Logic Lecture Slides

Logic and Discrete Mathematics
Section 6.7
Recurrence Relations and Their Solution
Slides version: January 2015

Definition
A recurrence relation for a sequence a0 , a1 , a2 , . . . is a formula

giving an in terms of one or more of the terms preceding it (i.e
in terms of a0 , a1 , a2 , . . . , an−1 ), for any n greater than some
initial integer k .
The values of the first few terms of the sequence needed to

start computing with the recurrence relation are called the initial
conditions.
The Fibonacci sequence
Let F0 = 1, F1 = 1, F2 = 2, F3 = 3, F4 = 5, etc.
Let F0 = 1, F1 = 1, F2 = 2, F3 = 3, F4 = 5, etc.
We find that
Fn = Fn−1 + Fn−2
for all n ≥ 2.
The necessary initial conditions are F0 = 1, F1 = 1.
Let F0 = 1, F1 = 1, F2 = 2, F3 = 3, F4 = 5, etc.
We find that
Fn = Fn−1 + Fn−2
for all n ≥ 2.
The necessary initial conditions are F0 = 1, F1 = 1.
This sequence is known as the Fibonacci sequence and

the terms of the sequence are known as Fibonacci numbers.
The Lucas Numbers
The sequence of Lucas numbers
2, 1, 3, 4, 7, 11, . . .
has the initial conditions L1 = 2 and L2 = 1

and the recursion Ln = Ln−1 + Ln−2 .
The Lucas Numbers
The sequence of Lucas numbers
2, 1, 3, 4, 7, 11, . . .
has the initial conditions L1 = 2 and L2 = 1

and the recursion Ln = Ln−1 + Ln−2 .
Note that this is the same recursion relation as for the

Fibonacci numbers.
The Catalan Numbers
The sequence of Catalan numbers is given recursively by
Cn+1 = C1 Cn + C2 Cn−1 + · · · + Ci Cn+1−i + · · · + Cn C1 ,
with C0 = C1 = 1.
The first eleven terms are
1, 1, 2, 5, 14, 42, 132, 429, 1430, 4862, 16796
Cn can also be given non-recursively as follows:

(2n)! 1 2n
Cn+1 = = , n = 0, 1, 2, . . . .
(n + 1)!(n!) n+1 n
Significance of The Catalan Numbers (1)
The number of ways to cut an (n + 2)-sided polygon into
triangles by connecting its corners with n − 1 non-intersecting
straight line segments is equal to Cn .
Figure : Dividing the square.
Figure : Dividing the pentagon.

Significance of The Catalan Numbers (2)
The number of complete binary trees with n leaves is equal to
Cn−1 .
One leaf: Two leaves:
Three leaves:
Four leaves:
Figure : Complete binary trees with one, two, three and four leaves
Definitions
Definition
A recurrence relation of the form
an = c 1 (n)an−1 +c 2 (n)an−2 +c 3 (n)an−3 + · · · + c m (n)an−m +f (n)
where c1 , c2 , . . . , cm and f are functions of n, is called a linear

recurrence relation of order m.
Definitions
Definition

• The relation is homogeneous if f (n) = 0 for all n

Definitions
Definition


• The relation has constant coefficients and is of order m
when the ci (n)’s are all constant and cm 6= 0.
Definitions
Definition


• The relation has constant coefficients and is of order m
when the ci (n)’s are all constant and cm 6= 0.
We consider recurrence relations that have the form
an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

Characteristic Equation and Roots
Definition
x m − c1 x m−1 − c2 x m−2 − · · · − cm = 0
is called the characteristic equation of the recurrence relation

Definition
x m − c1 x m−1 − c2 x m−2 − · · · − cm = 0
• The characteristic equation has m (not necessarily distinct)

roots, r1 , r2 , r3 , . . . rm , referred to as characteristic roots.
Definition
x m − c1 x m−1 − c2 x m−2 − · · · − cm = 0

• These roots may be real or complex.
Definition
x m − c1 x m−1 − c2 x m−2 − · · · − cm = 0

• These roots may be real or complex.
• As cm 6= 0, they are all different from 0.
Theorem: Solutions
Theorem
The sequence an = r n is a solution of the recurrence
if and only if r is a root of the characteristic equation
x m − c1 x m−1 − c2 x m−2 − · · · − cm = 0
Hence, for each characteristic root r , an = r n is a solution.

These are called fundamental solutions.
Operations on Solutions
Let (an ) and (bn ) be sequences.
Addition of the sequences: (an ) + (bn ) is the sequence

(an + bn ),
i.e.
(a1 , a2 , a3 , . . .) + (b1 , b2 , b3 , . . .) = (a1 + b1 , a2 + b2 , a3 + b3 , . . .).

Operations on Solutions
Let (an ) and (bn ) be sequences.
Addition of the sequences: (an ) + (bn ) is the sequence

(an + bn ),
i.e.
(a1 , a2 , a3 , . . .) + (b1 , b2 , b3 , . . .) = (a1 + b1 , a2 + b2 , a3 + b3 , . . .).
A constant multiple, k (an ), of a sequence (an ) is
k (an ) = (ka1 , ka2 , ka3 , . . .).

Theorem: Linear Combinations of Solutions
Theorem
If (an ) and (bn ) satisfy a homogeneous linear recurrence
relation with constant coefficients, then (an ) + (bn ) and k (an )
satisfy the same recurrence relation.
General Solution — Distinct Roots
Theorem
Let the roots r1 , r2 , r3 , . . . , rm of the characteristic equation be
distinct. Then an = k1 r1n + k2 r2n + k3 r3n + · · · km rmn is a general
solution of the homogeneous linear recurrence relation, i.e.
every solution can be expressed in this form for some choice of
constants k1 , k2 , k3 . . . , km .
I.e. the fundamental solutions of a recurrence relation form a
basis of the vector space of solutions of that relation.
Solution Procedure — Distinct Roots
1. Write down the characteristic equation for the relation.


2. Find the m roots of the characteristic equation. If the roots
are distinct we obtain m fundamental solutions.

3. Take an arbitrary linear combination of the fundamental
solutions to obtain a general solution.

3. Take an arbitrary linear combination of the fundamental
solutions to obtain a general solution.
4. Using the initial conditions, construct and solve a system of
equations to obtain the specific solution.
Examples: Distinct Roots
Solve the following recurrence relations:

1. an = 5an−1 − 6an−2 for n > 2, and with a1 = 2 and a2 = 10.

2. an = −an−1 + 12an−2 for n > 1, and with a0 = 1 and
a1 = 2.

2. an = −an−1 + 12an−2 for n > 1, and with a0 = 1 and
a1 = 2.
3. an = −4an−1 + 21an−2 for for n > 2, and with a1 = 2 and
a2 = 3.
Complex Roots
Theorem
If the roots of a characteristic equation of a homogeneous
linear recurrence relation of degree two are complex numbers
a + bi and a − bi (i.e. the roots are complex conjugates) then
the general solution of the recurrence relation is given by
an = ρn (k1 cos nθ + k2 sin nθ)

√ a
where ρ = a2 + b2 , cos θ = ρ and sin θ = bρ .
Complex Roots — Examples
Solve the following recurrence relation:

√
an = 2 2an−1 − 4an−2 for n > 1, with a0 = 1 and a1 = 2.
Multiplicity of Roots
A characteristic root r has multiplicity j if it appears as a root of

the characteristic equation
x m − c1 x m−1 − c2 x m−2 − · · · − cm = 0
exactly j times.
This is equivalent to (x − r )j being a divisor of the polynomial
x m − c1 x m−1 − c2 x m−2 − · · · − cm but (x − r )j+1 not being a
divisor.
Repeated Roots — Fundamental Solutions
If r is a characteristic root of multiplicity j, then

an = r n , an = nr n , an = n2 r n , . . . an = nj−1 r n are linearly
independent and can be taken as fundamental solutions.
Repeated Roots — Example
Solve the recurrence relation given by

an = 6an−1 − 9an−2 with a0 = 1 and a1 = 4.

Logic Lecture Slides

Uploaded by

Copyright:

Available Formats

Logic Lecture Slides

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Logic Lecture Slides

Uploaded by

Copyright:

Available Formats

Logic and Discrete Mathematics

Slides version: January 2015

A recurrence relation for a sequence a0 , a1 , a2 , . . . is a formula

The values of the first few terms of the sequence needed to

This sequence is known as the Fibonacci sequence and

The sequence of Lucas numbers

has the initial conditions L1 = 2 and L2 = 1

The sequence of Lucas numbers

has the initial conditions L1 = 2 and L2 = 1

Note that this is the same recursion relation as for the

The sequence of Catalan numbers is given recursively by

Cn+1 = C1 Cn + C2 Cn−1 + · · · + Ci Cn+1−i + · · · + Cn C1 ,

1, 1, 2, 5, 14, 42, 132, 429, 1430, 4862, 16796

Cn can also be given non-recursively as follows:

Figure : Dividing the square.

Figure : Dividing the pentagon.

an = c 1 (n)an−1 +c 2 (n)an−2 +c 3 (n)an−3 + · · · + c m (n)an−m +f (n)

where c1 , c2 , . . . , cm and f are functions of n, is called a linear

an = c 1 (n)an−1 +c 2 (n)an−2 +c 3 (n)an−3 + · · · + c m (n)an−m +f (n)

where c1 , c2 , . . . , cm and f are functions of n, is called a linear

• The relation is homogeneous if f (n) = 0 for all n

an = c 1 (n)an−1 +c 2 (n)an−2 +c 3 (n)an−3 + · · · + c m (n)an−m +f (n)

where c1 , c2 , . . . , cm and f are functions of n, is called a linear

• The relation is homogeneous if f (n) = 0 for all n

an = c 1 (n)an−1 +c 2 (n)an−2 +c 3 (n)an−3 + · · · + c m (n)an−m +f (n)

where c1 , c2 , . . . , cm and f are functions of n, is called a linear

• The relation is homogeneous if f (n) = 0 for all n

an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

• The characteristic equation has m (not necessarily distinct)

an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

• The characteristic equation has m (not necessarily distinct)

an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

• The characteristic equation has m (not necessarily distinct)

an = c1 an−1 + c2 an−2 + c3 an−3 + · · · + cm an−m , cm 6= 0.

if and only if r is a root of the characteristic equation

Hence, for each characteristic root r , an = r n is a solution.

Let (an ) and (bn ) be sequences.

Addition of the sequences: (an ) + (bn ) is the sequence

(a1 , a2 , a3 , . . .) + (b1 , b2 , b3 , . . .) = (a1 + b1 , a2 + b2 , a3 + b3 , . . .).

Let (an ) and (bn ) be sequences.

Addition of the sequences: (an ) + (bn ) is the sequence

(a1 , a2 , a3 , . . .) + (b1 , b2 , b3 , . . .) = (a1 + b1 , a2 + b2 , a3 + b3 , . . .).

A constant multiple, k (an ), of a sequence (an ) is

k (an ) = (ka1 , ka2 , ka3 , . . .).

1. Write down the characteristic equation for the relation.

1. Write down the characteristic equation for the relation.

1. Write down the characteristic equation for the relation.

1. Write down the characteristic equation for the relation.

Solve the following recurrence relations:

Solve the following recurrence relations:

Solve the following recurrence relations:

an = ρn (k1 cos nθ + k2 sin nθ)

Solve the following recurrence relation:

A characteristic root r has multiplicity j if it appears as a root of

If r is a characteristic root of multiplicity j, then

Solve the recurrence relation given by

You might also like