the formula for Maclaurin’s series. dx3 Since differentiation here is with respect to x . Since the question requires the series up to x 3 . Hence find the series expansion of e y . dx3 256 Find the Maclaurin’s series of y up to and including the term in x 3 . When x 0 . up to and including the term in x 2 . by the Product Rule and Chain Rule: 2 d dy d2 y dy y y dx dx dx2 dx 2 d y dx 2 dy dx d y dx 2 2 d2y y 2 dx d3y y 3 dx dy 3 dx dy dx d 2y dx 2 d 2y y 2 dx dy dx Evaluate y and its derivatives at x using the equations found above.

differentiate both sides of dx 1 2 1 3 y x x x 2 2 Since –x . 65 . dy e x sec2 e x 1 From part (i): dx 1 2 1 3 x x From part (ii): y x 2 2 Solution (iii) Comments dy e x sec2 e x 1 dx d 1 2 1 3 x x x dx 2 2 3 2 1 x x 2 x 2 x sec2 (e x 1) dy looks similar to . up to and including the term in x 3 . sec2 e x 1 Deduce the first three terms of the Maclaurin’s series for . (iii) d 2 y dy 1 2ex y . substitute x with 1 x 3 2 x 2 3 2 1 x x 2 Questions on binomial expansion or power series commonly include differentiation and integration concepts.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (ii) ey 2 e 2 3 15 2 x x 4 64 3 e2 e 4 x 15 2 x 64 Substitute the series for y which has 2 been found into the expression. To obtain the series ex dx dy of . 2 dx dx Find Maclaurin’s series for y in ascending powers of x . e 2 must be extracted. Hence. 3 15 2 e 1 x x 4 64 e2 1 3 15 2 x x 4 64 1 3 15 2 x x 2! 4 64 1 3 2 4 2 ey 1 y 1 2 1 3 y y 2! 3! x2 e2 3 2 15 2 2 ex ex 4 64 will have infinitely many constant terms. since each y contains a 2 . Otherwise. ex We will focus on part (iii) only. e sec e 1 sec2 e x 1 e x sec2 (e x 1) has an e ex x term. Example 3 [YJC08/Prelim/I/4iii] It is given that y (i) (ii) Prove that tan ex 1 . this series cannot be simplified.

including only terms up to x k . Extract a n to convert the expression such that it contains 1 as a term. For x 1 x 1 x 1 x 1 1 2 1. where a and b are constants and x c is a positive power of x . 3. For the expansion to be valid. and x 1 . A1: 1. Simplify the coefficients. Use the above “shortcut” expansions if possible to save time. x 1. Use the GC to check your answer by substituting some within the range of validity. Use the binomial expansion for rational index.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Binomial Expansion Binomial Expansion for Rational Index 1 x n n 1 2 n n 1 n r 1 r (In MF15) x x 2! r! is either a non-integer rational number or a negative integer. a bx c n bx c a 1 a n n 2. Knowing “shortcut” expansions (NOT in MF15) can be convenient. 1 x x2 1 x x2 1 2 x 3 x2 x3 x3 4 x3 xr 1 xr r 1 xr r FAQ Q1: Expand a bx c n in ascending powers of x up to the term in x k . Write out the entire expression before simplifying to reduce careless mistakes. The actual value and the value given by the expansion should be approximately equal. 66 . n 1 nx where The expansion is an infinite series.

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Q2: State the range of values of x for which the expansion of a A2: bx c n is valid. a bx c n bx c n a 1 bx c n Q6: Find the r th term (or its coefficient) in the expansion of an expression. Multiply (or add) the expansions for each term together. 2. For each factor. Manipulate the equation to obtain an approximation for b . Express the expression as a product (or sum) of terms of the form a using partial fractions if required. A7: 1. find an approximation for b . Simplify the inequality to give a range of values of x. 3. Substitute x a into both the original expression and the expansion. It should be of the form [1 f ( x )]n . 1 f ( x) 1 3. The set of values of x for which the entire expression is valid is then the intersection of the sets of values of x for each term. follow the 3 steps in Answer (1) above. Q4: State the range of values of x for which the expansion of an expression is valid. find the range of values of x for which the expansion of each term of the form a bx c n bx c n . Q5: Expand a A5: bx c n in descending powers of x up to the term in x k . 3. A6: Use the fact that the r 1 th term in the expansion is n n 1 n r 1 r x r! Q7: By putting x a . 2. i. the range of validity is f ( x ) 1 . For the expansion of [1 f ( x )]n . A4: 1. 1. A3: 1. is valid. 4. where a and b are constants. Ensure a is within the range of values of x such that the expansion is valid. but for Step 1: Extract bx c n instead of a n to convert each factor such that it contains 1 as a term. Discard any terms which have a power of x greater than k. 67 . 2. Look at the expression immediately before the binomial expansion was used. Q3: Expand an expression in ascending powers of x up to the term in x k . Same as Answer (1) above. Using Answer (2) above. 2.e.

Use the binomial expansion for rational index. including only terms up to x 2 . since the expansion should be up to x 2 . Extract a n to convert the expression such that it contains as a term... 1 1 x 1 2 2 1 x 1 x2 1 1 1 x 2 2 1 x 2 2 2. Expand f ( x) in ascending powers of x. Simplify the coefficients. Use the “shortcut” expansions of 1 1 1 x and 1 x to save time. 1 3 9 2 x x 2 4 8 68 . 1 x 1 x2 1 x 1 x 2 1 2 1 1 x 2 1 1 For each term. 1 1 1 2 x x . 1 (i) (working omitted) 1 x 1 f ( x) 2 1 x 2 x f ( x) 1 x 1 x2 2 x 1 (ii) Express f ( x) as a sum of terms of the form a bx c n . 1 x 1 x2 3.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 1 [N2006/I/12 (Modified)] (i) Express f ( x) 1 x 2 x2 2 x 1 x2 (ii) (iii) (iv) in partial fractions. 1. Deduce the equation of the tangent to the curve y f ( x) at the point where x 0 . Solution Comments Use the partial fractions decomposition for a non-repeated quadratic factor in the denominator (in MF15). 2 4 8 1 x2 x 1 1 1 2 x x 2 4 8 Ignore x x2 x3 when simplifying. Write “ ” after the first few terms of an expansion (after any expansion in any step). Find the set of values of x for which the expansion is valid. up to and including the term in x 2 .

find y when x 0 .e. Example 2 [HCI08/Prelim/I/6] 1 x Expand 1 x n in ascending powers of x up to and including the term in x 2 . find dy when x 0 . p 19 . Use set-builder notation.1 . y Questions on binomial expansion or power series commonly include differentiation and integration concepts. in the form .e. State the set of values of x for which the series expansion is valid. 2 Find the intersection of the 2 sets of values. both binomial expansions involved must be 1 x 1 and x 2 1 . and give the final answer in set notation. Hence find an approximation to the fourth root of positive integers. since the question specified “set of values”. NOT interval notation i. To find the point of dx tangency. where p and q are q 21 69 . For the entire expansion to be valid. y1 has equation: y y1 m( x x1 ) 1 dy 3 and . A straight line with gradient m and passing through a point with coordinates x1 . 1. the expressions were 1 1 x 2 1 and 1 x 2 1 . (iv) y dy dx 1 2 3 9 2 x x 4 8 3 9 x 4 4 To find the gradient of the tangent. i. 2 dx 4 Equation of required tangent: 1 3 y x 0 2 4 3 1 y x 4 2 When x 0 . valid.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iii) 1 x 1 and x 2 1 2 x 2 and x 1 2 x 2 and 1 x 1 For the expansion to be valid. : 1 x 1} Set of values of x {x Immediately before the binomial expansion was used.

nx . Do the same with the other 2 terms in the first bracket. including only terms up to x 2 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution Comments 1 x 1 x n 1 x n 1 x n Express f x as a product of terms of the form a bx c n .e. While doing so. the expressions n n were 1 x and 1 x . : 1 x 1} Immediately before the binomial expansion was used. and 2 in the second bracket in turn. 70 . but both inequalities are equivalent to x 1. For the expansion to be valid. i. – x 1 and x 1 . discard any terms which have a power more than 2. both binomial expansions involved must be valid. For the entire expansion to be valid. 1 n 1 1 nx x n x n n 1 2! n 2! x2 x2 x 2 Use the binomial expansion for rational index. x2 n 1 Write out the entire expression before simplifying to reduce careless mistakes. since the expansion should be up to x 2 . x Set of values of x {x 1. n n 1 2 2 1 nx n n 1 1 nx n n 1 2 1 2nx 2n 2 x 2 x 2 nx n 2x 2 n n 1 2 x2 Choose 1 from the first bracket and n n 1 2 x multiply it with 1.

You may come up with . write “ ” instead of “ ” where the approximation occurs. It remains to 21 make 19 1 x . 20 Try to write 1 1 1 1 1 2 2 4 20 4 20 3121 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society 1 1 and n . 20 4 1 1 . since raised to the power 4 21 4 19 is the fourth root of . 20 19 21 19 in a similar form as 21 1 x 20 1 . x . Since x 20 Let x 1 x 1 x n 1 1 20 1 1 20 1 4 1 4 1 19 1 . 21 1 x with x 1 for the expansion to be valid. Substitute x 71 . where p 3121 and q 3200 . n Solve for x directly to get x OR 20 1 20 1 4 1 . 1 x 20 1 which suggests dividing both top and bottom by 20 to get 1 1 20 1 1 20 1 Hence. the expansion is valid. Since only the first few terms are taken. 3200 2 2 1 back into the 20 expansion.

then OP Angle. then: OP OA OB OA OB . 1 acute and 1 obtuse. 2 planes.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Vectors When representing vectors as single letters. 72 . Between 2 vectors. write a squiggle underneath: a . . thus there is no modulus around the dot product. l2 : r a 2 b2 : 2 planes p 1. 2 In particular. usually used to specify ˆ direction. The modulus around the dot product in finds the acute angle. Basic Formulae Distance between 2 points a1 If 2 points have position vectors a 2 b1 b2 . The unit vector a of a vector a is given by: a ˆ a a Ratio theorem If the point P divides AB such that AP : PB : . between: 2 vectors a and b : 2 lines l1 : r a1 b1 . there is only 1 unique angle. or a line and a plane. then: b3 2 a2 and b a3 2 AB a1 b1 a2 b2 a3 b3 Unit vector A unit vector is a vector with modulus 1. 1 : r n1 p2: 2 : r n2 a line l : r a b and a plane : r n p : cos 1 a b a b b1 b 2 b1 b 2 n1 n 2 n1 n 2 cos 1 cos 1 sin 1 b n b n Between 2 lines. there are generally 2 angles. if P is the midpoint of AB .

get the scalar product form. e. b . and you know the vectors a and b . You can then find the lengths AC and BC . Cartesian n1 x n 2 y n 3 z Use b c n to find a normal Set special conditions. . Q on the plane. Expand the dot product to the plane. c. then combine them to get the Use a Cartesian form.g. P . as shown. Calculation a1 b1 a2 a3 a a b2 b3 a 2 a1b1 a2b2 a3b3 a2 a3 b2 b3 a3b1 a1b3 a1b2 a2b1 Lengths & Areas A Say the vectors a . Dot Product VS Cross Product Dot Product ( a b ) Scalar Cross Product ( a b ) Vector perpendicular to both a and b Magnitude: a b sin Direction: Right-Hand Grip Rule a1 b1 a2b3 a3b2 Output Formula a b a b cos . b a Length AC Length of projection of AB onto b AB cos ˆ ab Length BC Perpendicular distance from B to b AB sin ˆ a b 73 C c B . then a n p to x y 0 to find 3 points to get the Cartesian equation. A . The vector form is then r OA AP AQ . Cartesian x a1 y a2 b1 b2 Make the subject for each of the 3 equations. c form a right-angled triangle.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Equations of Lines & Planes Line Vector r a b. Plane Vector r a b z a3 b3 a1 a2 and b a3 p b1 b2 b3 Scalar Product r n a n p to get vector form. where is the angle between a and b . Let A .

Solve for Say a point P has position vector p . and a line l has equation r a b . F . Also. Since PF and l are perpendicular. PF AP b Say a point P has position vector p . f a b for some . as shown. f p n for some . Let C be the point such that OACB is a parallelogram. PF AP n as shown. with position vector f : Since F lies on l . Find f as above. PF 3. F .f n p p n n p to find f . a b p b 0 F A Solve for to find f . By Pythagoras’ Theorem. P From P to : n Foot of perpendicular. PF f p 2. Say point A lies on l .Mathematics Revision Notes 2009 RI(JC) Mathematics Society A a C Let the points A and B have position vectors a and b respectively. l :r a b Perpendicular distance. with position vector f : Since PF is perpendicular to . f p b 0 . PF is parallel to n . since P lies on PF : . Say point A lies on . PF Use any of these 3 methods. l PF : r p n. 1. PF : AP 2 AF 2 . Since F lies on l PF . Perpendicular distance. PF and a plane has equation r n p . Find AP and AF . Area of triangle OAB 1 a b 2 O b B Area of parallelogram OACB a b Foot of Perpendicular & Perpendicular Distance P From P to l : Foot of perpendicular. F A :r n p Since F lies on . PF : 74 .

l1 : r Parallel: Intersecting at a point: a1 b1 and l2 : r a2 b2 b 1 and b 2 are parallel (see above). q1 Point. 2 Vectors. A . Q . Use PlySmlt2 on the GC to solve SLEs. b 1 and b 2 are NOT parallel (see above). b 1 and b 2 are NOT parallel (see above). AND The SLE given by a1 for and . p: q n p b has a real solution Lies on a plane Lies on a plane :r n b :r a c : The SLE given by q a solution for and . ab 0 2 Lines in 3 dimensions. OP a1 Skew: 0 1 b (or a 2 0 b 2 ). p: n1q1 n2 q 2 n3 q 3 p b c has a real Lies on a plane : n1 x n2 y n3 z 75 . C Collinear AB . AC .Mathematics Revision Notes 2009 RI(JC) Mathematics Society SLE denotes system of linear equations. P : If the solutions are 0 and 0 . with position vector q q2 q3 Lies on a line l : r a b : The SLE given by q a for . b1 a2 b 2 has no real solution 3 Points. (on the same line): You may pick any 2 and show that they are parallel (see above). AND The SLE given by a1 solution for and . b1 a2 b 2 has a unique real Point of intersection. a and b Parallel: The SLE given by a Perpendicular: b has a non-zero real solution for . BC are all parallel. B .

: :r n p p l lies on l is parallel to but does not lie on l intersects b n 0 AND a n b n 0 AND a n p at one point: bn 0 l is perpendicular to if b and n are parallel (see above). n and m are NOT parallel (see above) Line of intersection.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Line & Plane. l : r a b and : . 1 :r n p and 2 :r m q . OP a 0 b n1 m1 2 Planes. a 2 . x3 represent x . z respectively and a1 . Point of intersection. x 2 . y . l : Expand the dot products into Cartesian equations to obtain the SLE: n1 x n2 y n3 z m1 x m 2 y m 3 z p q If PlySmlt2 on the GC solves the SLE to give: x1 x2 x3 a1 b1 x3 a2 b2 x3 x3 where x1 . 1 76 . If . b1 . b2 are constants. P : Solve a the solution is 0 b n p for . where n n2 n3 and m m2 m3 Parallel: Intersecting along a line: n and m are parallel (see above). a1 l a2 0 b1 b2 .

Find (i) (ii) (iii) the acute angle between l1 and the z -axis. where 2 planes intersect along a line. x0 y0 z0 OP Intersect along one common line: x and y can be solved in terms of z . at any common point or line: Intersect at one common point: Unique solution for x . y . Solution (i) Let x 8 2 5 y 2 z 1. the position vector of the foot of the perpendicular. y 0 . Do not intersect No real solution for x . P : If the solution to the SLE is x0 . Comments Convert Cartesian equations of lines to vector equations.Mathematics Revision Notes 2009 RI(JC) Mathematics Society 3 Planes Use their Cartesian equations to obtain a SLE with 3 equations in 3 variables: x . z . Example 1 [TJC06/Prelim/I/14] x 8 The lines l1 and l 2 have equations 2 5 y 2 6 z 1 and r 4 respectively. z 0 . Hence. where 1 . 77 . z . from the origin to l1 . which are usually much more useful. y . the point of intersection of l1 and l 2 . find a vector equation of the reflection of line l 2 in the line l1 . Line of intersection. l : See above. z . Point of intersection. N . y .

1 Express x . ON 8 2 5 2 1 2 2 1 (Since ON is perpendicular to l1 ) 0 16 10 1 4 3 8 2 3 ON 5 2 3 1 3 4 0 Expand the dot product. z in terms of . for some . 1 0 0 . and the z -axis passes through the origin. (ii) ON 8 2 5 2 1 2 2 1 (Since N lies on l1 ) . then solve for . while x and y do not.) Since we are finding the acute angle.5 (1 d. y .Mathematics Revision Notes 2009 RI(JC) Mathematics Society x Then: y z l1 : r 8 5 1 8 2 5 2 1 2 2 . into the original equation for 78 . b b 2 0 cos 1 1 2 b1 b 2 2 0 cos 1 1 2 2 1 1 0 0 1 70. explaining its vector equation. Let be the acute angle between l1 and For the acute angle between 2 lines. 2 1 2 Substitute ON .p. use: the z -axis. place a modulus around the dot product. z -axis: r z varies along the z -axis.

From the diagram. solve using PlySmlt2 on the GC. 8 5 1 . another point on l3 . OA 2ON 4 PA 2 4 4 2 4 9 6 1. 79 . and RHS has a constant. With reference to the diagram. OP Hence. 1 point is P . The question asks for “the point”. we can use ON from part (ii) to find the position vector of A .5 5 8 5. 6 1.5 1.5 4 1.6. and label the points clearly. part (ii) gives ON and part (iii) gives OP .5 . l2 O Draw a diagram first to illustrate the question.5 0. use information from previous parts. P N l1 A l3 Let l3 be the reflection of l 2 in l1 . Write the system of linear equations such that LHS has and in that order.5 To obtain the vector equation of l3 . From the GC. 1. Then OP for some 2 2 6 4 8 2 5 2 1 6 4 . P is 9. .5 1. we can find 2 points which lie on l3 . Then. Now PA is a direction vector of l3 . In a “Hence” question.5 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iii) Let P be the point of intersection between l1 and l 2 . so give the coordinates of the required point instead of its position vector.5 . and part (ii) gives OP . 9 6 1. since ON AN .

z gives the coordinates of the point of intersection.9 y 17 z 16. (i) (ii) (iii) (iv) Find a vector equation of l . find and . find the 16.6 . y . choose OA over OP ) and leave b in the simplest integer form.6 Use the Cartesian equations of the 3 planes to obtain a SLE with 3 equations in 3 variables: x . From the GC. choose the simplest vector for a (here. z . p 3 are 2 x 5 y 3z 3 3x 2 y 5 z 5 5x y 17 z respectively.9 . not its position vector. When coordinates of the point at which these planes meet. p2 . 11 PA l3 : r When writing the equation of a line l : r a b . Solve it using GC. 20. Give the coordinates of the point. Given that all three planes meet in the line l .9 and 16. y . y . 1. The planes p1 and p2 intersect in a line l .6 . 2 x 5 y 3z 3 3x 2 y 5 z 5 5 x 20. z x 11 11 7 11 Coordinates of the point at which p1 .3 . This makes your working easier in any later parts. 4 4 . Example 2 [N08/I/11] The equations of three planes p1 . then the unique solution for x . 20. Solution When Comments Substitute the given values for and . Given instead that the three planes have no point in common.Mathematics Revision Notes 2009 RI(JC) Mathematics Society 1 2 10 16 11 4 2 4 10 16 . what can be said about the values of and ? Find the Cartesian equation of the plane which contains l and the point 1. p2 and p 3 meet: 4 4 7 . where and are constants. 11 11 11 80 . .

l must lie on p3 . Solve it using GC.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (i) 2 x 5 y 3z 3 3x 2 y 5 z 5 From the GC. If all 3 planes intersect in l . 1 1 0 1 0 1 17 gives the position vector of a point on l . 1 1 1 1 5 s 1 1 . . with equation found in part (i). If the GC gives the solutions for x and y in terms of z . as variable. Write z as a parameter. lies on l also lies on p3 . with and constant. for all . to get the . where z 1 1 0 1 1 . l :r (ii) p3 : r 5 17 p1 and p2 already intersect in the line l . x 1 y 1 . The method shown here uses the fact that Since l lies on p3 . comparing coefficients. say vector equation of the line. 22 5 22 . Since 22 0 every point on l lies on p3 . 1 Use the Cartesian equations of the 2 planes to obtain a SLE with 2 equations in 3 variables: x . every point which for each value of . z . 5 5 0 17 22 5 Since the above equation is true for all Consider 22 5 to have . the 2 planes intersect along a line. every value of 5 fulfils the equation. 17 17 81 . y .

then any point on l will be a point in common to p1 . p3 will intersect. if and only if l is parallel to p3 and l Case 1: l is not parallel to p3 . p2 and p3 .e. either every point on l or no point on l lies on p3 . Hence. then l and does not intersect p3 . p2 and p 3 have no point in common p1 and p2 already intersect in the line l . Case 2: l lies on p3 . 17 82 . l is parallel to p3 if and only if: 1 1 1 5 0 17 l is parallel to p3 if and only if the direction vector of l is perpendicular to the normal to p3 . 22 and 22 and 5 This corresponds to the situation when l is parallel to p3 and l does not intersect p3 . it is enough to check if a point on l lies on p3 . both must be true. When l is parallel to p3 . l intersects p3 if and only if: 1 1 0 5 17 5 Hence. p1 . This intersection point will be a point in common to p1 . p2 and p3 . 5 17 0 22 When l is parallel to p3 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iii) p1 . then since p1 and p2 only intersect in l . their dot product is 0. p2 and p 3 have no point in common. Case 3: l is parallel to p3 and l does not intersect p3 . i.

3 to find another vector which lies on . the direction vector of l also lies on . as required. Expand the dot product to get the Cartesian equation of . Once n is known. 1. Use a point which lies on l and the given point 1. p4 : r 3 1 2 1 1 3 3 1 2 2 2. 1. find the constant p using r n a n p where a is the position vector of any point on . Since l lies on . x y z 3 1 2 2 Cartesian equation of 3x y 2 z 2 : 83 . 1 2 1 0 3 3 2 0 3 3 1 2 To get the scalar product equation of the plane in the form r n p .Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iv) Let 1 1 0 1 1 1 be the required plane. Find the normal n by taking the cross product of 2 vectors which lie on the plane.

o Calculate the basic angle. 0 a Re On an Argand diagram. the horizontal axis is the real axis. Let z a bi . Im z o o o o Re z Im z a b (not bi ) Re z is the x -coordinate of Z Im z is the y -coordinate of Z Modulus. arg z o Angle that OZ makes with the positive real axis. where a. b b z arg z consists of a real part a and an imaginary part bi . Real part. A tan A A 0 A A Re arg z A arg z A 84 . On an Argand diagram. let Z and Z * be the points representing z and z* respectively. and the vertical axis is the imaginary axis. Any complex number z a bi can be represented geometrically as a point on the complex plane. b and i 1. Re z and imaginary part. a. b be a complex number in Cartesian form.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Complex Numbers Im A complex number in Cartesian form: z a bi Z a. in an anticlockwise sense arg z o o arg z A Im arg z A arg 0 is undefined 1 b a o Draw the Argand diagram: arg z depends on both A and the quadrant which the point Z is in. z o z a 2 b2 o Distance of Z from the origin (length of OZ ) Argument.

subtraction. b r sin o Between Polar and Exponential: Use the same r and 85 . arg z o Polar or Exponential to Cartesian: a r cos . i. z* w* . multiplication. multiply both numerator and denominator by the complex conjugate of the denominator. and division: z z* * * . z z* arg z r cos i sin Conversion between forms o Cartesian to Polar or Exponential: r z. z * o Re z * o o o o o Im z * z* z arg z 2 a bi Re z Im z arg z * zz * z z z* 2 Re z o z z * 2i Im z o The conjugate distributes over addition. division and exponentiation of complex numbers If z rei o o o r cos r re i i sin .e. z w zw zw . z is in Polar or Exponential form. z w z* w* * w w o To simplify a fraction with a complex number in its denominator. which are more useful. * * * * Forms of Complex Numbers Cartesian form o z a bi o Useful for addition and subtraction of complex numbers Polar form (also Trigonometric form or Modulus-Argument form) o z r cos i sin o Convert from polar to Cartesian or exponential forms. o Z is Z * reflected about the x -axis. Exponential form o z re i o Useful for multiplication.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Conjugate.

1 1 On an Argand diagram. A clockwise pivot by 2 radians about the origin and scaling by a factor of z r2 gives the position vector of the point representing 1 . the . Hence. o OZ1 OZ 2 OB . Addition & Subtraction Im A Z1 B Z2 0 z1 z2 a1 a2 b1 b2 i Re o b1 b2 i o z1 z2 a1 a2 o Complex numbers can be represented as vectors on an Argand diagram. In particular. addition and subtraction of complex numbers can be represented as addition and subtraction of vectors.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Operations on Complex Numbers Let z1 a1 bi re i 1 and z2 a2 b2i r2e i 2 . let Z 1 and Z 2 be the points representing z1 and z2 respectively. the position vector of the point representing z1 z2 . 86 2 . An anticlockwise pivot by 2 radians about the origin and scaling by a factor of r2 gives the position vector of the point representing z1 z 2 . Multiplication & Division o z1 z2 z1 z2 r1r2 o z1 z2 z1 z2 z1 z2 r1 r2 o arg z1 z2 o arg arg z1 arg z1 arg z2 arg z2 1 2 1 2 o Starting from O Z 1 . o OZ1 OZ2 OA . the position vector of the point representing z1 z 2 . z2 An anticlockwise pivot by radians about the origin gives the position vector of the point representing z1ei . when position vector of the point representing iz1 is given.

then * is also a root of the equation.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Complex Roots of Polynomial Equations Non-real complex roots of a polynomial equation with real coefficients occur in conjugate pairs. Example: Since z 3 5z 2 17 z 13 0 is a polynomial equation with real * 2 3 is a root. while A and B are fixed points. Locus of Z : The area outside the circle with centre A and radius r . If is a root of a polynomial equation with real coefficients. Equivalently. 87 . Locus of Z : The area inside the circle with centre A and radius r . Im z a r z a r r 0 A Re What it means: The distance between Z and A is a constant. A . b respectively. z 2 3i is also a factor. i sin n r n cos n Loci of Complex Numbers A locus (plural loci) is a set of points which fulfill a given condition. arg z a is the angle made by the vector AZ with the positive real axis. z a is the distance between Z and A . r 0 Re z a r The distance between Z and A is more than r . On an Argand diagram. coefficients and z since z 2 3i is a factor of the polynomial. a . z i 2 3i is also a root. B be the points representing z . Z is a variable point which can move. De Moivre’s Theorem If z rei zn r n ei n r cos i sin . r Locus of Z : A circle with centre A and radius r Im z a r z a r The distance between Z and A is less than r . let Z .

88 . anticlockwise to the locus arg z a 2 . A 0 2 1 Locus of Z : The area bounded from the locus arg z a Re 1 . Im arg z a arg z a What it means: The angle made by AZ with the positive real axis is a constant. z b is the area containing B . Re A 0 Locus of Z : The half-line starting at (but not including) A and making an angle with the positive real axis Im 1 arg z a 2 1 arg z a 2 The angle made by AZ with the positive real axis is between 1 and 2 . z a z b Z is closer to B than it is to A .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Im A z a z b z a z b What it means: Z is an equal distance from each of A and B B 0 Re Locus of Z : The perpendicular bisector of the line joining A and B Im A z a z b z a z b Z is closer to A than it is to B . B 0 Re Locus of Z in each case: z a z b separates the plane into 2 areas: z a z a z b is the area containing A .

z n re 2k i .1. 3. Use (A) (B) to eliminate the 2 z term from both equations. where a and b are real. 2. z 2k n i 6. Say w rei . Using (A). z r e r e 1 n 2k n i 5.(B) 2z 1 i w 6 2i giving each answer in the form a bi . Q: Solve for z and w in 2 simultaneous linear equations. is a constant. Substitute z in (B) with the expression in terms of w . then solve for z . Solve for z . n z rei 2. By De Moivre’s Theorem. then solve for z . and n is a positive integer. A: 1. 1. Reject those values of k which make any expression undefined (denominator 0 ). A: Method 1 1. Introduce 2k to .Mathematics Revision Notes 2009 RI(JC) Mathematics Society FAQ n Q: Find the roots of the equation z w.n 1. where k . where z and w are complex numbers. However.(A) 40 ----. k 0. . Convert w to exponential form. Write the possible values of k . substitute values of k to check if is in the required range. if the roots must be given in the form rei with . Solve for w . n 1 rn e 1 n 2k i 1 n 4. Example [PJC07/Prelim/I/1] Solve the simultaneous equations iw z 1 i ----. 2. There will be n consecutive integer values of k . Method 2 1 i iw . express z in terms of w : z 2. Raise both sides to the power of z 1 . 3. Multiply (A) by 2 such that the coefficient of z in (A) and (B) are both 2. In general. 89 . 3. Solve for w .

2 . A: Interpret what expressions mean geometrically on the Argand diagram. the equation can be expressed as a product of k and n linear factors of the form z . n are the n (possibly repeated) roots of the equation. If the polynomial equation has real coefficients. When the equation is in this form with n linear factors of the form z k z 1 z z 0 2 n then 1 : . 1. A: Draw all circles. for z a r . For loci of the form arg z a . rather than interpret them algebraically. * is also a root. ruler and protractor. and label required regions/sets of points. since arg a a arg 0 is undefined. . Q: Find the minimum/maximum value of z a . For strict inequalities. arg z a or z w . 4. draw a small circle at the point representing a to show that it is excluded.g. e. 90 . find the Cartesian equations of the lines or circles and solve the equations simultaneously. z is a factor of the polynomial. If is a root of the equation. Label all loci with their equations. Ensure the real and imaginary axes are to the same scale. A: If the polynomial equation in z has degree n (the highest power of z is n ). minimum value of z w is the minimum distance between any point on the locus of points representing z and any point on the locus of points representing w .g. draw the locus z a r as a dashed circle. 3. e. As a last resort for finding points of intersection. Using any given roots or factors. Find lengths using the formula for the distance between 2 points. long division or comparing coefficients. lengths and angles accurately and to scale using a compass. Use the general quadratic formula to factorize any quadratic factors. express the polynomial as a product of its factors. OR Find the points of intersection of the loci of points representing z and w . 2. Given that is a non-real complex root of the equation. draw the loci of the equality cases as dashed to show that they are excluded.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Q: Find the complex (real and non-real) roots of a polynomial equation. and the coefficient of z n is k . where z and w are variable while a is fixed. Q: Sketch loci on an Argand diagram.

1 ei 2 1 8 1 i . values of arg z a . o Draw a line from point A through the centre of the circle. cos i sin cos 2 2 cos sin sin 91 . The line and circle will have 2 points of intersection. 2 i 2 Re e 2i Im e 2 2 i The aim is to create expressions of the form z z * or z z * . 8 w 1 1 1 i . giving each root in the form rei Find all the roots of the equation z 2 where r 0 and ./max. Use the distance from A to these 2 points to find the min.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Find right-angled triangles to use trigonometric ratios. z b r and the fixed complex number a is represented by the point A outside the circle. Example 1 [SRJC08/Prelim/II/3] (i) (ii) (iii) 1 ei can be reduced to i cot . o Use tangents to the circle to find the min. which can be written as Re z 2cos 2i sin 2 2 or 2i Im z respectively. Here. values of z a ./max. If the locus of z is a circle. Hence deduce all the roots of w 1 2 Show that Solution Comments (i) 1 ei 1 ei e e 2 i e i 2 i “Factoring out 2 e i 2 i ” is a common theme in 2 Complex Numbers proving questions. it takes the form of factoring out e from both the numerator and denominator.

1 . 2. express w in terms of z . Since w 1 z 1 ei 1 z 1 ei i cot k takes the same values as in part (ii). 2. To deduce the roots of the equation in w . since the values of k are for the roots of the equation in z from part (ii).43. z e 8k 1 32 i 3. . 2. and 8 By De Moivre’s Theorem. 2. where k . Write out the possible values of k . 8k 1 . Raise both sides to the power of 4.Mathematics Revision Notes 2009 RI(JC) Mathematics Society 1 cot i 2 i cot 2 i 1 i i i2 i i 1 (ii) z8 e e 1 1 i 2 8k 1 4 i e 4 1. 0. 1. w 1 equate the two. 2 64 3. 0. Convert 1 1 i to exponential form. where k 3. 1. 2. 3.1 . 32 w 1 z ei w 1 zw z w 1 1 z w1 z w 1 w 1 8 w 1 corresponds with z in part (ii). Use De Moivre’s Theorem. 2 2k 4 i . 1 . there will be exactly 8 consecutive integer values of k which will give in the required range. Since the question requires . 4 where k w i cot 92 . 5. Substitute values of k to check which values are valid. Introduce 2k to 4 . (iii) 1 1 i 2 8k 1 Let .

3i 2 w 2 0 . convert the coefficients of z in P z to their conjugates. Expand z i z 2 az b and compare the coefficients. to avoid careless mistakes. since it will be difficult to compare coefficients. Hence. 4 1 ( 2i ) Use the general quadratic formula to factorize the quadratic factor. it should leave a quadratic factor (degree 2). az b . z 3 1 3i z 2 2 3i z 2 * 0 * Observe that the coefficients of w in Q w are the conjugates of the coefficients of z in P z . Not all of the coefficients of P z are real. including square roots. Let Q w w3 1 3i w2 3i 2 w 2 . Hence. i a root. determine the other roots. When z 2 1 2i z 2i 1 2i 2(1) 3 4i 2 0. the other roots are 2i and 1 . 93 . Hence find the roots of the equation w Solution Let z 3 z i 1 3i w2 Comments 1 3i z 2 z2 2 3i z 2 Let P z z3 1 3i z 2 2 3i z 2 . 1 3i a i a 1 2i 2 2 bi b 2i i z 2 az b z 2 1 2i z 2i After P z (degree 3) is divided by z i (degree 2). for some complex numbers a and b .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 2 [AJC08/Prelim/I/6] If z i is a root of the equation z 3 1 3i z 2 3 2 3i z 2 0 . Use the GC to perform calculations with complex numbers. Do NOT write P z in terms of linear factors first: z i z a z b . Taking conjugates on both sides. z (1 2i) 2i 1 2 2i or 1 Hence. Alternative method Use long division to find the remainder when P z is divided by z i . though i is a root. * i may not be z 3 3 1 3i z 2 2 2 3i z 2 z az bz iz2 aiz bi z 3 a i z 2 b ai z bi Comparing coefficients.

The locus of points representing w is the perpendicular bisector of the line segment joining the origin and C . Plot any such point A . The complex numbers z and w are such that a and w w ia Sketch. which extends to: z3 Hence. the loci of the points representing z and w . a D A a 0 a Re C The locus of points representing z is a circle centred at B with radius a . in a single diagram. 2i . A is in the first 2 quadrant of the complex plane. Example 3 [AJC07/Prelim/II/3] In an Argand diagram. Let C be the point representing –ia . each root of Q w corresponds to the conjugate of a root of P z . * z z z * z* z* z* z* Since the equation in z* is now similar to the equation in w . z 2ia a B w w ia If 0 arg a . the point A represents the fixed complex number a . the roots of w3 1 3i w2 3i 2 w 2 0 are the conjugates of the roots of z 3 1 3i z 2 2 3i z 2 0. 94 . the range of values of arg Solution Im z 2ia 1 in terms of arg a . i . z Comments Let B be the point representing 2ia .Mathematics Revision Notes 2009 RI(JC) Mathematics Society z3 z* * 1 3i * z2 2 * 2 3i z* 2 0 3i 2 z * 2 0 * Simplify the LHS using z w zw * * z* w* . where 0 arg a 2 .e. and 3 3 1 3i z * z * w* . Find: (i) (ii) the minimum value of z w in terms of a . and 1 . i.

If Z is the point representing z . value of arg z a when the locus of points representing z is a circle. a relevant part of it is copied here for easier reference. arg z Then arg(m) arg z arg n . as shown. a . an anticlockwise pivot about the origin by radians and a stretching 2 by factor 2 gives OB . origin. Let M and N represent the complex numbers m and n respectively.Mathematics Revision Notes 2009 RI(JC) Mathematics Society From OA . z 2ia a M a B A a D N a 0 a Re Let M and N be the points at which the Using tangents to the circle is a tangent to the locus of z passes through the common technique to find the min. a clockwise pivot about the origin by (i) Minimum value of z w OD a 3 a 2 1 OC 2 1 a 2 2 radians gives OC . the minimum value of z w is the minimum distance between a point on the locus of points representing Z and a point on the locus of points representing W . 95 . From OA . Do not redraw the diagram. it divides OC into 1 2 segments of equal length. Geometrically. 2 (ii) arg 1 z arg 1 arg z arg z Im Express the condition in terms of a known variable. arg z . is the angle made by the vector OZ with the positive real axis. Since the locus of points representing w is a bisector of OC ./max.

The set of points Q represents another complex number w given by (ii) arg w 2 3i Q. arg a 3 arg z arg a Convert the range of values of arg z to the range of values of arg 1 . of Hence. and add/subtract from it. Example 4 [NJC08/Prelim/I/6] (i) The set of points P in an Argand diagram represents the complex number z that satisfies z 2 i 2 . Sketch the locus of P . In the case where such that the locus of Q meets the locus of P tan 1 3 . 4 96 . and flip the inequality signs. The line OB passes through the centre of the circle. z arg a arg a 3 2 3 arg z arg 1 z arg a arg a 2 3 3 Multiply each term in the inequality by 1 . Give a geometrical description of the locus of Find the range of values of more than once. use the angle OB makes with the positive real axis as a reference. (a) (b) where . OA makes an angle of arg a with the 6 arg n arg a arg a arg a BON positive real axis. Hence. OB makes an arg a 2 3 2 with the positive real axis. and BOA angle 2 6 2 3 2 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society BOM BON sin 1 a 2a 6 arg m arg a arg a arg a 2 2 3 2 BOM Use arg a since the question requires the final answer to be in terms of arg a . find the least value of z w in exact form.

1 3 2 3 97 . and add/subtract from it. part (ii). (ii)(a) Im z 2 i (2. arg w 2 3i (ii) The locus of Q is a half-line starting at (but not including) 2.1 with radius 2 . Remember to write “(but not including)”. 0 Re Show on the diagram that the circle is tangent to the imaginary axis. 3) 2 4 2 4 The line passing through the centre of the 1 2 2 sin sin 2 1 2 2 6 6 1 2 3 2 3 circle makes an angle of with the 2 positive real axis.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution (i) Im Comments z 2 i (2. Hence. 3 and making an angle with the positive real axis. use this angle as a reference. Examiners will look out for the relative position of lines and circles to the axes.1 ) 2 2 4 2 0 Re 2 1 (2.1 ) 2 2 z 2 i 2 The locus of P is a circle centred at 2. Using tangents to the circle is the same technique as in Example 3.

where DCB and ACB are right angles. the least value of z w is the minimum distance between a point on the locus of P and a point on the locus of Q . The only useful right-angled triangle is ABC . Since DC AC 2 . Since tan . From the diagram. aim to find AC first. Knowing AB and ABC .1 ) 2 2 The locus of Q (which represents w ) should have gradient arg w 2 3i 6 D tan tan Re 1 3 4 3 4 0 4 C 3 B(2. draw a right-angled triangle 3 involving . the 4 side opposite and the side adjacent to have lengths in the ratio 3: 4 . 98 . AC can be found using trigonometric ratios on the right-angled triangle ABC . 3) tan 1 4 Geometrically. this minimum distance is DC . 3 4 AB 4 ABC 2 tan 1 3 4 sin ABC AC AC AB AB sin ABC 4sin 2 tan 1 3 4 4 16 5 5 Least value of z w AC 4 Let DC AC 2 6 5 tan 1 3 . 4 5 2 3 4 To calculate sin in exact 2 form.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (ii)(b) Im z 2 i A(2.

Mathematics Revision Notes 2009 RI(JC) Mathematics Society

Permutations & Combinations
Basic Counting Principles Addition Principle: If there are a ways to do something and b ways to do another thing, and the two things cannot be done simultaneously, then there are a b total ways to do exactly one of the things. Multiplication Principle: If there are a ways to do something and b ways to do another thing, then there are a b ways to do both things in succession. Choosing & Permuting Is the order of objects important? Look for keywords in the question, e.g. o Important: arrange, order, in a row, around a table o Not Important: choose, select, form groups/committees/teams Yes Can each object be repeated any number of times? Yes No. of arrangements of r objects (possibly repeated) from n distinct objects: nr e.g. Form a string of 4 letters using the English alphabet: n 26 , r 4 , No. of ways 26 4 Are all the objects distinct? Yes No. of arrangements of r objects (all distinct) from n distinct objects:
n

No No. of ways to choose r objects from n distinct objects:
n

No

Cr

n! r! n r !

e.g. Choose 2 monitors from a class of 24 students: n 24 , r 2 No. of ways No
24

C2

No. of arrangements of n objects, where: n1 are of type 1,

Pr

n! n r !

n2 are of type 2, … nk are of type k :
n! n1 !n2 ! nk ! e.g. Arrange 2 red balls, 3 green balls, 5 blue balls in a row. Since 2 3 5 10 ,
No. of ways
10 ! 2 ! 3! (5 !)

e.g. Choose 1 monitor, 1 treasurer, 1 secretary (all distinct) from a class of 24 students: n 24, r 3 , No. of ways 24 P3

99

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Useful Techniques Circular permutations: No. of permutations of distinct objects in a circle = ( − 1)! , assuming the positions around the circle are identical. Grouping adjacent objects: If some objects must be adjacent, group them and treat them as a single object. When arranging all the objects/groups, remember to also arrange the grouped objects within each group. Slots: e.g. If 3 women and 5 men are to be seated in a row such that no 2 women are adjacent, arrange the 5 men first. There are 6 ‘slots’ (denoted | ) between and beside the 5 men (denoted M): |M|M|M|M|M| Then, arrange the 3 women in 3 distinct ‘slots’. Hence, the number of ways is 5!
6

P3 .

Complement method: Use it when direct computation is tedious, especially if key phrases like ‘at least one’ appear in the question. If A is a condition, the number of objects with A equals the total number of objects, minus the number of objects without A : n A n total n (A ') Case-by-case method: Use it if all else fails. Check that each possibility is counted in one and exactly one case, such that no possibilities are double-counted or missed out.

Example 1 [RJC08/Tutorial 12/Section D/11] Find the number of permutations of all eight letters of the word BELIEVED in which (i) there is no restriction; (ii) the three letter ‘E’s are next to each other; (iii) the three letter ‘E’s are always separated. Solution (i)
8! 6720 3!

Comments The letters in BELIEVED are 3 ‘E’s and 1 each of ‘B’, ‘L’, n! ‘I’, ‘V’, ‘D’. Since only ‘E’ is repeated, use n1 !n2 ! nk ! with n 8 and n1 3 ( n2 , , n8 1 ). Group the 3 ‘E’s as 1 object. Arrange the 6 resulting objects.
24 00

(ii) (iii)

6! 720
5!
6

C3

Arrange B, L, I, V, D first, for 5! permutations. There are 6 ‘slots’ between and beside these 5 letters: | B | L | I | V | D |. Place the 3 ‘E’s in 3 of these 6 slots. Since the order of the ‘E’s does not matter, use 6C3 instead of 6 P3 . 100

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 2 [RJC08/Tutorial 12/Section D/12] 3 boys, 2 girls, and a dog sit at a round table. How many ways can they be arranged if (i) (ii) (iii) there is no restriction? the dog is to be seated between the 2 girls? the dog is to be seated between any 2 boys? Solution (i) (ii) Comments Use n 1 ! directly with n 6 . Group the 2 girls and the dog as 1 object. Arrange the 4 resulting objects, then multiply by 2 as the objects within the group can be arranged in 2 ways. Denoting the girls by G1 and G2 and the dog by D , the group can be G1 DG2 or G2 DG1 in clockwise order. (iii)

6 1 ! 120
(4 1)! 2 12

4 1 ! 3! 36

Group some 2 boys and the dog as 1 object. Arrange the 4 resulting objects to get 4 1 ! . Since the 3 boys are distinct, multiply by 3! to arrange them.

Example 3 [HCJC01/I/7] Thirteen cards each bear a single letter from the first thirteen letters A, B, …, L, M of the alphabet. Seven cards are selected at random from the thirteen cards, and the order of selection is not relevant. How many selections can be made if the seven cards (i) (ii) contain at least two but not more than four of the letters A, B, C, D and E? contain four but not five consecutive letters, e.g. A, B, C, D, G, H, I? Solution (i)
5

Comments
5

C2

8

C5

C3

8

C4

5

C4

8

C 3 Use a case-by-case method.

1580

The 7 cards chosen contain: Case (1): exactly 2 of the letters A, B, C, D, E. Case (2): exactly 3 of those letters. Case (3): exactly 4 of those letters. e.g. for Case (1) when there are 2 of those 5 letters, choose those 2 first to get 5C 2 . We are left with 5 letters to choose and 8 letters to choose from, to get 8 C5 . 101

and: Case (1): Contain A.: If B. i. 2 (not 1) other letters cannot be chosen. Hence. I cannot be chosen. K.g. K. E are fixed. except after the 4 consecutive letters are fixed. since it would form 5 consecutive letters.e. C. We are left with 3 letters to choose and 8 letters to choose from. E cannot be chosen. 1 other letter cannot be chosen. we get 7C3 for each sub-case. The 7 cards chosen have no 5 consecutive letters.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (ii) 8 C3 2 7 C3 8 392 Since there are only 7 cards. D or J. A and F cannot be chosen. M (2 sub-cases) Case (2): Contain any other 4 consecutive letters (8 sub-cases) For Case (1). D.: If A. D are fixed. B. to get 8C3 for each sub-case. C. L. 102 . Use a case-bycase method. after the 4 consecutive letters are fixed. B. e. L. they can contain at most 1 string of 4 consecutive letters. M are fixed. If J. Case (2) is similar. C.

o P A| B P A o o P B| A P A B P B P A P B Useful Techniques Use set diagrams to illustrate the problem. Conditional probability: P A | B is the conditional probability of A given B . B is the probability that only A occurs. P A P A B is the probability that both A and B occur. o P A| B P A B . Mutually exclusive: A and B are mutually exclusive if A and B cannot both occur. especially if key phrases like ‘at least one’ appear in the question. Equivalently. Check that each possibility is counted in one and exactly one case. then since P A P A B P A P B 103 . such that no possibilities are double-counted or missed out. or both A and B occur. P A B P A P B P A B B 0. In particular. if A and B are mutually exclusive. P A 1 P A' Case-by-case method: Use it if all else fails. only B occurs. where P B P B 0. o P A B 0 o P A B P A P B Independent: A and B are independent if the occurrence or non-occurrence of one event has no influence on the occurrence or non-occurrence of the other.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Probability Definitions Let A and B be random events. Complement method: Use it when direct computation is tedious. P A | B is the probability of A occurring. For any 2 random events A and B . given that B has occurred.

e. or P B o Consider the reduced sample space given that B has occurred. 1 B A 1 2 1 B' B 1 A' 1 2 B' is the probability of A . To find P A probabilities. o P&C.Mathematics Revision Notes 2009 RI(JC) Mathematics Society To find the probability of A occurring given that B has occurred. Probability Tree (Tree Diagram) Use a probability tree when the fixed probabilities or conditional probabilities of 2 or 3 events are given.g. P A' P A' B . multiply the 2 relevant P B | A' To find P B or P B ' . while those on the second level are conditional probabilities. o The Binomial distribution. e. P B| A 1 .g. when objects are chosen with replacement. P A 1 B' . P B P A P B| A P A' P B | A' 1 1 2 104 . P A ' B B . where n X is the number of outcomes for X to occur. while 1 is the conditional probability of B given A . P A B o Calculate P A | B . P A' 1 B ' . add the probabilities of the branches which lead to B or B ' . the probabilities on the first level of the diagram are for only 1 event. Example: P A Probability tree: . n B Use concepts from Permutations & Combinations (P&C) to count cases. and P B | A ' 2 . In general. then calculate n A B P A| B . when objects are chosen without replacement. Use concepts from other topics for alternative approaches to questions.

else to account for order. (ii) 1 1 4 3 2 1 14 10 9 8 7 97 105 1 1 14 4 10 C4 C4 97 10 5 Using the complement method: P G 1. at least 2 green counters are drawn. 4 counters are drawn at random from the bag.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 1 [D90/II/6] A bag contains 4 red counters and 6 green counters. State with a reason whether or not the events ‘at least 2 green counters are drawn’ and ‘at least 1 counter of each colour is drawn’ are independent. 3 1 P G 1 P G 0 0 P (G (i) 4) Similar to the Probability approach. 105 . 3 37 42 1 97 14 105 6 C1 10 4 C3 C4 37 42 P G 1 Again. assume the counters are distinct. Probability approach (i) 6 5 4 3 10 9 8 7 1 14 P&C approach 6 10 C4 C4 1 14 Let G be the number of green counters out Assume the counters are distinct. Out of these. so that of the 4 counters drawn. at least 1 counter of each colour is drawn. 6C4 There is no need to multiply by anything combinations are all green. (iii) 1 97 6 4 3 2 4! 14 105 10 9 8 7 3! P G P G 2. There are 6C1 combinations of 1 green counter. since all 4 counters are the same type (green). without replacement. counters have an equal probability of being chosen. and 4 (i) (ii) P(G 1) 4! when finding P(G 1) . 3. Calculate the probability that: (i) (ii) (iii) (iv) all the counters drawn are green. this is each of the 10C4 combinations of 4 P(G 4) . 2. 2. given that at least 1 counter of each colour is drawn. Then. C3 of 3 red counters. to 3! account for order when there are different types of items (green and red). 4 4 P G 1. at least 2 green counters are drawn. Multiply by giving 6C1 4C3 total combinations of 1 green and 3 red.

42 P A| B P B .3 . 2. 85 Since P A | B 97 37 and P A . and P( A | B) are known from parts (ii). using the formula for conditional probability: P (G 2.3) Let A be the event that at least 2 green counters are drawn. (iii).3. 4 P(G 4) (iii) (i) (ii) P (G 1.3 . even though the question may not require it explicitly.3) P G 2. P ( A) . 2. thus A and B are not independent.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution (iv) Comments 37 1 42 14 97 105 85 97 In probability questions with many parts.3) P (G 1. 2. choose the condition P A P( A | B) to check if A and B are independent. Since the intersection of these 2 events is G 2. We want to find the conditional probability of G 2. 4 given that G 1. and (iv) respectively. look out for ways to use your answers to earlier parts. Since P ( B ) . Let B be the event that at least 1 counter of each colour is drawn. 106 .3.

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Example 2 [J88/II/6] 4 girls, Amanda, Beryl, Claire, and Dorothy, and 3 boys, Edward, Frank, and George, stand in a queue in random order. Find the probability that: (i) (ii) (iii) (iv) (v) the first 2 in the queue are Amanda and Beryl, in that order, either Frank is first or Edward is last (or both), no 2 girls stand next to each other, all 4 girls stand next to each other, all 4 girls stand next to each other, given that at least 2 girls stand next to each other. Probability approach (i)
1 1 7 6 1 42

P&C approach
5! 7! 1 42

Amanda is in position 1 with probability 1 , then once there are 6 positions 7 remaining, Beryl is in position 2 with 1 probability . 6 (ii)
1 7 1 7 1 1 7 6 11 42

There are 7! permutations of the 7 people, each with equal probability. 5! of these permutations have Amanda in position 1 and Beryl in position 2.

6! 6! 5! 7!

11 42

Let F be the event that Frank is first, Similar to the probability approach. and E be the event that Edward is last. Use P F E P F P E P( F E) . (iii)
4 3 2 1 7 6 5 4 1 35 4! 3! 7! 1 35

The 7 people must be in the order Since we are counting permutations with ‘gbgbgbg’, where ‘g’ denotes a girl and this approach, we must permute both boys ‘b’ a boy. and girls in the ‘gbgbgbg’ order, then divide it by the total of 7! permutations. Position the girls first, then the positions of the boys are fixed and the condition is satisfied. (iv)
4 3 2 1 4 7 6 5 4 4 35 4! 3! 4 7! 4 35

This is similar to part (iii), except there Similar to the probability approach. are 4 possible positions of ‘gggg’ within or beside ‘bbb’.

107

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution (v)
4 35 1 1 35

Comments Again, in probability questions with many parts, look out for ways to use your answers to earlier parts, even though the question may not require it explicitly. Let A be the event that all 4 girls stand next to each other. Let B be the event that at least 2 girls stand next to each other. Then A B A , since B is a subset of A .

2 17

P( B' ) is given by the answer to (iii), since ‘at least 2 girls stand next to each other’ and ‘no 2 girls stand next to each other’ are complementary events. Using the formula for conditional probability:
P A| B P( A B) P( B) P ( A) P( B) P ( A) 1 P ( B ') (iv) 1 (iii)

Example 3 [N01/II/11a] Events A and B are such that P A Find (i) (ii)
1 , P B| A 3 1 , and P A ' 3 B' 1 . 6

P A P B .

B ,

Solution (i)
P A B 1 P A' B' 5 6

Comments On a Venn diagram, A B is the area within one or both of A and B , and A ' B ' is the area not within either A or B . Hence, P A
B 1 P A' B' .

(ii)

P B P A P A B B P A P A P A B P A P( B | A)

Rewrite the given equations in terms of P ( A) , P ( B ) , and P( A B) for easier manipulation. Use
P A B P A P B P A B

5 1 6 3 11 18

1 3

1 3

then
P B| A P( A B) P ( A)

108

Mathematics Revision Notes 2009 RI(JC) Mathematics Society

Distributions
Expectation Since the expectation is a summation, you can apply summation rules to it. Let a and b be constants, and X and Y be random variables.
E a E aX E aX E aX b bY a aE X aE X aE X b bE Y

Variance The variance is always positive. When combined, 2 variances are always added together.
Var a Var aX Var aX b 0 , since a constant has no variation. a 2 Var X a2 Var X

If X and Y are independent, Var aX bY a2 Var X b2 Var Y Standard deviation
Var 2 X

Var X
X2 2 Var X . (Example 2)

4 Var X , but Var X 1

Properties of Distributions Binomial: X ~ B n, p o

P X

x

n x p 1 p x
np

n x

(In MF15) for x

0,1, 2,..., n

o Mean, E X

o Variance, Var X np 1 p o Assumptions: n independent trials Constant probability of success Each outcome is either success or failure Occur randomly

109

p . . 1. GC syntax. o Mean. E X o Variance. 3. a. p . then press the down key until normalcdf( is highlighted. . Var X o Assumptions: Occur independently Uniform rate of occurrence Occur singly Occur randomly Normal: X ~ N . Press the [+] key to display the syntax. a is given by normalcdf X a a X b a is given by normalcdf a . b. to find the syntax for normalcdf(. b is given by normalcdf a . use CtlgHelp. [APPS] 3:CtlgHelp [ENTER] 2. 1..1. p . a 2 If X ~ Po P X P X .. the standard normal random variable. E X o Variance.1 .a . E 99. . The GC will assume that 0 and and can be omitted in the 110 . Var X GC Syntax If you forget GC syntax. 2 . E 99. a a is given by binomcdf n. P X P X a is given by binompdf n. For example.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Poisson: X ~ Po x o P X x e x! (In MF15) for x 0. If X ~ B n. a is given by poissonpdf a is given by poissoncdf .a If X ~ N P X P X P a P X P X P a . 2 o Mean. [2ND][VARS].. 2. P X P X a a X b P a X b P a X b P a For Z ~ N 0.

X is continuous and follows a normal distribution. If If If Y p is close to 1.e. approximate X to a normal distribution. p . use the value of p to decide whether to approximate to a Poisson or normal distribution. X ~ B n. i. and approximate to a Poisson distribution. such that np 5 Poisson X ~ Po np Continuity correction Poisson X ~ Po 10 If X follows a binomial distribution. 111 . Q: “Using a suitable approximation.5). define a new variable Y which has p close to 1. approximate X to a Poisson distribution. p n is large such that np 5 and n 1 p 5 Continuity correction Normal X ~ N np . Legend: “ ” denotes “approximates to” Binomial X ~ B n.…” or “Find the approximate number of…” or “Estimate…” A: Show that the original distribution fulfills the conditions for approximation. p is close to 0. p is moderate (close to 0. X follows a binomial distribution.1 ). np 1 p Normal X ~N . If X can have any integer value from 0 onwards to an upper limit n .Mathematics Revision Notes 2009 RI(JC) Mathematics Society FAQ Q: Which distribution does a random variable X follow? A: If X can have non-integer values. If X can have any integer value from 0 onwards with no upper limit. n is large ( 50 ). X follows a Poisson distribution. p is small ( 0.

2.25 . If X ~ Po X Y ~ Po 1 1 and Y ~ Po 2 2 . 2 .1 and P X x P Z X .120. then let Z . Since for Z . 2 1 . Using a suitable approximation. . 2 1 2 1 2 b~ N a b . 112 . find the probability that at least 40 of these observations are less than -3. find the probability that at least 197 of these observations are less than 7. A: and . or both are unknown. Y~N . always standardize X to Z . For Binomial or Poisson. if 1. 4. 2 2 2 2 . X use InvNorm on the GC to find an equation for . Then Z ~ N 0.460 0. x . Solve simultaneous equations if necessary. X . use GC or use the formula for P X For Normal. 200 independent observations of X are taken.460. and Y ~ N 1 1 1 2 2 . a2 b bY ~ N a . Subtraction ( X If X ~ N o o o X aX aX 1 Y ) is not allowed. e. both mean (0) and standard deviation (1) are known.01 and P X 3.120 0. a2 2 1 b2 2 2 Q: Solve for multiple unknowns. A: Let X and Y be independent random variables. If X ~ N .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Q: Sum random variables. (i) (ii) Using a Poisson approximation.g. Example 1 [N01/II/8] The random variable X has a normal distribution with P X 7. Find the standard deviation of X . 3.

If you do not specify and . Then Y ~ B(200. 7.f. Comments Since and are unknown for X . write “out of 200 2.) (working omitted) From the GC.460 2. P X 3. InvNorm can be used as and are known for Z .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution Let X ~ N ( .25) . P Z 0.67449 From the GC. (i) Let Y be the number of observations not less than 7. as with the standard 0 and normal distribution. (ii) Let W be the number of observations less than 3.01) . Remember to observations”. and standardization is necessary.) Write the system of linear equations and use PlySmlt2 on the GC to solve for and . 113 . P 200 Y P Y 3 197 0.53 (3 s.01 Once you get an equation in this form.f.120 0. as required.460 out of 200 observations. 0. Then W ~ B(200. we cannot immediately use InvNorm. x Let Z .01 is small such that np 2 5 . Similarly.f.460 3. because p must be small for a Poisson approximation.460.460 P Z 7.857 (3 s. standardize and use InvNorm for the other equation from the question. Since n 200 is large (>50) and p 0.99) .) 7. P X 7. so define the variable as W instead. 3. We define Y with “not less than” rather than “less than”. 0. 2 ) .3263 0. Z is reserved for the variable with the standard normal distribution.120 0.25 .120 3.f.460 0.460 7.67449 (5 s. the GC will assume 1 .3263 (5 s. Y ~ Po(2) approximately.) 200 Y is the number of observations less than 7.120 out of 200 observations.99 From the GC. Key in InvNorm(0.1) . then Z ~ N (0.

01 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Since n 200 is large such that np 50 5 and n 1 p 150 5 . 0. in grams.25 .5 approximately.5 .g. To approximate a binomial variable with moderate p e. 0.957 (3 s. Let Y represent the weight of a standard egg.32 Required probability is P Y E Y Var Y 4 X 5 4 X 5 50 32 4 (65) 5 16 4 25 2 4 X 5 2 19.5) (by continuity correction) 0. 42 and Y ~ N 50. 37.03 that the average weight of the eggs in one pack is more than the average weight of the eggs in the other pack by at least m grams. 5 Standard eggs are sold in packs of 12 while large eggs are sold in packs of 5. Find the value of m (to the nearest 0. not P W 40 P (W 39.5 for 37. (a) One large egg and one standard egg are chosen at random. A customer bought two packs of large eggs. in grams. Then X ~ N 65. W ~ N 50.24 114 . use a Poisson approximation.g. If p is extreme e. (i) Find the probability that the weight of a pack of standard eggs differs from twice the weight of a pack of large eggs by at most 5 grams. Remember to key in normalcdf on the GC. Comments (b) (ii) Solution (a) Let X represent the weight of a large egg. Find the probability that 4 the weight of the standard egg is more than the weight of the large egg. Example 2 [AJC05/II/27] The weights of large eggs are normally distributed with mean 65 grams and standard deviation 4 grams.1 gram) if there is a probability of 0. use a normal approximation.f.) 2 37. The weights of standard eggs are normally distributed with mean 50 grams and standard deviation 3 grams.

f. X 5 50 12 2 5 65 50 Var Y1 . X 1 . 6...1g Even though one pack is more than the other.. (iii) Using a suitable approximation. there are at least 35 squares that contain at most 5 bullfrogs each. The student wishes to model the distribution of bullfrogs in each square by using the Poisson distribution with mean .324 (3 s..19. 115 .4 5 5 X 1 .. m 5. 0... Find the probability that two of these squares contain no bullfrogs and each of the other two squares contains at least 2 bullfrogs...5 g nearest 0.f. Y12 2 X 1 . X 5 X 6 .) (b)(i) Required probability P Y1 Y2 . hence each egg is an independent observation of Y... X 5 50.. One part of the field is identified and divided into 50 equal squares. Y12 2 X 1 . Y12 2 X 1 .03 P Q m 0.24 P Y 4 X 5 0 0.C.. X 5 E Y1 . 428 5 5 5 A pack of standard eggs contains 12 different eggs. out of 50 squares...015 From G. X 5 X 6 .. Y12 2 X 1 ..... X 5 ~ N P Y1 Y2 . Y12 2 X 1 .Mathematics Revision Notes 2009 RI(JC) Mathematics Society Y 4 X~N 5 4 P Y X 5 2. Example 3 [NJC05/II/30] A Geography student is studying the distribution of bullfrogs in a large rural field where there is an average of 500 bullfrogs per 400 km2. each with sides measuring 2 km..... X 10 ~ N 0. find the probability that.. Y12 2 X 1 . use the modulus sign because either pack could be the heavier one... X 5 P 5 Y1 .....0109 3 s.. X 10 Let Q 5 5 P | Q | m 0. (ii) Four squares are selected at random. X 5 12 32 4 5 16 428 Y1 .. (i) State the value of and one assumption made in using the Poisson distribution.

try Let Y be the number of squares that contain at binomial to Poisson. Otherwise. (ii) Let X be number of bullfrogs in a square.f. Then X ~ Po 5 Required probability [P X 0 ]2 1 P X 1 2 Use P & C: 4! 2!2! 4! permutations of 4 2!2! objects where 2 objects are of type A and the other 2 are of type B. 400 Total squares in field 100.000251 3 s.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iv) Determine the approximate probability that the total number of bullfrogs in 10 randomly selected squares is between 40 and 60 (not inclusive). 0. n 1 p 35 34. try approximating P X 5 0.828 approximately.2 5.f. Solution (i) Comments Do not write the “probability” that a bullfrog lies in a square is independent. There are 0. Use the word “event”. 116 . since these most 5 bullfrogs. bullfrogs in the field is constant. 19.61596 binomial to normal. Comment on the suitability of the model in this situation.61596 Since n np P Y P Y 50 is large. Alternative assumption Area of a square is 4km2 .8 5. (iii) If np > 5. 30. 0. are the only two possible cases.798. 4 500 5 100 Assumption: The event that a bullfrog lies in a Average rate of occurrence of square is independent of other bullfrogs.141 3 s.5 (by continuity correction) Y ~ N 30. Y ~ B 50. out of 50.11. (v) The student suggests using the same model on another rural field in another country.

5) (by continuity correction) 0. Then Q ~ Po 50 Since .821 3 s. Q ~ N 50.5 Q 59. 117 .f.50 approximately. the model may not be suitable. Therefore. P 40 Q 60 P (40. (v) The number of bullfrogs may not be randomly distributed in another country.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (iv) Let Q denote number of bullfrogs in 10 squares.

…. and 60 are above 40 years old. the 3rd. 11th. Select a sample of 50 workers 200 As 4 . randomly select one worker from using a list of random 50 numbers. 7th. 40 workers are under 30 years old.e. except: groups with sample size proportional to the Sample size need not be (but can relative size of each age group. the first 4. i.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Sampling & Testing Sampling Methods Example: Say we want to select a sample of 50 workers from a company with 200 workers. If say the 3rd worker is chosen. include every 4th worker from the 3rd worker in the sample.: be) proportional Non-random samples are drawn Age Sample size from each age group 20 30 30 – 40 50 30 40 Pros Likely to give more representative Lower cost sample Less time-consuming Does not require sampling frame Cons Requires sampling frame Non-random Cannot replace unavailable sampling Likely to give less representative units sample Strata may not be clearly defined Biased as researcher may just pick members who are easier to collect data from 118 . 199th workers.e. Pros Even spread over population Cons Requires sampling frame Requires sampling frame Cannot replace unavailable Cannot replace unavailable sampling units sampling units Biased if members of the population have a periodic or cyclic pattern Time-consuming to locate selected units Stratified sampling Quota sampling Random Non-random Method Select random samples from the age Same as stratified. Simple random sampling Systematic sampling Random Random Method Assign each worker a number from 1 to 200. 100 are 30 to 40 years old. i.

Mathematics Revision Notes 2009 RI(JC) Mathematics Society FAQ (Sampling) Questions on sampling methods Use the context of the question in your answer. 2 1 n 1 x 2 x n 2 1 n 1 x x 2 Let y x b. Identify type of sampling method used Explain why: o Method is inadequate: list cons o Method is biased/unbiased o Sample is likely to be representative or not Suggest a better method if required Describe your suggested method 1 x x n s 2 Given or x x and x . y x a x a y2 b n(b a ) x b 2 a 2 Find y and s y (see above). x Find x and s 2 . and population distribution is unknown sample is small ( n 50 ). 2 given ( x a) and x b . and 2 is unknown 2 is unknown sample is large ( n 50 ). Find unbiased estimate x of population mean Find unbiased estimate s of population variance 2 Find x and s 2 . 2 y b . and population distribution is unknown 119 . 2 x2 . L2 Testing Do this: Use t -test instead of z -test Use s 2 instead of 2 Use Central Limit Theorem Assume population distribution is normal Only if: sample is small ( n 50 ). given a list of data. Individual data o Values in L1 o Key in: 1-Var Stats L1 Grouped data o Values in L1 . square the value to get s x . frequencies in L2 o Key in: 1-Var Stats L1 . sx 2 sy 2 GC gives sx .

/ n increases. However. the t -test would be lead to a conclusion of H 0 being rejected for more values of x . write “approximately” after the distribution of X . Central Limit Theorem is not required. / n increases. or population mean on outcome x P Z x 0 / n . State the assumptions made. and no assumptions about its distribution are needed. FAQ (Testing) Explain what is meant by ‘a % significance level’ in the context of the question A % significance level means that the probability of 100 concluding that H 0 is rejected given that H 0 is true is .75kg is 0. if any If the sample is large ( n 50 ): o Central Limit Theorem holds.Mathematics Revision Notes 2009 RI(JC) Mathematics Society If the sample is large. given the same significance level. p value increases x P Z x 0 If p value o If x P X 0 / n . p value decreases 120 . and 2 is unknown. hence the population follows a normal distribution approximately. If p value o If x P X 0 Compare effect of change in significance level. sample mean. If the sample is small ( n 50 ): o Assume the distribution of the population is normal Compare outcomes of z -test and t -test.05. Example: A 5% significance level means that the probability of concluding that the mean mass of a chicken is greater than 1. given same data Since the graph of the t -distribution has “thicker tails” than that of the z -distribution about 0 .75kg when it is actually 1. population is normal. since s 2 must be used as an unbiased estimate of 2 .

5. 14. x .1.05 ) o If H 0 is rejected. Based on this sample. 14. only if the mean 0 or variance is unknown. 2P X x x State the condition on the p -value for the outcome (For e. value 0. Find the range of values for the average tar content per cigarette that the manufacturer should claim in order that it will not be rejected as an underestimate at the 1% level of significance if a z -test is carried out.2.9. s . or n 0 Use InvNorm to find the range of . 14. a 5% significance level. 14.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Find the range of . or n such that H 0 is rejected (or not rejected) at a % significance level. the surveyor took another random sample of 50 ‘Genius’ cigarettes and measured the tar content x in mg/cig. (i) (ii) Find unbiased estimates of the population mean and variance of the tar content in mg/cig.1.2.2455 . or satisfies the condition on the p -value Conclude in the context of the question that Example [TJC07/Prelim/II/9 (Modified)] The manufacturer of the ‘Genius’ brand of cigarettes claimed that the average tar content of a ‘Genius’ cigarette is not more than 14mg per cigarette (mg/cig). A surveyor from a rival company took a random sample of 8 ‘Genius’ cigarettes and lab measurements of the tar content yielded 14. i. The following results were obtained: 2 x 10 208.4 mg/cig.g. p value value Standardize. To confirm his findings. when finding the range of s2 of X n 0 .e.2515 . s . 14. 0 . 121 . 2P X . x .0. p value value o If H 0 is not rejected. Will he be able to reject the manufacturer’s claim at 5% level of significance? State any necessary assumption made by the surveyor in carrying out the test. 13. 14. x 14 5. Find an expression for the p -value o 1-tailed: P X o 2-tailed: If x If x x or P X x 0 0 .

Specify its units.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution Let X be the tar content of a cigarette in mg/cig. Since p value 0.) 4. surveyor will be able to reject H 0 and conclude that there is sufficient evidence at the 5% significance level that X 14 and the manufacturer’s claim is false. and significance level. the 5. x and s 2 will appear on the GC screen with the p -value. it must be assumed to be normal. Since n 8 is small. ~ t ( n 1) . which can have many values. T X μ0 .0206 (3 s. p value 0. 1. the surveyor must assume that X follows a normal distribution. x is the mean of a particular sample. . Decide on the test statistic (t or z) to be used and determine its distribution. Under H 0 . 3. State the null hypothesis H 0 and the alternative hypothesis H1 . since σ is s/ n σ/ n unknown and we estimated it with s . Write down 1-tailed or 2-tailed test. 0. usually given in the question. the level of significance α% . H0 : 14 vs H1 : 14 Comments Always define X if the question does not. Use the GC to calculate p -value.f. not X μ0 X is a random variable representing the sample mean.0206 0. so we do not need to calculate them. Conclude in the context of the question. Since the sample is small and the distribution of the population is unknown.19821 .175 . if any (mg/cig). Perform a 1-tailed test at the 5% 2.05 . 0 14 . Using a t -test. █ 122 . T where x s X 0 s/ n 14. n 8 .

2515 5.165 (5s f. Since μ 0 is unknown for X . and if it μ μ were “rejected as an underestimate. InvNorm can be used as μ and σ are known for Z . Perform a 1-tailed test at the 1% significance level.165 0 P Z x 0 0. n 50 . first. e.2455 Always let y be the expression which is squared in the summary statistics. we cannot immediately use InvNorm. x 14 8.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (i) Let y x 14 .01 Once you get an equation in this form.079260 / 50 Solving. y y2 x 14 x 10 50 4 2 8. let 2 x 14 is given.165 P Z x 0 s/ n 0.25152 5. . y x 14 when Give values which will be used in later parts of the question to 5 s.) 0.1 (3 s.01 . .) s Translate the question into mathematical terms: What is the range of μ 0 such that H 0 is not rejected when a z-test is carried out at the 1% significance level? The manufacturer’s claim is H 0 : μ μ 0 . Then Z ~ N (0. then to 3 s. . s2 X ~ N 0.f.g. p value 0. 0 Let Z X s/ n If H 0 is not rejected.2455 50 1 50 █ 0. where n x 14. s 0.) 14. by the Central Limit Theorem since n 50 is large.) 0.)█ 2 sx 1 8.079260 (5s f. (ii) H0 : 0 vs H1 : 0. Under H 0 . .f. H1 : 0 would be true. s/ n From the GC. and standardization is necessary.079260 .1) . approximately.3263 (5 s. P X 14.0793 (3 f . 0 14. .f.99 2.165 . 14.f.2 (3s f.2515 x y 14 14 50 14.) █ 123 .

Mathematics Revision Notes 2009 RI(JC) Mathematics Society Correlation & Regression Product Moment Correlation Coefficient (PMCC) r>0 Positive Linear Correlation r<0 Negative linear correlation r ≈ 0 No linear correlation (Note that no linear correlation does not imply no relation) The closer the absolute value of r is to 1.g. points on a parabola can be chosen to give a PPMC of close to zero Therefore.g. 124 . the stronger the linear correlation between 2 variables. Limitations of PMCC Does not necessarily prove that there is a linear relationship between two variables o e. there may be an outlier Unable to prove any non-linear relationship between variables using the coefficient o e. use BOTH the scatter diagram and the r-value together when making judgements/statements/comparisons.

Ensure the relative position of the data points is correct. If the data is far away from the origin. GC Tips & FAQ 1. How do I key in the data set into the GC? Key STAT EDIT Type in your data 125 . Display the origin at the correct relative position. use a jagged line to indicate an omitted range of values.) Scatter Diagrams Label each axis with the variable and its units.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Relevant Formulae (In MF15) Estimated Product Moment Correlation Coefficient r ( x x )( y ( x x )2 y) (y y )2 x 2 xy x n 2 x n y y y n 2 2 Estimated regression line of y on x y y b(x x) b ( x x)( y y) ( x x) 2 Although not in MF15. the formula for the regression line of x on y can be obtained by swapping x with y in the above formula for the regression line of y on x) Linear Regression The least square regression line is the line which minimises the sum of squares of the residuals. (Note that the a residual is the difference between the observed & predicted values.

When a question requires me to plot ln y against x. Key all your values into the lists. with x values in L1. we would look the amount of extrapolation. Otherwise. then that estimate is reliable. If it is close to the experimental range of values. but the question requires you to plot ln y against x.Mathematics Revision Notes 2009 RI(JC) Mathematics Society 2. an interpolation would be very reliable. key ln(L2) STO L3 3. we look at whether it is an interpolation or an extrapolation. If it is close to 1 or – 1. Then. if the data given is a list of values of y. and their percentage score for the mathematics final examination (y). In the case of an extrapolation. The results are tabulated below: x y 20 16 35 25 45 35 60 50 70 60 80 65 100 70 110 75 120 80 140 85 126 . What is the significance of the intersection point of regression lines of x on y and y on x? The intersection point of regression lines of x on y and y on x is ( ̅ . Example 1 [SAJC08/Prelim/II/6] Each of a random sample of 10 students are asked about the average number of minutes spent on doing mathematics tutorials in a week (x). but it only provides data set for x and y. y values are in L2. we look at the r-value. ). For example. 4. then the estimate would be reliable. you can do so using GC. Assuming r is close to 1 or – 1. y values in L2. Next. it is not reliable. How do you assess whether an estimate is reliable? First. Note that you should use the scatter plot in line with the r-value. what can I do other than manually evaluating each value of ln y? For example.

591x 10. The question is indirectly saying that y is the dependent variable and x is dependent..1x 16.6 0.] (ii) Find the linear product moment correlation coefficient between y and x. linear regression should be done with L1. x and y. Example 2 [HCI08/Prelim/II/8 (Modified)] A teacher decided to investigate the association between students’ performances in Mathematics and Physics. f . in Mathematics and Physics tests respectively.C values calculated earlier r 0. y 15.5 0. the estimate is not reliable as 10 minutes is well outside the given range of x values so the approximation would be unreliable. She selected 5 students at random and recorded their scores.0 3s. for this set of data. Using G. Remember that the intercept of the regression lines (y on x and x on y) is x.5 0. y 0. Use the G. L2 where L1 is the independent variable and L2 is the dependent variable. Combination of concepts from sampling x 5 25 125 and 5 21 105 127 . f . Remember to turn on the diagnostic function.6 0. the equation of the regression line of y on x was y 18. and comment on the relationship between x and y. [Whether you use LinReg(ax+b) or LinReg(a+bx).972 3s . Key all the values into the lists in GC. (iii) Use the appropriate regression line to estimate the percentage score of a student who spends 10 minutes doing mathematics tutorial in a week.4 y x 25 and y x 125 and y x 21 y y 105 . From G. f . She found that.0 3s. Show that 18.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (i) Find the equation of the regression line of y on x.C.C.1x and the equation of the regression line of x on y was x 16.9 However.4 y . y 0.591x 10. y . Comment on the reliability of the estimate.

07 90 1. y 18. l.1 x When x 26. y 21(nearest integer) Since x is given.10 15 0. After conducting the experiment. Determine which model is more appropriate. A:T a b ln l a bl B :T 2 Calculate the product moment correlation coefficients for both models. (iii) Comment on the reliability of your estimate.45 135 2. The student proposes two models. it is close to 0. [3] (iii) Using the model determined in part (ii). The estimate is not reliable because | r | is small. he obtained the following set of data: l/cm T/s (i) (ii) 150 2. and give a reason for the use of the chosen equation. estimate the value of l when T = 3s to 1 decimal place. hence the linear correlation between x and y is very weak.e. y is the dependent variable and therefore we use the regression line of y on x.35 30 1. [3] (iv) Find a value of l and its corresponding value of T such that the equation of the regression line for the chosen model will remain the same after the addition of this pair of values.779 [2] Obtain the scatter plot of this set of data.22 105 2. T. and the corresponding period.Mathematics Revision Notes 2009 RI(JC) Mathematics Society (i) The Physics score of a sixth student was mislaid but his Mathematics score was known to be 26.91 75 1. indicating that the linear model is not suitable. giving your answer to 4 decimal places. Example 3 [NYJC08/Prelim/II/11] A student wishes to determine the relationship between the length of a pendulum.56 45 1.74 60 1. i.31 120 2. [3] 128 .5 0. Use the appropriate line of regression to estimate this student’s Physics score. Comment on the suitability of this method.

to evaluate the values. T T 1.Mathematics Revision Notes 2009 RI(JC) Mathematics Society Solution (i) Comments Note that your scatter plot must be to scale. Hence it is not a good estimate of l.9 (1d. the value of l is 223. (iii) Since T is given. Although r is close to 1. (4 d.p..f. it is NOT the regression line of T2 From G. This means that your points should be relatively correct to each other. on the length of the pendulum. Axis should be labelled correctly. Since the value of r for model B is closer to 1 as look at r values. compared to model A. The intersection of the two lines will obviously be the point you are looking for.C. we should use the regression line Even though it is in the Physics mindset that the period depends of l on T2. Use G.p.C.) Be careful as question asks for 4 decimal places.9996.f. 129 .. (iv) The values to be used are l and T 2 . Note that you can plot the regression lines for both l on T2 and T2 on l. plot the regression line. the value of T = 3s lie Be aware that model B uses T2 well outside of the range of the T values used to instead of T.C. (ii) Using a G. (4 d.) r for model B is 0. model B is more appropriate.5 and T 2 3.9871.3301 5 s. r for model A is 0. l l 82.) on l.82 3 s.p. For appropriateness of model.

Sign up to vote on this title
UsefulNot useful