Arab Academy for Science, Technology & Maritime

Transport

Math 5 (BA323)
(Gamma and Beta Functions)

Fall 2020-2021
Chapter 2 Special Functions

Chapter 2

Gamma Function

Introduction

Gamma function is used to generalize the factorial function to nonintegral values, introduced by the
Swiss mathematician Leonhard Euler in the 18th century. The gamma function belongs to the category of
the special functions. It also appears in various area as asymptotic series, definite integration and
hypergeometric series.

Definitions of the Gamma function

To extend the factorial to any real number 𝑥 > 0 (whether or not 𝑥 is a whole number), the gamma
function is defined as

∞
Γ(𝑥) = ∫ 𝑡 𝑥−1 𝑒 −𝑡 𝑑𝑡.
0

By using techniques of integration, it can be shown that

∞
∞
Γ(1) = ∫ 𝑒 −𝑡 𝑑𝑡 = −𝑒 −𝑡 | 0 = 0 + 1 = 1 = 0!
0

Similarly, using a technique from calculus known as integration by parts, it can be proved that the gamma
function has the following recursive property: if 𝑥 > 0, then

∞
∞
Γ(2) = ∫ 𝑡 𝑒 −𝑡 𝑑𝑡 = (−𝑡𝑒 −𝑡 − 𝑒 −𝑡 ) | 0 = 0 + 1 = 1 = 1!,
0

∞
∞
Γ(3) = ∫ 𝑡 2 𝑒 −𝑡 𝑑𝑡 = (−𝑡 2 𝑒 −𝑡 − 2𝑡𝑒 −𝑡 − 2𝑒 −𝑡 ) | 0 = 0 + 2 = 2 = 2!,
0

-1-
Chapter 2 Special Functions

and
∞
∞
Γ(4) = ∫ 𝑡 3 𝑒 −𝑡 𝑑𝑡 = (−𝑡 3 𝑒 −𝑡 − 3𝑡 2 𝑒 −𝑡 − 6𝑡𝑒 −𝑡 − 6𝑒 −𝑡 ) | 0 = 0 + 6 = 6 = 3!,
0

and so on. Generally, if 𝑥 is a natural number (1, 2, 3, … ), then

Γ(𝑥) = (𝑥 − 1)!

While the gamma function behaves like a factorial for natural numbers (a discrete set), its extension to the
positive real numbers (a continuous set) makes it useful for modeling situations involving continuous change,
with important applications to calculus, differential equations, complex analysis, and statistics.

Now, we want to prove that

Γ(1/2) = √𝜋.

To prove that, we remember the Gamma function

∞
Γ(𝑥) = ∫ 𝑡 𝑥−1 𝑒 −𝑡 𝑑𝑡.
0
2
Making the substitution 𝑡 = 𝑢 gives the equivalent expression
∞
2
Γ(𝑥) = 2 ∫ 𝑢2𝑥−1 𝑒 −𝑢 𝑑𝑢.
0
1 1
A special value of the gamma function can be derived when 2𝑥 − 1 = 0 (𝑥 = 2). When (𝑥 = 2) 2 , Γ(1/2)
simplifies as
∞
2
Γ(1/2) = 2 ∫ 𝑒 −𝑢 𝑑𝑢.
0
To derive the value for Γ(1/2), the following steps are used. First, the value of Γ(1/2) is squared. Second,
the squared value is rewritten as a double integral. Third, the double integral is evaluated by transforming to
polar coordinates. Fourth, the Γ(1/2) is explicitly solved for. First, square the value for Γ(1/2) and rewrite
as a double integral.
1 2 ∞
2
∞
2
∞ ∞
2 2
[Γ ( )] = Γ(1/2)Γ(1/2) = [2 ∫ 𝑒 −𝑢 𝑑𝑢 ] [2 ∫ 𝑒 −𝑣 𝑑𝑣] = 4 ∫ ∫ 𝑒 −𝑢 −𝑣 𝑑𝑣 𝑑𝑢.
2 0 0 0 0

The region 𝑅 which defines the first quadrant, is the region of integration. The polar transformation 𝑢 𝑟 cos 𝜃,
𝑣 = 𝑟 sin 𝜃 will transform the integral problem from cartesian coordinates to polar coordinates, (𝑟, 𝜃). These
𝜋
new variables will range from 0 ≤ 𝑟 ≤ ∞ and 0 ≤ 𝜃 ≤ 2 for the first quadrant.

Hence,
𝜋 𝜋
1 2 ∞ ∞
2 2 2
∞
2 2 1 ∞
[Γ ( )] = 4 ∫ ∫ 𝑒 −𝑢 −𝑣 𝑑𝑣 𝑑𝑢 = 4 ∫ ∫ 𝑒 −𝑟 𝑟 𝑑𝑟 𝑑𝜃 = 4 ∫ (− 𝑒 −𝑡 ) | 0 𝑑𝜃
2 0 0 0 0 0 2
𝜋 𝜋
2
= 2 ∫ 𝑑𝜃 = 2𝜃 | 2 = 𝜋.
0 0
Hence,

1 2 1
[Γ ( )] = 𝜋 ⇒ Γ ( ) = √𝜋
2 2
-2-
Chapter 2 Special Functions

Also, we can prove that

Γ(𝑥 + 1) = 𝑥 Γ(𝑥).
The Gamma function is
∞
Γ(𝑥) = ∫ 𝑡 𝑥−1 𝑒 −𝑡 𝑑𝑡,
0
then,
∞
Γ(𝑥 + 1) = ∫ 𝑡 𝑥 𝑒 −𝑡 𝑑𝑡.
0
by applying the integration by parts yields,
∞ ∞ ∞
∞
Γ(𝑥 + 1) = ∫ 𝑡 𝑥 𝑒 −𝑡 𝑑𝑡 = (−𝑡 𝑥 𝑒 −𝑡 ) | 0 + 𝑥 ∫ 𝑡 𝑥−1 𝑒 −𝑡 𝑑𝑡 = 𝑥 ∫ 𝑡 𝑥−1 𝑒 −𝑡 𝑑𝑡 = 𝑥 Γ(𝑥)
0 0 0

Graph and Table of the Gamma function 𝚪(𝒙) in a real domain.

The following graph represents the Gamma function in real value of 𝑥. Remain that the value of Gamma
function at zero an negative integer equal ±∞, or

Γ(0) = Γ(−1) = Γ(−2) = ⋯ = ±∞

Because Γ(𝑥 + 1) = 𝑥 Γ(𝑥), we can write

Γ(𝑥 + 1)
Γ(𝑥) =
𝑥

So,
Γ(1) Γ(0) Γ(−1)
Γ(0) = = ±∞, Γ(−1) = = ±∞, Γ(−2) = = ±∞, ⋯
0 −1 −2

-3-
Chapter 2 Special Functions

The following table shows Gamma function on the interval [1, 2].

Gamma Table 𝚪(𝒙), 𝟏 ≤ 𝒙 ≤ 𝟐

𝚪(𝒙) 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
1.0 1.00000 0.99432 0.98884 0.9835 0.97843 0.9735 0.96874 0.96415 0.95972 0.95545
1.1 0.95135 0.94739 0.9435 0.93993 0.93641 0.93304 0.9298 0.92669 0.92372 0.92088
1.2 0.91816 0.91557 0.91310 0.91075 0.90852 0.90640 0.90439 0.90250 0.90071 0.89904
1.3 0.89747 0.89600 0.89464 0.89337 0.89221 0.89115 0.89018 0.88931 0.88853 0.88785
1.4 0.88726 0.88676 0.88635 0.88603 0.8858 0.88566 0.88560 0.88563 0.88574 0.88594
1.5 0.88622 0.88659 0.88703 0.88756 0.88817 0.88886 0.88963 0.89049 0.89142 0.89242
1.6 0.89351 0.89468 0.89592 0.89724 0.89864 0.90011 0.90166 0.90329 0.90500 0.90678
1.7 0.90863 0.91057 0.91258 0.91466 0.91682 0.91906 0.92137 0.92376 0.92622 0.92876
1.8 0.93138 0.93407 0.93684 0.93969 0.94261 0.94561 0.94868 0.95184 0.95507 0.95837
1.9 0.96176 0.96523 0.96877 0.97239 0.97609 0.97988 0.98374 0.98768 0.99170 0.99581

Example 1 Find the values of the following Gamma functions

𝟑 𝟑 𝟓 𝟓
𝚪 (𝟐) , 𝚪 (− 𝟐) , 𝚪 (𝟐), 𝚪 (− 𝟐) , 𝚪(𝟑. 𝟑𝟓), 𝚪(−𝟐. 𝟓𝟐), 𝚪(𝟓. 𝟕𝟗), 𝚪(−𝟏. 𝟓𝟒),
Solution:

3 1 1 1
Γ ( ) = Γ ( ) = √𝜋.
2 2 2 2

1 1
3 Γ (− 2) Γ (2) √𝜋 4 √ 𝜋
Γ (− ) = = = = .
2 3 3 1 3 3
−2 −2 × −2 4

5 3 3 3 1 1 3 1 3
Γ ( ) = Γ ( ) = × Γ ( ) = × √𝜋 = √𝜋.
2 2 2 2 2 2 2 2 4

3 1 1
5 Γ (− 2) Γ (− 2) Γ (2) √𝜋 8√𝜋
Γ (− ) = = = = =− .
2 5 5 3 5 3 1 15 15
−2 −2 × −2 −2 × −2 × −2 − 8

Γ(3.35) = 2.35 Γ(2.35) = 2.35 × 1.35 Γ(1.35) = 2.35 × 1.35 × 0.89115 (from Gamma Table)
= 2.35 Γ(2.35) = 2.8271.

Γ(−1.52) Γ(−0.52) Γ(0.48)

Γ(−2.52) = = =
−2.52 −2.52 × −1.52 −2.52 × −1.52 × −0.52
Γ(1.48) 0.88574
= = (from Gamma Table)
−2.52 × −1.52 × −0.52 × 0.48 −0.95606

= 2.35 Γ(2.35) = 2.8271.

-4-
Chapter 2 Special Functions

Γ(5.79) = 4.79 Γ(4.79) = 4.79 × 3.79 Γ(3.79) = 4.79 × 3.79 × 2.79 Γ(2.79)
= 4.79 × 3.79 × 2.79 Γ(2.79) = 4.79 × 3.79 × 2.79 × 1.79 Γ(1.79)
= 4.79 × 3.79 × 2.79 × 1.79 × 0.92876 (from Gamma Table)
= 84.2045.

Γ(−0.54) Γ(0.46) Γ(1.46)

Γ(−1.54) = = ==
−1.54 −1.54 × −0.54 −1.54 × −0.54 × 0.46
0.88574
= (from Gamma Table)
0.382536

= 2.31508.

Example 2 Find the following integrals using Gamma Function

∞ ∞ ∞ 𝟑
(𝐢) ∫𝟎 𝒙𝟓 𝒆−𝒙 𝒅𝒙 (𝐢𝐢) ∫𝟎 𝒙𝟑 𝒆𝟏𝟎−𝟐𝒙 𝒅𝒙 (𝐢𝐢𝐢) ∫𝟎 𝟏𝟎−𝟐𝒙 𝒅𝒙
Solution:
(i)
∞
∫ 𝑥 5 𝑒 −𝑥 𝑑𝑥 = Γ(6) = 5! = 120.
0

(ii)
∞ ∞
∫ 𝑥 3 𝑒 10−2𝑥 𝑑𝑥 = 𝑒 10 ∫ 𝑥 3 𝑒 −2𝑥 𝑑𝑥
0 0

To convert the resulted integral to Gamma function, we use the substitution

𝑡 1
𝑡 = 2𝑥 ⇒ 𝑥 = ⇒ 𝑑𝑥 = 𝑑𝑡
2 2
At 𝑥 = 0 ⇒ 𝑡 = 0, at 𝑥 = ∞ ⇒ 𝑡 = ∞
So, we can write the previous integral as

∞ ∞
𝑒 10 ∞ 𝑡 3 −𝑡 𝑒 10 ∞ 3 −𝑡 𝑒 10
∫ 𝑥 3 𝑒 10−2𝑥 𝑑𝑥 = 𝑒 10 ∫ 𝑥 3 𝑒 −2𝑥 𝑑𝑥 = ∫ ( ) 𝑒 𝑑𝑡 = ∫ 𝑡 𝑒 𝑑𝑡 = Γ(4)
0 0 2 0 2 16 0 16

𝑒 10 3𝑒 10
= × 3! = .
16 8
(iii)
∞ ∞ 3
∞
−2𝑥 3 ln 10−2𝑥 3 ln 10
∫ 10 𝑑𝑥 = ∫ 𝑒 𝑑𝑥 = ∫ 𝑒 −2𝑥 𝑑𝑥
0 0 0

To convert the resulted integral to Gamma function, we use the substitution

2
−
3
𝑡 3 1 𝑡 3
𝑡 = 2𝑥 ln 10 ⇒ 𝑥 = √ ⇒ 𝑑𝑥 = ( ) 𝑑𝑡
2 ln 10 6 ln 10 2 ln 10
At 𝑥 = 0 ⇒ 𝑡 = 0, at 𝑥 = ∞ ⇒ 𝑡 = ∞
So, we can write the previous integral as

2
∞ ∞ ∞ ∞ −
−2𝑥 3 ln 10−2𝑥
3
−2𝑥 3 ln 10
1 𝑡 3
∫ 10 𝑑𝑥 = ∫ 𝑒 𝑑𝑥 = ∫ 𝑒 𝑑𝑥 = ∫ 𝑒 −t ( ) 𝑑𝑡
0 0 0 0 6 ln 10 2 ln 10
-5-
Chapter 2 Special Functions
2 2
∞ − ∞ −
−2𝑥 3
1 1 3
−
2
−t
1 1 3 1
∫ 10 𝑑𝑥 = ( ) ∫ 𝑡 𝑒 𝑑𝑡 =
3 ( ) Γ( )
0 6 ln 10 2 ln 10 0 6 ln 10 2 ln 10 3

−
2 4 −
2
1 1 3 Γ (3) 1 1 3 0.89337
= ( ) × = ( ) × = 0.53696 .
6 ln 10 2 ln 10 1 6 ln 10 2 ln 10 1
3 3

Example 2 Find the following integrals using Gamma Function

𝟏 𝟏 𝟏 𝟑 𝟏 𝟏 𝟐
(𝐢) ∫𝟎 √− 𝐥𝐧 𝒙 𝒅𝒙 (𝐢𝐢) ∫𝟎 (𝐥𝐧 ) 𝒅𝒙 (𝐢𝐢𝐢) ∫𝟎 √𝒙𝟐 − 𝟐𝒙 + 𝟏 (𝐥𝐧 ) 𝒅𝒙
𝒙 √𝟏−𝒙
Solution:
(i)
1
∫ √− ln 𝑥 𝑑𝑥
0

To convert the integral to Gamma function, we use the substitution

2 2
𝑡 = √− ln 𝑥 ⇒ 𝑥 = 𝑒 −𝑡 ⇒ 𝑑𝑥 = −2𝑡 𝑒 −𝑡 𝑑𝑡
At 𝑥 = 0 ⇒ 𝑡 = ∞, at 𝑥 = 1 ⇒ 𝑡 = 0
So, we can write the previous integral as

1 0 ∞
2 2
∫ √− ln 𝑥 𝑑𝑥 = ∫ 𝑡 (−2𝑡 𝑒 −𝑡 )𝑑𝑡 = 2 ∫ 𝑡 2 𝑒 −𝑡 𝑑𝑡
0 ∞ 0

We need another substitution to get the pervious integral in Gamma function

1 1
𝑢 = 𝑡 2 ⇒ 𝑡 = √𝑢 ⇒ 𝑑𝑡 = 𝑑𝑢
2 √𝑢
At 𝑡 = 0 ⇒ 𝑢 = 0, at 𝑡 = ∞ ⇒ 𝑢 = ∞
1 ∞ ∞ ∞ 1
2 2 1 3 1
∫ √− ln 𝑥 𝑑𝑥 = 2 ∫ 𝑡 2 𝑒 −𝑡 𝑑𝑡 = ∫ (√𝑢) 𝑒 −𝑢 𝑑𝑢 = ∫ 𝑢2 𝑒 −𝑢 𝑑𝑢 = Γ ( ) = √𝜋
0 0 0 √𝑢 0 2 2

(ii)
1 3
1
∫ (ln ) 𝑑𝑥
0 𝑥
To convert the integral to Gamma function, we use the substitution
1
𝑡 = ln ⇒ 𝑥 = 𝑒 −𝑡 ⇒ 𝑑𝑥 = − 𝑒 −𝑡 𝑑𝑡
𝑥
At 𝑥 = 0 ⇒ 𝑡 = ∞, at 𝑥 = 1 ⇒ 𝑡 = 0
So, we can write the previous integral as

1 3 0 ∞
1
∫ (ln ) 𝑑𝑥 = − ∫ 𝑡 𝑒 𝑑𝑡 = ∫ 𝑡 3 𝑒 −𝑡 𝑑𝑡 = Γ(4) = 3! = 6
3 −𝑡
0 𝑥 ∞ 0

-6-
Chapter 2 Special Functions
(iii)
1 2
1
∫ √𝑥 2 − 2𝑥 + 1 (ln ) 𝑑𝑥
0 √1 − 𝑥
To convert the integral to Gamma function, we use the substitution
1
𝑡 = ln ⇒ 𝑥 = 1 − 𝑒 −2𝑡 ⇒ 𝑑𝑥 = 2 𝑒 −2𝑡 𝑑𝑡
√1 − 𝑥
At 𝑥 = 0 ⇒ 𝑡 = 0, at 𝑥 = 1 ⇒ 𝑡 = ∞

So, we can write the previous integral as

1 2 1 2
1 1
∫ √𝑥 2 − 2𝑥 + 1 (ln ) 𝑑𝑥 = ∫ √(𝑥 − 1)2 (ln ) 𝑑𝑥
0 √1 − 𝑥 0 √1 − 𝑥
∞ ∞
−2𝑡 2 (2 −2𝑡 )𝑑𝑡
=∫ 𝑒 𝑡 𝑒 = 2 ∫ 𝑡 2 𝑒 −4𝑡 𝑑𝑡
0 0

We need another substitution to get the pervious integral in Gamma function form
𝑢 1
𝑢 = 4𝑡 ⇒ 𝑡 = ⇒ 𝑑𝑡 = 𝑑𝑢
4 4
At 𝑡 = 0 ⇒ 𝑢 = 0, at 𝑡 = ∞ ⇒ 𝑢 = ∞
1 2 ∞ ∞
1 𝑢 2 1
∫ √𝑥 2 − 2𝑥 + 1 (ln ) 𝑑𝑥 = 2 ∫ 𝑡 2 𝑒 −4𝑡 𝑑𝑡 = 2 ∫ ( ) 𝑒 −𝑢 ( ) 𝑑𝑢
0 √1 − 𝑥 0 0 4 4
1 ∞ 2 −𝑢 1 1
= ∫ 𝑢 𝑒 𝑑𝑢 = Γ(3) =
32 0 32 16

Example 3 Prove that

𝟏 (𝟐𝒏−𝟏)!
𝚪 (𝒏 + 𝟐) = 𝟐𝒏−𝟏 √𝝅
𝟐 (𝒏−𝟏)!
Solution:
1 1 1 1 3 3
Γ (𝑛 + ) = (𝑛 − ) Γ (𝑛 − ) = (𝑛 − ) (𝑛 − ) Γ (𝑛 − )
2 2 2 2 2 2
1 3 5 5
= (𝑛 − ) (𝑛 − ) (𝑛 − ) Γ (𝑛 − )
2 2 2 2
1 3 5 3 1 1
= (𝑛 − ) (𝑛 − ) (𝑛 − ) ⋯ ( ) ( ) Γ ( )
2 2 2 2 2 2
2𝑛 − 1 2𝑛 − 3 2𝑛 − 5 3 1
=( )( )( ) ⋯ ( ) ( ) √𝜋
2 2 2 2 2
(2𝑛 − 1)(2𝑛 − 3)(2𝑛 − 5) ⋯ (3)(1)
= √𝜋
2𝑛
(2𝑛 − 1)(2𝑛 − 2)(2𝑛 − 3)(2𝑛 − 4)(2𝑛 − 5) ⋯ (3)(1)
= √𝜋
2𝑛 (2𝑛 − 2)(2𝑛 − 4) ⋯ (2)
(2𝑛 − 1)(2𝑛 − 2)(2𝑛 − 3)(2𝑛 − 4)(2𝑛 − 5) ⋯ (3)(1)
= √𝜋
2𝑛 2𝑛−1 (𝑛 − 1)(𝑛 − 2) ⋯ (1)
(2𝑛 − 1)!
= 2𝑛−1 √𝜋.
2 (𝑛 − 1)!
-7-
Chapter 2 Special Functions

Beta Function

Introduction

Beta functions are a special type of function, which is also known as Euler integral of the first kind. It
is usually expressed as B(𝑥, 𝑦). Beta is a special function that is closely related to the gamma function and
to binomial coefficients.

Definitions of the Gamma function

Beta function defined by the integral
1
𝐵(𝑚, 𝑛) = ∫ 𝑡 𝑚−1 (1 − 𝑡)𝑛−1 𝑑𝑡 , 𝑛, 𝑚 > 0
0

By using the substitution 𝑡 = sin2 𝜃, we get

𝑑𝑡 = 2 sin 𝜃 cos 𝜃 𝑑𝜃,

and the limits are
𝜋
𝑡 = 0 ∶ sin2 𝜃 = 0 ⇒ 𝜃 = 0 and 𝑡 = 1 ∶ sin2 𝜃 = 1 ⇒ 𝜃 = 2
So,
𝜋
1
2
𝐵(𝑚, 𝑛) = ∫ 𝑡 𝑚−1 (1 − 𝑡)𝑛−1 𝑑𝑡 = ∫ (sin2 𝜃)𝑚−1 (1 − sin2 𝜃)𝑛−1 2 sin 𝜃 cos 𝜃 𝑑𝜃
0 0

𝜋
2
∴ 𝐵(𝑚, 𝑛) = 2 ∫ (sin 𝜃)2𝑚−1 (cos 𝜃)2𝑛−1 𝑑𝜃
0

or
𝜋
2 1
∫ (sin 𝜃)2𝑚−1 (cos 𝜃)2𝑛−1 𝑑𝜃 = 𝐵(𝑚, 𝑛)
0 2

Relation with Gamma Function

The given beta function can be written in the form of gamma function as follows:

Γ(𝑚)Γ(𝑛)
𝐵(𝑚, 𝑛) =
Γ(𝑚 + 𝑛)
Beta function is symmetric such that
Γ(𝑚)Γ(𝑛)
B(𝑚, 𝑛) = B(𝑛, 𝑚) =
Γ(𝑚 + 𝑛)
-8-
Chapter 2 Special Functions

Example 4 Find the following integral

𝟏
∫𝟎 𝒙𝟒 (𝟏 − 𝒙)𝟑 𝒅𝒙
Solution:
To solve this problem, we need expand the bracket (1 − 𝑥)3 as
(1 − 𝑥)3 = 1 − 3𝑥 + 3𝑥 2 − 𝑥 3
Then, the integration can be written as
1 1 1
∫ 𝑥 4 (1 − 𝑥)3 𝑑𝑥 = ∫ 𝑥 4 (1 − 3𝑥 + 3𝑥 2 − 𝑥 3 ) 𝑑𝑥 = ∫ (𝑥 4 − 3𝑥 5 + 3𝑥 6 − 𝑥 7 ) 𝑑𝑥
0 0 0

8 1
𝑥 5 3𝑥 6 3𝑥 7 𝑥 1 1 3 1 1
=[ − + − ] = − + − = .
5 6 7 8 0 5 2 7 8 280

We can use Beta function to find the value of the pervious example without this calculation as follows
1 1
∫ 𝑥 4 (1 − 𝑥)3 𝑑𝑥 = ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥, 𝑚 − 1 = 4, 𝑛−1=3
0 0

or
1 1
∫ 𝑥 4 (1 − 𝑥)3 𝑑𝑥 = ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥 = B(𝑚, 𝑛), 𝑚 = 5, 𝑛=4
0 0

therefore
1
Γ(5)Γ(4) 4! × 3! 1
∫ 𝑥 4 (1 − 𝑥)3 𝑑𝑥 = B(5, 4) = = =
0 Γ(9) 8! 280

Example 5 Find the following integrals using Beta Function

−𝟑 𝟓
𝟏 𝟏 𝟓
(𝐢) ∫𝟎 √𝒙 − 𝒙𝟐 𝒅𝒙 (𝐢𝐢) ∫𝟎 𝒙𝟒 (𝟏 − √𝒙) 𝒅𝒙 (𝐢𝐢𝐢) ∫ (𝟐 + 𝒙) (𝟑 + 𝒙)𝟐 ⅆ𝒙
−𝟐
Solution:
(i)
1
∫ √𝑥 − 𝑥2 𝑑𝑥
0

To convert the integral to Beta function, we take 𝑥 common outside the square root, yields
1 1 1 1 1
∫ √𝑥 − 𝑥 2 𝑑𝑥 = ∫ √𝑥√1 − 𝑥 𝑑𝑥 = ∫ 𝑥 2 (1 − 𝑥)2 𝑑𝑥
0 0 0
1
1 1
= ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥, 𝑚−1= , 𝑛−1=
0 2 2
1
3 3
= ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥 = B(𝑚, 𝑛), 𝑚= , 𝑛=
0 2 2
1 1
3 3 Γ(3/2)Γ(3/2) (2 √𝜋) (2 √𝜋) 𝜋
= B( , ) = = = .
2 2 Γ(3) 2! 8

-9-
Chapter 2 Special Functions

(ii)
1
5
∫ 𝑥 4 (1 − √𝑥)
0
To convert the integral to Beta function, we should use a substitution to change √𝑥 to 𝑡, then

𝑡 = √𝑥 ⇒ 𝑥 = 𝑡 2 ⇒ 𝑑𝑥 = 2 𝑡𝑑𝑡

At 𝑥 = 0 ⇒ 𝑡 = 0, at 𝑥 = 1 ⇒ 𝑡 = 1

So, we can write the previous integral as

1 1 1
5 4
∫ 𝑥 4 (1 − √𝑥) = ∫ (𝑡2 ) (1 − 𝑡)5 2 𝑡𝑑𝑡 = 2 ∫ 𝑡9 (1 − 𝑡)5 𝑑𝑡
0 0 0
1
= 2 ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥, 𝑚 − 1 = 9, 𝑛−1=5
0
1
= 2 ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥 = 2B(𝑚, 𝑛), 𝑚 = 10, 𝑛=6
0

Γ(10)Γ(6) 9! × 5! 1
= 2 B(10, 6) = 2 × =2× = .
Γ(16) 15! 15015

(iii)
−3 5
∫ (2 + 𝑥)3 (3 + 𝑥)2 ⅆ𝑥
−2

To convert the integral to Beta function, we should use a substitution to change 3 + 𝑥 to 𝑡, then

𝑡 =3+𝑥 ⇒ 𝑥 =𝑡−3 ⇒ 𝑑𝑥 = 𝑑𝑡

At 𝑥 = −2 ⇒ 𝑡 = 1, at 𝑥 = −3 ⇒ 𝑡 = 0

So, we can write the previous integral as

−3 5 0 5 1 5
∫ (2 + 𝑥)3 (3 + 𝑥)2 ⅆ𝑥 = − ∫ (1 − 𝑡)3 (𝑡)2 𝑑𝑡 = ∫ 𝑡 2 (1 − 𝑡)3 𝑑𝑡
−2 1 0
1
5
= ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥, 𝑚−1= , 𝑛−1=3
0 2
1
7
= ∫ 𝑥 𝑚−1 (1 − 𝑥)𝑛−1 𝑑𝑥 = B(𝑚, 𝑛), 𝑚= , 𝑛=4
0 2
7 5 3 1
Γ (2) Γ(4)
7
= B ( , 4) = = 2 × 2 × 2 × √𝜋 × 3! =
32
.
2 15 13 11 9 7 5 3 1 3003
Γ( 2 ) 2 × 2 × 2 × 2 × 2 × 2 × 2 × √𝜋

- 10 -
Chapter 2 Special Functions

Example 6 Find the following integrals using Beta Function

𝝅 𝝅 𝟓 𝝅
𝐬𝐢𝐧 𝜽 𝟏
(𝐢) ∫𝟎𝟐 𝐜𝐨𝐬 𝟐 𝜽 𝐬𝐢𝐧𝟖 𝜽 𝒅𝜽 (𝐢𝐢) ∫𝟎𝟐 𝒅𝜽 (𝐢𝐢𝐢) ∫𝟎𝟐 𝐬𝐢𝐧𝟐 𝜽 𝐬𝐢𝐧𝟐 𝟐𝜽 𝒅𝜽
𝝅 𝝅
𝐭𝐚𝐧 𝜽+𝐭𝐚𝐧𝟑 𝜽 𝟐

(iv) ∫𝟎𝟐 √𝐜𝐨𝐭 𝜽 𝒅𝜽 (v) ∫𝟎𝟐 𝐬𝐢𝐧𝟖 𝟐𝜽 𝒅𝜽

Solution:
(i)
𝜋
2
∫ cos 2 𝜃 sin8 𝜃 𝑑𝜃
0

We can solve this integral using the second form of Beta function as
𝜋 𝜋
2 2
∫ cos 2 𝜃 sin8 𝜃 𝑑𝜃 = ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 , 2𝑚 − 1 = 2, 2𝑛 − 1 = 8
0 0
𝜋
2 1 3 9
= ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 = 𝐵(𝑚, 𝑛) , 𝑚= , 𝑛=
0 2 2 2
3 9 1 7 5 3 1
1 3 9 1 Γ (2) Γ (2) 1 2 × √𝜋 × 2 × 2 × 2 × 2 × √𝜋
= B( , ) = × = ×
2 2 2 2 Γ(6) 2 5!
7𝜋
= .
512

(ii)
𝜋
2 sin5 𝜃
∫ 3
𝑑𝜃
0 tan 𝜃 + tan 𝜃

We can solve this integral using the second form of Beta function as
𝜋 𝜋 𝜋 𝜋
2 sin5 𝜃 2 sin5 𝜃 2 sin5 𝜃 2 sin5 𝜃
∫ 𝑑𝜃 = ∫ 𝑑𝜃 = ∫ 𝑑𝜃 = ∫ 𝑑𝜃
3
0 tan 𝜃 + tan 𝜃
2
0 tan 𝜃 (1 + tan 𝜃)
2
0 tan 𝜃 sec 𝜃
sin 𝜃 1
cos 𝜃 × cos 2 𝜃
0

𝜋
2
= ∫ sin4 𝜃 cos3 𝜃 𝑑𝜃
0
𝜋
2
= ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 , 2𝑚 − 1 = 4, 2𝑛 − 1 = 3
0
𝜋
2 1 5
= ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 = 𝐵(𝑚, 𝑛) , 𝑚= , 𝑛=2
0 2 2
5 3 1
1 5 1 Γ (2) Γ(2) 1 × × √𝜋 × 1! 2
= B ( , 2) = × = × 2 2 = .
2 2 2 9 2 7 × 5 × 3 × 1 × 𝜋 35
Γ (2)
2 2 2 2 √

- 11 -
Chapter 2 Special Functions
(iii)
𝜋
21
∫ sin2 𝜃 sin2 2𝜃 𝑑𝜃
0 2
We can solve this integral using the second form of Beta function as
𝜋 𝜋 𝜋
21 1 2 2
∫ sin2 𝜃 sin2 2𝜃 𝑑𝜃 = ∫ sin2 𝜃 (2 sin 𝜃 cos 𝜃)2 𝑑𝜃 = 2 ∫ sin4 𝜃 cos 2 𝜃 𝑑𝜃
0 2 2 0 0
𝜋
2
= 2 ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 , 2𝑚 − 1 = 4, 2𝑛 − 1 = 2
0
𝜋
2 5 3
= 2 ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 = 𝐵(𝑚, 𝑛) , 𝑚= , 𝑛=
0 2 2
5 3 3 1 1
5 3 Γ (2) Γ (2) × 2 × √𝜋 × 2 × √𝜋 𝜋
= B( , ) = = 2 = .
2 2 Γ(4) 3! 16

(iv)
𝜋
2
∫ √cot 𝜃 𝑑𝜃
0

We can solve this integral using the second form of Beta function as
𝜋 𝜋 𝜋
2 cos 𝜃 2 2 1 1
∫ √cot 𝜃 𝑑𝜃 = ∫ √ 𝑑𝜃 = ∫ sin−2 𝜃 cos 2 𝜃 𝑑𝜃
0 0 sin 𝜃 0
𝜋
2 1 1
= ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 , 2𝑚 − 1 = − , 2𝑛 − 1 =
0 2 2
𝜋
2 1 1 3
= ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 = 𝐵(𝑚, 𝑛) , 𝑚= , 𝑛=
0 2 4 4
1 3
1 1 3 Γ (4) Γ (4) 1 3.6256 × 1.2254
= B( , ) = = × = 2.2214 .
2 4 4 Γ(1) 2 0!

(v)
𝜋
2
∫ sin8 2𝜃 𝑑𝜃
0

We can solve this integral using the second form of Beta function as
𝜋 𝜋 𝜋
2 2 2
∫ sin8 2𝜃 𝑑𝜃 = ∫ (2 sin 𝜃 cos 𝜃)8 𝑑𝜃 = 28 ∫ sin8 𝜃 cos 8 𝜃 𝑑𝜃
0 0 0
𝜋
2
= 28 ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 , 2𝑚 − 1 = 8, 2𝑛 − 1 = 8
0

- 12 -
Chapter 2 Special Functions
𝜋 𝜋
8
2 2 2 9 9
∫ sin8 2𝜃 𝑑𝜃 = 28 ∫ (𝑠𝑖𝑛 𝜃)2𝑚−1 (𝑐𝑜𝑠 𝜃)2𝑛−1 𝑑𝜃 = 𝐵(𝑚, 𝑛) , 𝑚= , 𝑛=
0 0 2 2 2
2
9 9 7 5 3 1
8
2 9 9 Γ (2) Γ (2) (2 × 2 × 2 × 2 × √𝜋) 35𝜋
= B( , ) = = 27 × = .
2 2 2 Γ(9) 8! 256

Example 7 Prove that

𝚩(𝒑, 𝒒) = 𝚩(𝒑 + 𝟏, 𝒒)𝚩(𝒑, 𝒒 + 𝟏)
Solution:
Γ(𝑝 + 1)Γ(𝑞) Γ(𝑝)Γ(𝑞 + 1)
R. H. S = Β(𝑝 + 1, 𝑞)Β(𝑝, 𝑞 + 1) = +
Γ(𝑝 + 𝑞 + 1) Γ(𝑝 + 𝑞 + 1)
Γ(𝑝 + 1)Γ(𝑞) + Γ(𝑝)Γ(𝑞 + 1) 𝑝Γ(𝑝)Γ(𝑞) + Γ(𝑝)𝑞Γ(𝑞) (𝑝 + 𝑞)Γ(𝑝)Γ(𝑞)
= = =
Γ(𝑝 + 𝑞 + 1) (𝑝 + 𝑞)Γ(𝑝 + 𝑞) (𝑝 + 𝑞)Γ(𝑝 + 𝑞)
Γ(𝑝)Γ(𝑞)
= = B(𝑝, 𝑞) = L. H. S
Γ(𝑝 + 𝑞)

Example 8 Prove that

𝟑𝒏+𝟒
𝐁(𝟑𝒏 + 𝟑, 𝟑𝒏 + 𝟓) = 𝟏𝟐𝒏+𝟏𝟒 𝐁(𝟑𝒏 + 𝟑, 𝟑𝒏 + 𝟑)
Solution:
Γ(3𝑛 + 3) Γ(3𝑛 + 5)
L. H. S = B(3𝑛 + 3, 3𝑛 + 5) =
Γ(6𝑛 + 8)
Γ(3𝑛 + 3) × (3𝑛 + 4) × (3𝑛 + 3) × Γ(3𝑛 + 3)
=
(6𝑛 + 7) × (6𝑛 + 6) × Γ(6𝑛 + 6)

Γ(3𝑛 + 3) × (3𝑛 + 4) × (3𝑛 + 3) × Γ(3𝑛 + 3)

=
(6𝑛 + 7) × 2 × (3𝑛 + 3) × Γ(6𝑛 + 6)

(3𝑛 + 4) × Γ(3𝑛 + 3) × Γ(3𝑛 + 3)

=
2 × (6𝑛 + 7) × Γ(6𝑛 + 6)

(3𝑛 + 4) × Γ(3𝑛 + 3) × Γ(3𝑛 + 3)

=
(12𝑛 + 14) × Γ(6𝑛 + 6)

3𝑛 + 4
= B(3𝑛 + 3, 3𝑛 + 3)
12𝑛 + 14

- 13 -
Chapter 2 Special Functions
Sheet 2
Beta and Gamma Functions

1. Evaluate
7 7 4 Γ(7)
(i) Γ ( ) , Γ (− ) , Γ ( ) , Γ(2.7) (ii)
2 2 3 Γ(4)Γ(6)
3
1 3 5 Γ(3) Γ ( )
(iii) Γ ( ) Γ ( ) Γ ( ) (iv) 2
2 2 2 9
Γ( )
2
2. Find the following integrals using Gamma Function
∞ ∞ 1
(i) ∫ 𝑥 3 𝑒 −𝑥 𝑑𝑥 (ii) ∫ 𝑥 2 𝑒 2−5𝑥 𝑑𝑥 (iii) ∫ √− ln 𝑥 𝑑𝑥
0 0 0
2
∞
−𝑥 4
∞
−3𝑥 2
1 𝑥
1
(iv) ∫ 𝑒 𝑑𝑥 (v) ∫ 4 𝑑𝑥 (vi) ∫ ln ( ) 𝑑𝑥
0 0 0 𝑥
𝜋
∞ 1
2
(vii) ∫ 𝑒 (𝑥−𝑒 𝑥 ) (viii) ∫ (tan 𝑥 + tan 𝑥 )𝑒 3 5 − tan2 𝑥 (ix) ∫ √ln(1/𝑥) 𝑑𝑥
𝑑𝑥 𝑑𝑥
−∞ 0 0
∞ 1−𝑥 2 1 ∞
𝑥𝑒 1 1
(x) ∫ 𝑑𝑥 (xi) ∫ (𝑥 − 1)2 ln 𝑑𝑥 (xii) ∫ 3 𝑑𝑥
1 √𝑥 2 − 1 0 (1 − 𝑥 ) 2 1 𝑥 2 √ln 𝑥
3. Find the following integrals Beta function
1 1 4
3( 4
(i) ∫ 𝑥 1 − 𝑥 )2 𝑑𝑥 (ii) ∫ √(1/𝑥) − 1 𝑑𝑥 (iii) ∫ 𝑥 2 √4 − 𝑥 𝑑𝑥
0 0 0
3 𝜋
∞ 3
1 𝑥2 2
(iv) ∫ 𝑑𝑥 (v) ∫ 3 𝑑𝑥 (vi) ∫ cos4 𝜃 𝑑𝜃
0 √𝑥(1 + 𝑥) 0 √27 − 𝑥3 0
𝜋 𝜋
5 5 2
2 2
(vii) ∫ tan3/2 𝜃 𝑑𝜃 (viii) ∫ sin2 𝜃 cos2 𝜃 𝑑𝜃 (ix) ∫ 𝑥 3 √5 − 𝑥 𝑑𝑥
0 0 0
𝜋 𝜋
3 2
sin 𝜃
2 2
2
1
(x) ∫ 2
𝑑𝜃 (xi) ∫ cos (2𝜃) 𝑑𝜃 (xii) ∫ 𝑑𝑥
0 1 + tan 𝜃 0 0 √4 − 𝑥 2
𝜋
1 2
𝑥3 2
5 2
1
(xiii) ∫ 𝑑𝜃 (xiv) ∫ cos 𝜃 (sec 𝜃 − 1) 𝑑𝜃 (xv) ∫ 𝑥 (8 − 𝑥 3 )3 𝑑𝑥
0 √1 − 𝑥 2 0 0

4. Prove that the following

2
1 (Γ(𝑛))
(i) Β(𝑛, 𝑛 + 1) = .
2 Γ(2𝑛)
(ii) Β(𝑝, 𝑞)Β(𝑝 + 𝑞, 𝑚) = Β(𝑞, 𝑚)Β(𝑞 + 𝑚, 𝑝).
(iii) Β(𝑚, 𝑛) = Β(𝑚 + 1, 𝑛) + Β(𝑚, 𝑛 + 1) and then deduce that
𝑛
Β(𝑚, 𝑛 + 1) = Β(𝑚, 𝑛).
𝑚+𝑛

1
(iii) Β (𝑛, 𝑛 + ) = 22𝑛 Β(2𝑛, 2𝑛).
2
- 14 -
Chapter 2 Special Functions

Bessel Function

Bessel’s Differential Equation

The second order differential equation given as

𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (𝑥 2 − 𝜈 2 )𝑦 = 0

is known as Bessel’s equation. The solution of Bessel’s equation yields Bessel functions
y G . S  A J (x )  BY  (x )

for 𝜈 is fraction, where J (x ) is Bessel’s function of the first kind of order 𝜈 and it can be defined by the
series
2 r 
x 
 1  
r

J (x )   2
r 0 r ! Γ  r    1

The solution Bessel’s equation for 𝜈 = 𝑛 is integer is,

y G . S  A J n (x )  BY n (x )

where 𝑌𝑛 (𝑥) is Bessel’s function of the second kind of order 𝑛 and it can be defined by the series
cos 𝑛𝜋 1
𝑌𝑛 (𝑥) = J n (x ) − J  n (x )
sin 𝑛𝜋 sin 𝑛𝜋

The graphs of some Bessel function J n (x ) of the first kind for 𝑛 = 0, 1, 2, 3 and 4 are shown in the following

figure

J n (x )

1 J0
J1
J2 J3 J
4

- 15 -
Chapter 2 Special Functions

The graphs of some Bessel function of the second kind 𝑌𝑛 (𝑥) for 𝑛 = 0, 1, 2, 3, 4 and 5 are shown in the
following figure

Y n (x )

Y0 Y1 Y2 Y3 Y4

Example 1 Find the general solution of the following differential equations

𝟏
(𝐢) 𝒙𝟐 𝒚′′ + 𝒙𝒚′ + (𝒙𝟐 − 𝟒)𝒚 = 𝟎 (𝐢𝐢) 𝒙𝟐 𝒚′′ + 𝒙𝒚′ + (𝒙𝟐 − )𝒚 = 𝟎
𝟒
Solution:
(i) 𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (𝑥 2 − 4)𝑦 = 0
This equation is called Bessel differential equation with 𝑛 = 2. So we can write the general solution
as
y G . S  A J 2 (x )  B Y 2 ( x )
1
(ii) 𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (𝑥 2 − 4) 𝑦 = 0
1
This equation is called Bessel differential equation with 𝑛 = 2. So we can write the general solution
as
y G . S  A J 1/2 (x )  B J 1/ 2 (x )

Example 2 Find the general solution of the following differential equations

𝟏
(𝐢) 𝒙𝟐 𝒚′′ + 𝒙𝒚′ + (𝝀𝟐 𝒙𝟐 − 𝝂𝟐 )𝒚 = 𝟎 (𝐢𝐢) 𝒙𝟐 𝒚′′ + 𝒙𝒚′ + (𝟐𝟓𝒙𝟐 − )𝒚 = 𝟎
𝟗
Solution:
(i) 𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (𝜆2 𝑥 2 − 𝜈 2 )𝑦 = 0
This equation is not called directly Bessel differential equation. To convert it into Bessel Differential
equation, we can use the substitution
𝑢 1
𝑢 = 𝜆𝑥 ⇒ 𝑥= ⇒ 𝑑𝑥 = 𝜆 𝑑𝑢,
𝜆
𝑑𝑦 𝑢 𝑑𝑦 𝑑𝑢 𝑢 𝑑𝑦 𝑑𝑦
𝑥𝑦 ′ = 𝑥 = ∙ = ∙𝜆 =𝑢
𝑑𝑥 𝜆 𝑑𝑢 𝑑𝑥 𝜆 𝑑𝑢 𝑑𝑢
𝑑 𝑑𝑦 𝑑𝑢 𝑑 𝑑𝑦 𝑑𝑢 𝑢 2 𝑑 𝑑𝑦 𝑑2𝑦
𝑥 2 𝑦 ′′ = 𝑥 2 ( ) = 𝑥2 ∙ ( ∙ )=( ) 𝜆 ( ∙ 𝜆) = 𝑢2 2
𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑑𝑢 𝑑𝑢 𝑑𝑥 𝜆 𝑑𝑢 𝑑𝑢 𝑑𝑢
Then we can rewrite the equation as

- 16 -
Chapter 2 Special Functions
2
𝑑 𝑦 𝑑𝑦
2
+𝑢 𝑢2
+ (𝑢2 − 𝜈 2 )𝑦 = 0
𝑑𝑢 𝑑𝑢
that is the Bessel equation has the general solution as
y G . S  A J (u )  B J  (u )
or
y G . S  A J ( x )  B J  ( x )

1
(ii) 𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (25𝑥 2 − 9)𝑦 = 0
1
By comparing this equation by the final form in (i) yields, 𝜆 = 5 and 𝜈 = 3, then the general

solution is
y G . S  A J 1/3 (5x )  B J 1/ 3 (5x )

Theorem
2 2
J 1/2 (x )  sin x , J 1/2 (x )  cos x
x x
Proof:
Because
2 r 
x 
 1
r
  
J (x )   2
r 0 r ! Γ  r    1

therefore
1 1
2r  2r 

 1  
x r x 
 
1  
r 2 2
1
 1
2 r 1
 1/2 1/2 r
2  2 x   2  2 2  x 
J ( x )  
1  
   
  1   2   x   x  r 0  1   2 
r 0
r ! Γ  r   1 r 0
  r !   Γ  r   1 r !   Γ  r  1  
 2   2   2

1 √𝜋 (2𝑛)!
∵ Γ (𝑛 + 2) = 2𝑛
2 𝑛!

1 √𝜋 (2𝑟+2)! √𝜋 (2𝑟+2)(2𝑟+1)! √𝜋 (2𝑟+2)(2𝑟+1)! √𝜋 (2𝑟+1)!

∴ Γ ((𝑟 + 1) + 2) = 2𝑟+2 = 2𝑟+2 = 2𝑟+2 = 2𝑟+1
2 (𝑟+1)! 2 (𝑟+1) 𝑟! 2 (𝑟+1) 𝑟! 2 𝑟!

1 1
 1 2 r 1
 1 2 r 1
r r
 2 2  x   2 2  x 
 J ( x )           
 x  r 0 r ! Γ  r  1  1   2   x  r 0   2r  1! 2
  r!
 2 22 r 1 r!
1 1
1  2  2   1 2 1  2  2   1 2  1
2 r 1 r 2 r 1 r r
x 2 r 1 x 2 r 1 2 
          x 2 r 1
  x  r 0  2r  1! 2   x  r 0  2r  1! 2 x  2r  1!
2 r 1 2 r 1
r 0

2
 sin x
x
- 17 -
Chapter 2 Special Functions
Example 3 Find the solution of the following differential equation
𝒅 𝟏 𝝅 𝝅
(𝒙𝟐 𝒚′ ) − 𝒙𝒚′ + 𝒙𝟐 𝒚 = 𝒚, 𝒚 ( ) = 𝟎, 𝒚′ ( ) = −𝟏
𝒅𝒙 𝟒 𝟐 𝟐
Solution:
To solve this equation, we find the derivative of 𝑥 2 𝑦′′ then simplify the differential equation yields
𝑑 2 ′ 1 1
(𝑥 𝑦 ) − 𝑥𝑦 ′ + 𝑥 2 𝑦 = 𝑦 ⇒ 𝑥 2 𝑦 ′′ + 2𝑥𝑦 ′ − 𝑥𝑦 ′ + 𝑥 2 𝑦 − 𝑦 = 0
𝑑𝑥 4 4
1 1
𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ + (𝑥 2 − ) 𝑦 = 0, 𝜆 = 1, 𝜈 =
4 2
The solution of this equation is
y G . S  A J 1/2 (x )  B J 1/ 2 (x )

2 2
∴ 𝑦𝐺.𝑆 = 𝐴 √𝜋𝑥 sin 𝑥 + 𝐵√𝜋𝑥 cos 𝑥

𝜋 𝜋
∵ 𝑦 ( 2 ) = 0, we can substitute into the general solution to get the particular solution. At 𝑥 = 2 , 𝑦 = 0,

we get
2 𝜋 2 𝜋 2 2
0 = 𝐴 √𝜋𝑥 sin 2 + 𝐵√𝜋𝑥 cos 2 ⇒ 𝐴 √𝜋𝑥 = 0 ⇒ ∴𝐴=0 ⇒ ∴ 𝑦 = 𝐵√𝜋𝑥 cos 𝑥

3
2 1 2
𝑦 ′ = −𝐵√𝜋𝑥 sin 𝑥 − 2 𝐵√𝜋 𝑥 −2 cos 𝑥

𝜋 𝜋
∵ 𝑦 ′ ( 2 ) = −1 then 𝑥 = at 𝑦 ′ = −1. By substituting into the derivative of 𝑦, we get
2

3
2 𝜋 1 2 𝜋 −2 𝜋 4 𝜋
−1 = −𝐵√ 𝜋 sin − 𝐵√ ( ) cos ⇒ −1 = −𝐵√ 2 ⇒ 𝐵=
𝜋 (2) 2 2 𝜋 2 2 𝜋 2

Differentiation Identities
d
dx
 x  J (x )   x  J 1 (x ) and
d
dx
 x  J (x )   x  J 1 (x )
Proof:

To prove the first relation, we multiply the series for

2 r 
x 
 1
r
  
J (x )   2
r 0 r ! Γ  r    1

by 𝑥 𝜈 and then differentiate the result with respect to 𝑥, this gives us

2 r 
x  x 2 r  2
 
1     1
r r

22 r    1  2r  2  x
 r 2 r  2 1
d
 x  J ( x )  
d 
x  2 
d
 
dx dx r  0 r ! Γ  r    1 dx r  0 r !   Γ  r    1 r 0 r ! 22 r    Γ  r    1

- 18 -
Chapter 2 Special Functions

 1  2r  2  x 2  1  r   x
2 r  2 1 2 r  1
 r  r
d
dx
 x  J ( x )   
r ! 22 r  Γ  r   1
 x 
 2 r  2
  r   Γ  r  
r 0 r 0 r ! 2

 1  r   x 2r  1  x    1  r   x 2 r  1
 r r

x 

r 0 r ! 2
2 r  2 1
 r   Γ  r   
r  0 r ! 2
2 r  2 1
  r   Γ  r  
2 r  1
x 2 r  1 r x  x 2 r  1
  1
r
  1     1
r

 x  22 r  2 1  x  2 22 r  2 1
r 0 r ! Γ  r  

r 0 r !   Γ  r   
 x 
r 0 r !   Γ  r   
2 r  1
x 
 
1  
r

 x  2  x  J 1 (x )
r 0 r ! Γ  r  
Integration Identities

x x

J 1 (x ) dx  x  J (x )  c and 
J 1 (x ) dx  x  J (x )  c
Proof:
By integrate the differentiation identities, we can get it.

 xJ 9 (x )J 10 (x )   x  J 92 (x )  J 102 (x ) 
d
Example 4 Show that
dx
Solution:

To prove this relation, we start with the left hand side and try to get the right hand side as follow
d
dx
 x J 9 (x ) J 10 (x )  
d
dx
 x 9
J 9 (x )    x 10 J 10 (x )  
  x 9 J 9 (x )   x 10 J 10 (x )    x 10 J 10 (x )   x 9 J 9 (x ) 
d d
dx dx
  x 9 J 9 (x )  x 10 J 9 (x )    x 10 J 10 (x )  x 9 J 10 (x ) 

 x J 92 (x )  x J102 (x )  x  J 92 (x )  J 102 (x )  .

Example 5 Evaluate the following integrals

(i)  x 4J1 (x )dx (ii)  x 2J 0 (x )J1 (x )dx
Solution:

x J1 (x )dx   x 2  x 2J1 (x )dx  x 2  x 2J 2 (x )   2x  x 2J 2 (x )dx  x 4J 2 (x )  2x 3J 3 (x )  c

4
(i)

(ii) x
2
J 0 (x )J1 (x )dx    x J1 (x )    x J 0 (x )  dx   x J 1 (x )    x J 1 (x )     x J 0 (x )    x J 1 (x )  dx

x J 0 (x )J 1 (x )dx  x 2 J 12 (x )   x 2J 0 (x )J 1 (x )dx
2

1 2 2
2 x 2 J 0 (x )J 1 (x )dx  x 2 J 12 (x )  x
2
J 0 (x )J 1 (x )dx  x J 1 (x )  c
2

- 19 -
Chapter 2 Special Functions
Recurrence Relations
n n
(i) J n ( x )  J n ( x )  J n 1 ( x ) (ii) J n ( x )  J n ( x )  J n 1 ( x )
x x
2n
(iii) 2J n ( x )  J n 1 ( x )  J n 1 ( x ) (iv) J n ( x )  J n 1 ( x )  J n 1 ( x )
x
Proof:

To this relations, we can recall the differentiation relations

d
dx
 x n J n (x )   x n J n 1 (x )  x n J n (x )  nx n 1J n (x )  x n J n 1 (x )

By divided the both sides by 𝑥 𝑛 yields,

n
J n (x )  J (x )  J n 1 (x ) (1)
x
and

d
dx
 x  n J n (x )   x  n J n 1 (x )  x  n J n (x )  nx  n 1J n (x )  x  n J n 1 (x )

By divided the both sides by 𝑥 −𝑛 yields,

n
J n (x )  J (x )  J n 1 (x ) (2)
x
By adding the two relation (1) and (2), we get

2J n (x )  J n 1 (x )  J n 1 (x ) (3)

By subtracting the two relation (1) and (2), we get

n
J (x )  J n 1 (x )  J n 1 (x ) (4)
x

Example 5 Express J 3/ 2 ( x ) and J 3/ 2 ( x ) in terms of 𝐬𝐢𝐧 𝒙 and 𝐜𝐨𝐬 𝒙

Solution:
2n
J n ( x )  J n 1 ( x )  J n 1 ( x )
x
2n
 J n 1 (x )  J n ( x )  J n 1 ( x )
x
1 1 1 2 2 2  sin x 
At 𝑛 = 2 ∶ J 3/ 2 ( x )  J 1/ 2 ( x )  J 1/ 2 ( x )  sin x  cos x    cos x 
x x x x x  x 
2n
J n ( x )  J n 1 ( x )  J n 1 ( x )
x
2n
 J n 1 ( x )  J n ( x )  J n 1 ( x )
x
1 1 1 2 2 2  cos x 
At 𝑛 = − ∶ J 3/ 2 ( x )   J 1/ 2 ( x )  J 1/ 2 ( x )   cos x  sin x    sin x 
2 x x x x x  x 

- 20 -
Chapter 2 Special Functions
Example 6 Express J 3 ( x ) in terms of J 0 ( x ) and J 1 ( x )
Solution:
2n
J n ( x )  J n 1 ( x )  J n 1 ( x )
x
2n
 J n 1 (x )  J n ( x )  J n 1 ( x )
x
4
At 𝑛 = 2 ∶ J 3 (x )  J 2 (x )  J 1 (x )
x
2
At 𝑛 = 1 ∶ J 2 (x )  J 1 (x )  J 0 (x )
x
By replacing each J 2 ( x ) in J 3 ( x ) , we get
42  8 4  8  4
J 3 (x )   J 1 (x )  J 0 (x )   J 1 (x )  2 J 1 (x )  J 0 (x )  J 1 (x )   2  1  J 1 (x )  J 0 (x )
x x  x x x  x
 8  4
J 3 ( x )    1 J 1 (x )  J 0 (x )
x 
2
x

Theorem For integer 𝑛 the following relations hold

J n (x )  (1)n J n (x )
Proof:
To prove this theorem, we replace each 𝜈 by −𝑛 in J (x ) , we get
2r n
x 
 1
r
  
J  n (x )   2
r 0 r ! Γ  r  n  1
2r n 2r n 2 r  n   n

 1  
r x 
 1  
x x
 
1  
r r n
 
J  n (x )   2  2  2
r 0 r ! Γ  r  n  1 r  n   r !   Γ  r  n  1 r 0    r  n !   Γ  r  n  n  1
2r n

 1  
r x

2
  1    1 J n (x )
n n

r 0 r ! Γ  r  n  1

Example 7 Prove that

 x J 3 (x ) J 2 (x )   x  J 22 (x )  J 32 (x ) 
d
dx
Solution:
To prove this relation, we start with the left hand side and try to get the right hand side as follow

 x J 3 (x ) J 2 (x )    (1)3 x J 3 (x )  (1) 2 J 2 (x )     x J 3 (x )  J 2 (x ) 
d d d
dx dx dx

d
dx
  
x 3 J 3 (x )    x 2 J 2 (x )     x 3 J 3 (x )  x 2 J 3 (x )    x 3 J 2 ( x )  x 2J 2 ( x ) 

 x J 32 (x )  x J 22 (x )  x  J 32 (x )  J 22 (x ) 

- 21 -
Chapter 2 Special Functions
Sheet 3
Bessel Function

1. Write the first four terms of

J 0 (x ), J 1 (x ), J 2 (x ), J 3 (x ), J 4 (x ), J 5 (x ), J 6 (x )

2. Evaluate J 3/2 (x ), J 3/2 (x ), J 5/2 (x ), J 5/2 (x ) in terms of sines and cosines

3. Prove that

(i) J 4 (x )  
 48  8  J (x )   24  1 J (x )
 1  2  0
x x  x 
3

 12   24 5 
(ii) J 3 (x )   2  1  J 0 (x )   3   J 0 ( x )
x  x x 

(iii) 4J 0(x )  3J 0 (x )  J 3 (x )  0
1
(iv) J n( x )   J n 2 ( x )  2 J n ( x )  J n  2 ( x ) 
4
1
(v) J 2 ( x )  J 0( x )  J 0 (x )
x
4. Prove that
d
(i)  xJ 3 (x ) J 2 (x )  x J 22 (x )  J 32 (x ) 
dx
d
(ii)  xJ 3 (x ) J 4 (x )  x J 32 (x )  J 42 (x ) 
dx
d 2 
(iii)  x 2 J 1 (x ) J 3 (x )   x 2  J 1 (x )  J 3 (x )   J 1 (x )  J 0 (x ) 
dx x 
d
(iv)  xJ n (x ) J n 1 (x )  x J n2 (x )  J n21 (x ) 
dx
d
(v)  x 2J n 1 (x ) J n 1 (x )   2x 2J n (x ) J n (x )
dx
5. Integrate the following

 J 4 (x )dx  J 5/2 (x )dx 

5 3/2 3
(i ) x (ii ) x (iii ) x J 0 (x )dx

 x (vi )  J (x )dx
6 2 n
(iv ) x J 3 (x )dx (v ) J n 1 (x )dx 1

(vii )  x (viii )  x (ix )  J (x )dx

2 4
J 0 (x )J 1 (x )dx J 1 (x )dx 3

- 22 -
Chapter 2 Special Functions

 x ln x x x
2 3
(x ) J 0 (x )dx ( xi ) J 0 (x )dx ( xii ) J 3 (x )dx

J
(xiii )  4 (x )dx (xiv )  x 9 J 3 (x )dx (xv )  x 7 J 1 (x )dx
x 
2  /2 2
 3 /2
1
(xvi )


/2
x J 1 (x )dx
2
(xvii )


/2 x
J 3 (x )dx

2

6. Solve each of the following differential equations:

d
(x y )  x
dy
x 
 1     0 , y      1
y  0 y   
2 2
(i)
dx dx  4 2 2

x y  
d 
(xy )   9x 
5     0 , y      2
y  0 y   
2 2
(ii)
dx  4 2 2

x y  
d
(xy )   x 
 5     0 , y      10
y  0 y   
2 2
(iii)
dx  4 2 2

x y   3xy ' 2
d
( xy )   16x 
 7     0 , y     3
y  0 y   
2 2
(iv)
dx  4 2 2

x y   2xy '
d
(xy )   4x 
 3     0 , y      2
y  0 y   
2 2
(v)
dx  4 2 2

y  
3
y '
2 d
(xy )   16 
 7 
y 0
    0 , y      4
2 
(vi) y   
 4x  2 2
2
x x dx

- 23 -
Chapter 2 Special Functions

Legendre Polynomials

Legendre Differential Equation

Consider the Legendre equation

(1 − 𝑥 2 )𝑦 ′′ − 2𝑥𝑦 ′ + 𝑛(𝑛 + 1)𝑦 = 0, 𝑛 ≥ 0

for 𝑛 an integer. We first note that there are singular points for 1 − 𝑥 2 = 0, or 𝑥 = ±1. Therefore, 𝑥 = 0
is an ordinary point and we can proceed to obtain solutions in the form of Maclaurin series expansions.
Assume the differential equation has a solution of the form

∞ ∞ ∞
𝑟 ′ 𝑟−1
𝑦 = ∑ 𝑐𝑟 𝑥 , 𝑦 = ∑ 𝑟𝑐𝑛 𝑥 anⅆ 𝑦 = ∑ 𝑟(𝑟 − 1)𝑐𝑛 𝑥 𝑟−2
′′

𝑟=0 𝑟=1 𝑟=2

By inserting the previous series expansions into the differential equation, we obtain
∞ ∞ ∞

(1 − 𝑥 2 ) ∑ 𝑟(𝑟 − 1)𝑐𝑟 𝑥 𝑟−2 − 2𝑥 ∑ 𝑟𝑐𝑟 𝑥 𝑟−1 + 𝑛(𝑛 + 1) ∑ 𝑐𝑟 𝑥 𝑟 = 0

𝑟=2 𝑟=1 𝑛=0

∞ ∞ ∞ ∞

∑ 𝑟(𝑟 − 1)𝑐𝑟 𝑥 𝑟−2 − ∑ 𝑟(𝑟 − 1)𝑐𝑟 𝑥 𝑟 − 2 ∑ 𝑟𝑐𝑟 𝑥 𝑟 + 𝑛(𝑛 + 1) ∑ 𝑐𝑟 𝑥 𝑟 = 0

𝑟=2 𝑟=2 𝑟=1 𝑟=0

Note that when we change ∑∞

𝑟=2 𝑟(𝑟 − 1)𝑐𝑟 𝑥
𝑟−2
to ∑∞ 𝑟
𝑟=0(𝑟 + 2)(𝑟 + 1)𝑐𝑟+2 𝑥 , the index in the sequence
increases by 2
Now , we can write the last equation as
∞ ∞ ∞ ∞

∑(𝑟 + 2)(𝑟 + 1)𝑐𝑟+2 𝑥 − ∑ 𝑟(𝑟 − 1)𝑐𝑟 𝑥 − 2 ∑ 𝑟𝑐𝑟 𝑥 + 𝑛(𝑛 + 1) ∑ 𝑐𝑟 𝑥 𝑟 = 0

𝑟 𝑟 𝑟

𝑟=0 𝑟=0 𝑟=0 𝑟=0

or
∞

∑[(𝑟 + 2)(𝑟 + 1)𝑐𝑟+2 − 𝑟(𝑟 + 1)𝑐𝑟 + 𝑛(𝑛 + 1)𝑐𝑟 ] 𝑥 𝑟 = 0

𝑟=0

and the recurrence relation is

(𝑟 + 2)(𝑟 + 1)𝑐𝑟+2 = (𝑟(𝑟 + 1) − 𝑛(𝑛 + 1))𝑐𝑟

𝑟(𝑟 + 1) − 𝑛(𝑛 + 1)
𝑐𝑟+2 = 𝑐𝑟 , 𝑟 ≥ 0
(𝑟 + 2)(𝑟 + 1)
- 24 -
Chapter 2 Special Functions
1 1
𝑟 = 0: 𝑐2 = 2 (−𝑛(𝑛 + 1))𝑐0 = − 2 𝑛(1 + 𝑛)𝑐0,
1 1
𝑟 = 1: 𝑐3 = 6 (2 − 𝑛(𝑛 + 1))𝑐1 = 6 (1 − 𝑛)(2 + 𝑛)𝑐1,
1 1
𝑟 = 2: 𝑐4 = 12 (6 − 𝑛(𝑛 + 1))𝑐2 = − 24 𝑛(2 − 𝑛)(1 + 𝑛)(3 + 𝑛)𝑐2,
1 1
𝑛 = 3: 𝑐5 = 20 (12 − 𝑛(𝑛 + 1))𝑐3 = 120 (1 − 𝑛)(3 − 𝑛)(2 + 𝑛)(4 + 𝑛)𝑐3,

The solution is

1 1
𝑦 = 𝑐0 (1 − 2 𝑛(1 + 𝑛)𝑥 2 − 24 𝑛(2 − 𝑛)(1 + 𝑛)(3 + 𝑛)𝑥 4 + ⋯ )
1 1
+𝑐1 (𝑥 + (1 − 𝑛)(2 + 𝑛)𝑥 3 + (1 − 𝑛)(3 − 𝑛)(2 + 𝑛)(4 + 𝑛)𝑥 5 +. . . . . . )
6 120

𝑦 = 𝑐0 𝑃𝑛 (𝑥) + 𝑐1 𝑄𝑛 (𝑥)

where 𝑃𝑛 (𝑥) is Legendre polynomials of the first kind of order 𝑛 and they can be defined by the series
𝑛
[ ]
2
(−1)𝑟 (2𝑛 − 2𝑟)!
𝑃𝑛 (𝑥) = ∑ 𝑛 𝑥 𝑛−2𝑟
2 𝑟! (𝑛 − 𝑟)! (𝑛 − 2𝑟)!
𝑟=0

𝑄𝑛 (𝑥) is Legendre polynomials of the second kind of order 𝑛 which are bounded at 𝑥 = ±1, and they can
be defined by the series

𝑀
1 1−𝑥 (−1)𝑛−2𝑖 (2𝑛 − 4𝑖 − 1)
𝑄𝑛 (𝑥) = 𝑃𝑛 (𝑥) ln −∑ 𝑃𝑛−2𝑖−1 (𝑥)
2 1+𝑥 (2𝑖 + 1)(𝑛 − 𝑖)
𝑖=0

where
(𝑛 − 1)/2 for 𝑛 oⅆⅆ
𝑀={
(𝑛 − 2)/2 for 𝑛 even, 𝑛 ≥ 1

Rodrigues's formula:
A representation of the Legendre polynomials involving differentiation is given by Rodrigues's formula

1 𝑑𝑛
𝑃𝑛 (𝑥) = (𝑥 2 − 1)𝑛 , 𝑛 = 0, 1, 2, 3, …
2𝑛 𝑛! 𝑑𝑥 𝑛
as follows

𝑃0 (𝑥) = 1
1 𝑑 2
𝑃1 (𝑥) = (𝑥 − 1) = 𝑥
2 𝑑𝑥
1 𝑑2 2 2
1 𝑑2 4 2
1
𝑃2 (𝑥) = 2 (𝑥 − 1) = (𝑥 − 2𝑥 + 1) = (−1 + 3𝑥 2 )
2 2! 𝑑𝑥 2 22 2! 𝑑𝑥 2 2
3
1 𝑑 1 𝑑3 1
𝑃3 (𝑥) = 3 3
(𝑥 2 − 1)3 = 3 3
(𝑥 6 − 3𝑥 4 + 3𝑥 2 − 1) = (−3𝑥 + 5𝑥 3 )
2 3! 𝑑𝑥 2 3! 𝑑𝑥 2

- 25 -
Chapter 2 Special Functions

We can express the polynomial functions in terms of Legendre polynomials. So

𝑘 = 𝑘 𝑃0 (𝑥)
𝑥 = 𝑃1 (𝑥)
1
𝑥 2 = 3 (2𝑃2 (𝑥) + 𝑃0 (𝑥))
1
𝑥 3 = 5 (2𝑃3 (𝑥) + 3𝑃1 (𝑥))
1
𝑥 4 = 35 (8𝑃4 (𝑥) + 40𝑃2 (𝑥) + 17𝑃0 (𝑥))
⋮

Example 1 Express 𝟐𝒙𝟑 + 𝟐𝒙𝟐 − 𝒙 − 𝟑 in terms of Legendre’s polynomials

Solution
1 1
We know that 1 = 𝑃0 (𝑥), 𝑥 = 𝑃1 (𝑥), 𝑥 2 = 3 (2𝑃2 (𝑥) + 𝑃0 (𝑥)) and 𝑥 3 = 5 (2𝑃3 (𝑥) + 3𝑃1 (𝑥)). Then
1 1
2𝑥 3 + 2𝑥 2 − 𝑥 − 3 = 2 (5 (2𝑃3 (𝑥) + 3𝑃1 (𝑥))) + 2 (3 (2𝑃2 (𝑥) + 𝑃0 (𝑥))) − 2𝑃1 (𝑥) − 𝑃0 (𝑥)
1 4 4 4
= − 𝑃0 (𝑥) − 𝑃1 (𝑥) − 𝑃2 (𝑥) + 𝑃3 (𝑥).
3 5 3 5

Generating Function for Legendre Polynomials:

On expanding the function

1
√1 − 2𝑥𝑡 + 𝑡 2

in power of 𝑡, the coefficient of 𝑡 𝑛 will be 𝑃𝑛 (𝑥). For this reason this function is called Legendre generating
function. For |𝑥| ≤ 1 and |𝑡| ≤ 1, the following relation holds
∞
1
= ∑ 𝑃𝑛 (𝑥) 𝑡 𝑛
√1 − 2𝑥𝑡 + 𝑡2 𝑛=0

Proof:
1
Consider the function 𝑤(𝑥, 𝑡) = √1−2𝑥𝑡+𝑡 2 . Recall the binomial expansion
1 3 5 3
(1 − 𝑧)−1/2 = 1 + 𝑧 + 𝑧 2 + 𝑧 +⋯
2 8 16
So,
1 1
−
= (1 − (2𝑥𝑡 − 𝑡 2 )) 2
√1 − 2𝑥𝑡 + 𝑡 2
1 1∙3 1∙3∙5
= 1 + (2𝑥𝑡 − 𝑡 2 ) + (2𝑥𝑡 − 𝑡 2 )2 + (2𝑥𝑡 − 𝑡 2 )3 + ⋯
2 2∙4 2∙4∙6
1 3 5
= 1 + 𝑥𝑡 − 𝑡 2 + (4𝑥 2 𝑡 2 − 4𝑥𝑡 3 + 𝑡 4 ) + (8𝑥 3 𝑡 3 − 12𝑥 2 𝑡 4 + 6𝑥𝑡 5 − 𝑡 6 ) + ⋯
2 8 16
1 1
= 1 + 𝑥𝑡 + (3𝑥 2 − 1)𝑡 2 + (5𝑥 3 − 𝑥)𝑡 3 + ⋯
2 2 ∞

= 𝑃0 (𝑥) + 𝑃1 (𝑥)𝑡 + 𝑃2 (𝑥)𝑡 2 + 𝑃3 (𝑥)𝑡 3 + ⋯ = ∑ 𝑃𝑛 (𝑥) 𝑡 𝑛

𝑛=0

- 26 -
Chapter 2 Special Functions

Example 2 Use the generating function of Legendre Polynomials to prove

𝑷𝒏 (𝟏) = 𝟏 and 𝑷𝒏 (−𝟏) = (−𝟏)𝒏
Solution
From the generating function, we have

∞
1 1 1
∑ 𝑃𝑛 (1) 𝑡 𝑛 = = = = (1 − 𝑡)−1
√1 − 2𝑡 + 𝑡 2 √(1 − 𝑡)2 1−𝑡
𝑛=0
∞

= 1 + 𝑡 + 𝑡2 + 𝑡3 + ⋯ = ∑ 𝑡𝑛
𝑛=0

which implies that 𝑃𝑛 (1) = 1 by comparing the coefficients of 𝑡 𝑛 in both sides. Similar, we can prove the
other relation as

∞
1 1 1
∑ 𝑃𝑛 (−1) 𝑡 𝑛 = = = = (1 + 𝑡)−1
√1 + 2𝑡 + 𝑡 2 √(1 + 𝑡)2 1+𝑡
𝑛=0
∞

= 1 − 𝑡 + 𝑡 − 𝑡 + ⋯ = ∑(−1)𝑛 𝑡 𝑛
2 3

𝑛=0
which implies that 𝑃𝑛 (−1) = (−1)𝑛 by comparing the coefficients of 𝑡 𝑛 in both sides.

Recurrence Formula:
The following relation is hold
(2𝑛 + 1)𝑥𝑃𝑛 (𝑥) = (𝑛 + 1)𝑃𝑛+1 (𝑥) + 𝑛𝑃𝑛−1 (𝑥)

Proof:

To prove this relation, we differentiate both sides of the generating function

∞
1 1
2 ) −2
= (1 − 2𝑥𝑡 + 𝑡 = ∑ 𝑃𝑛 (𝑥) 𝑡 𝑛
√1 − 2𝑥𝑡 + 𝑡2 𝑛=0

with respect to 𝑡, so we have

∞
1 3
− (−2𝑥 + 2𝑡)(1 − 2𝑥𝑡 + 𝑡 2 )−2 = ∑ 𝑛 𝑃𝑛 (𝑥) 𝑡 𝑛−1
2
𝑛=0
or
∞
𝑥−𝑡
3 = ∑ 𝑛 𝑃𝑛 (𝑥) 𝑡 𝑛−1
(1 − 2𝑥𝑡 + 𝑡 2 )2 𝑛=0

and the last form can write in the form

∞
𝑥−𝑡
= ∑ 𝑛 𝑃𝑛 (𝑥) 𝑡 𝑛−1
(1 − 2𝑥𝑡 + 𝑡 2 )√1 − 2𝑥𝑡 + 𝑡 2 𝑛=0
- 27 -
Chapter 2 Special Functions
or
∞
𝑥−𝑡 2)
= (1 − 2𝑥𝑡 + 𝑡 ∑ 𝑛 𝑃𝑛 (𝑥) 𝑡 𝑛−1
√1 − 2𝑥𝑡 + 𝑡2 𝑛=0

By recalling the generating function yields,

∞ ∞
𝑛 2)
(𝑥 − 𝑡) ∑ 𝑃𝑛 (𝑥) 𝑡 = (1 − 2𝑥𝑡 + 𝑡 ∑ 𝑛 𝑃𝑛 (𝑥) 𝑡 𝑛−1
𝑛=0 𝑛=0

by comparing the coefficients of 𝑡 𝑛 in both sides yields,

𝑥 𝑃𝑛 (𝑥) − 𝑃𝑛−1 (𝑥) = (𝑛 + 1)𝑃𝑛+1 (𝑥) − 2𝑛𝑥𝑃𝑛 (𝑥) + (𝑛 − 1)𝑃𝑛−1 (𝑥)

or

(2𝑛 + 1)𝑥 𝑃𝑛 (𝑥) = (𝑛 + 1)𝑃𝑛+1 (𝑥) + 𝑛𝑃𝑛−1 (𝑥).

The orthogonality of Legendre polynomials:

1
0, 𝑛 ≠ 𝑚.
∫ 𝑃𝑛 (𝑥)𝑃𝑚 (𝑥)𝑑𝑥 = { 2
, 𝑛 = 𝑚.
−1 2𝑛 + 1

Example 3 Use the orthogonality of Legendre polynomials to evaluate the following integrals
𝟏 𝟏 𝟏
(i) ∫−𝟏 𝑷𝟐 (𝒙)𝑷𝟑 (𝒙)𝒅𝒙 (ii) ∫−𝟏 𝑷𝟓 (𝒙)𝑷𝟓 (𝒙)𝒅𝒙 (iii) ∫−𝟏 𝑷𝟏 (𝒙)𝑷𝟒 (𝒙)𝒅𝒙
𝟏 𝟏 𝟏
(iv) ∫–𝟏 𝒙𝑷𝟐 (𝒙)𝑷𝟑 (𝒙)𝒅𝒙 (v) ∫–𝟏 𝒙𝟐 𝑷𝟓 (𝒙)𝑷𝟕 (𝒙)𝒅𝒙 (vi) ∫–𝟏 𝑷𝟏 (𝒙)𝑷𝟑 (𝒙)𝑷𝟒 (𝒙)𝒅𝒙
𝟏 𝟏 𝒙𝟐
(vii) ∫–𝟏(𝟏 − 𝟐𝒙 + 𝒙𝟑 )𝑷𝟑 (𝒙)𝒅𝒙 (viii) ∫−𝟏 𝒅𝒙
√𝟓−𝟒𝒙
𝟏
𝑷𝟏 (𝒙) 𝑷𝟑 (𝒙)
(ix) ∫ 𝒅𝒙
√𝟓𝟖−𝟒𝟐𝒙
−𝟏
Solution
1
(i) ∫−1 𝑃2 (𝑥)𝑃3 (𝑥)𝑑𝑥 = 0

1 2 2
(ii) ∫−1 𝑃5 (𝑥)𝑃5 (𝑥)𝑑𝑥 = 2(5)+1 = 11

1
(iii) ∫−1 𝑃1 (𝑥)𝑃4 (𝑥)𝑑𝑥 = 0

1
(iv) ∫−1 𝑥𝑃2 (𝑥)𝑃3 (𝑥)𝑑𝑥

To solve this integration, we must replace 𝑥𝑃2 (𝑥) by using the recurrence relation

(𝑛 + 1) 𝑛
𝑥 𝑃𝑛 (𝑥) = 𝑃𝑛+1 (𝑥) + 𝑃 (𝑥).
(2𝑛 + 1) (2𝑛 + 1) 𝑛−1

Then,
3 2
𝑥𝑃2 (𝑥) = 𝑃3 (𝑥) + 𝑃1 (𝑥)
5 5

- 28 -
Chapter 2 Special Functions

1 1 3 2
So, ∫−1 𝑥𝑃2 (𝑥)𝑃3 (𝑥)𝑑𝑥 = ∫−1 ( 𝑃3 (𝑥) + 𝑃1 (𝑥)) 𝑃3 (𝑥)𝑑𝑥
5 5

1 3 1 2 3 2 6
= ∫−1 5 𝑃3 (𝑥)𝑃3 (𝑥)𝑑𝑥 + ∫−1 5 𝑃1 (𝑥)𝑃3 (𝑥)𝑑𝑥 = (5) (7) + 0 = 35

1
(v) ∫–1 𝑥 2 𝑃5 (𝑥)𝑃7 (𝑥)𝑑𝑥

To solve this integration, we must replace 𝑥𝑃5 (𝑥) and 𝑥𝑃7 (𝑥) by using the recurrence relation

(𝑛 + 1) 𝑛
𝑥 𝑃𝑛 (𝑥) = 𝑃𝑛+1 (𝑥) + 𝑃 (𝑥).
(2𝑛 + 1) (2𝑛 + 1) 𝑛−1

Then,
6 5 8 7
𝑥𝑃5 (𝑥) = 11 𝑃6 (𝑥) + 11 𝑃4 (𝑥) and 𝑥𝑃7 (𝑥) = 15 𝑃8 (𝑥) + 15 𝑃6 (𝑥)

1 1 6 5 8 7
So, ∫−1 𝑥 2 𝑃5 (𝑥)𝑃7 (𝑥)𝑑𝑥 = ∫−1 ( 𝑃6 (𝑥) + 𝑃4 (𝑥)) ( 𝑃8 (𝑥) + 𝑃6 (𝑥)) 𝑑𝑥
11 11 15 15
6 7 1 6 7 2 28
= 11 × 15 ∫−1 𝑃6 (𝑥)𝑃6 (𝑥)𝑑𝑥 = 11 × 15 × 13 = 715

1 1
(vi) ∫–1 𝑃1 (𝑥)𝑃3 (𝑥)𝑃4 (𝑥)𝑑𝑥 = ∫–1 𝑥𝑃3 (𝑥)𝑃4 (𝑥)𝑑𝑥

To solve this integration, we must replace 𝑥𝑃3 (𝑥) by using the recurrence relation

(𝑛 + 1) 𝑛
𝑥 𝑃𝑛 (𝑥) = 𝑃𝑛+1 (𝑥) + 𝑃 (𝑥).
(2𝑛 + 1) (2𝑛 + 1) 𝑛−1

Then,
4 3
𝑥𝑃3 (𝑥) = 7 𝑃4 (𝑥) + 7 𝑃2 (𝑥)

1 1 4 3
So, ∫–1 𝑥𝑃3 (𝑥)𝑃4 (𝑥)𝑑𝑥 = ∫−1 ( 𝑃4 (𝑥) + 𝑃2 (𝑥)) 𝑃4 (𝑥)𝑑𝑥
7 7
1 4 1 3 4 2 8
= ∫−1 7 𝑃4 (𝑥)𝑃4 (𝑥)𝑑𝑥 + ∫−1 7 𝑃2 (𝑥)𝑃4 (𝑥)𝑑𝑥 = (7) (9) + 0 = 63

1
(vii) ∫–1(1 − 2𝑥 + 𝑥 3 )𝑃3 (𝑥)𝑑𝑥

To solve this integration, we must express (1 − 2𝑥 + 𝑥 3 ) in terms of Legendre’s polynomials by

1
using the relations 1 = 𝑃0 (𝑥), 𝑥 = 𝑃1 (𝑥) and 𝑥 3 = 5 (2𝑃3 (𝑥) + 3𝑃1 (𝑥)). Then
1 7 2
1 − 2𝑥 + 𝑥 3 = 𝑃0 (𝑥) − 2𝑃1 (𝑥) + (2𝑃3 (𝑥) + 3𝑃1 (𝑥)) = 𝑃0 (𝑥) − 𝑃1 (𝑥) + 𝑃3 (𝑥)
5 5 5
and
1 1 7 2
∫–1(1 − 2𝑥 + 𝑥 3 )𝑃3 (𝑥)𝑑𝑥 = ∫–1 (𝑃0 (𝑥) − 5 𝑃1 (𝑥) + 5 𝑃3 (𝑥)) 𝑃3 (𝑥)𝑑𝑥
2 1 2 2 4
= 5 ∫–1 𝑃3 (𝑥)𝑃3 (𝑥)𝑑𝑥 = 5 × 7 = 35

- 29 -
Chapter 2 Special Functions
1 𝑥2
(viii) ∫−1 √5−4𝑥 𝑑𝑥
1
To solve this problem, we need to replace in terms of Legendre’s polynomials using the
√5−4𝑥
generating function as follows

Assume

1 1 1
= ⇒ √𝛼 √1 − 2𝑥𝑡 + 𝑡 2 = √5 − 4𝑥
√5 − 4𝑥 √𝛼 √1 − 2𝑥𝑡 + 𝑡 2
or

𝛼(1 − 2𝑥𝑡 + 𝑡 2 ) = 5 − 4𝑥

By comparing the coefficients of 𝑥 in both sides of the pervious relation yields,

Coeff. of 𝑥 1 : − 2𝛼𝑡 = −4 ⇒ 𝛼𝑡 = 2 ----(1)
Coeff. of 𝑥 0 : 𝛼(1 + 𝑡 2 ) = 5 ----(2)

By dividing equation (1) by equation (2) yields,

𝑡 2
= ⇒ 5𝑡 = 2𝑡 2 + 2 ⇒ 2𝑡 2 − 5𝑡 + 2 = 0 ⇒ (2𝑡 − 1)(𝑡 − 1) = 0
1 + 𝑡2 5
1
The corrected value 𝑡 = 2 and the rejected value is 𝑡 = 2. By substituting into (1) yields 𝛼 = 4, then
∞
1 1 1 1 1 1 1 𝑛
= = = ∑ ( ) 𝑃𝑛 (𝑥)
√5 − 4𝑥 √𝛼 √1 − 2𝑥𝑡 + 𝑡 2 √4 5 2 2
√ −𝑥 𝑛=0
4
1 1 1 ∞
𝑥2 1 1 1 𝑛
∫ 𝑑𝑥 = ∫ ( ) 𝑥 𝑑𝑥 = ∫ ( ∑ ( ) 𝑃𝑛 (𝑥) ) 𝑥 2 𝑑𝑥
2
√5 − 4𝑥 √5 − 4𝑥 2 2
−1 −1 𝑛=0 −1
1
1 1 1 1
= ∫ (𝑃0 (𝑥) + 𝑃1 (𝑥) + 𝑃2 (𝑥) + ⋯ ) × (2𝑃2 (𝑥) + 𝑃0 (𝑥)) 𝑑𝑥
2 2 4 3
−1
1
1 1
= ∫ (𝑃0 (𝑥) + 𝑃2 (𝑥)) (2𝑃2 (𝑥) + 𝑃0 (𝑥)) 𝑑𝑥
6 4
−1
1 1
1 1 1 2 1 2 11
= ∫ 𝑃0 (𝑥)𝑃0 (𝑥) 𝑑𝑥 + ∫ 𝑃2 (𝑥)𝑃2 (𝑥) 𝑑𝑥 = × + × =
6 12 6 1 12 5 30
−1 −1

1
𝑃1 (𝑥) 𝑃3 (𝑥)
(ix) ∫ 𝑑𝑥
√58−42𝑥
−1
1
To solve this problem, we need to replace in terms of Legendre’s polynomials using the
√58−42𝑥
generating function as follows

Assume

- 30 -
Chapter 2 Special Functions
1 1 1
= ⇒ √𝛼 √1 − 2𝑥𝑡 + 𝑡 2 = √58 − 42𝑥
√58 − 42𝑥 √𝛼 √1 − 2𝑥𝑡 + 𝑡 2
or

𝛼(1 − 2𝑥𝑡 + 𝑡 2 ) = 58 − 42𝑥

By comparing the coefficients of 𝑥 in both sides of the pervious relation yields,

Coeff. of 𝑥 1 : − 2𝛼𝑡 = −42 ⇒ 𝛼𝑡 = 21 ----(1)
Coeff. of 𝑥 0 : 𝛼(1 + 𝑡 2 ) = 58 ----(2)

By dividing equation (1) by equation (2) yields,

𝑡 21
= ⇒ 58𝑡 = 21𝑡 2 + 21 21𝑡 2 − 58𝑡 + 21 = 0 ⇒ (7𝑡 − 3)(3𝑡 − 7) = 0
1 + 𝑡 2 58
3 7
The corrected value 𝑡 = 7 and the rejected value is 𝑡 = 3. By substituting into (1) yields 𝛼 = 49, then
∞
1 1 1 1 1 1 3 𝑛
= = = ∑ ( ) 𝑃𝑛 (𝑥)
√58 − 42𝑥 √𝛼 √1 − 2𝑥𝑡 + 𝑡 2 √49 58 6 7 7
√ − 𝑥 𝑛=0
49 7
1 1
𝑃1 (𝑥) 𝑃3 (𝑥) 1
∫ 𝑑𝑥 = ∫ ( ) 𝑃1 (𝑥) 𝑃3 (𝑥) 𝑑𝑥
√58 − 42𝑥 √58 − 42𝑥
−1 −1
1 ∞
1 3 𝑛 4 3
= ∫ ( ∑ ( ) 𝑃𝑛 (𝑥) ) ( 𝑃4 (𝑥) + 𝑃2 (𝑥)) 𝑑𝑥
7 7 7 7
−1 𝑛=0
1
1 3 9 27 81
= ∫ (𝑃0 (𝑥) + 𝑃1 (𝑥) + 𝑃2 (𝑥) + 3 𝑃3 (𝑥) + 4 𝑃4 (𝑥) + ⋯ )
7 7 49 7 7
−1

4 3
× ( 𝑃4 (𝑥) + 𝑃2 (𝑥)) 𝑑𝑥
7 7
1
1 1
= ∫ (𝑃0 (𝑥) + 𝑃2 (𝑥)) (2𝑃2 (𝑥) + 𝑃0 (𝑥)) 𝑑𝑥
7 4
−1
1 1
1 81 4 1 9 3
= × 4 × ∫ 𝑃4 (𝑥)𝑃4 (𝑥) 𝑑𝑥 + × × ∫ 𝑃2 (𝑥)𝑃2 (𝑥) 𝑑𝑥
7 7 7 7 49 7
−1 −1

1 81 4 2 1 9 3 2 3006
= × 4× × + × × × =
7 7 7 9 7 49 7 5 588245

Example 3 Prove that

𝟏 𝟐𝒏(𝒏+𝟏)
∫−𝟏 𝒙𝟐 𝑷𝒏+𝟏 (𝒙)𝑷𝒏−𝟏 (𝒙)𝒅𝒙 = (𝟒𝒏𝟐 −𝟏)(𝟐𝒏+𝟑)

Solution

To solve this integration, we must replace 𝑥𝑃𝑛+1 (𝑥) and 𝑥𝑃𝑛−1 (𝑥) by using the recurrence relation

- 31 -
Chapter 2 Special Functions
(𝑛 + 1) 𝑛
𝑥 𝑃𝑛 (𝑥) = 𝑃𝑛+1 (𝑥) + 𝑃 (𝑥).
(2𝑛 + 1) (2𝑛 + 1) 𝑛−1

Then,
𝑛+2 𝑛+1 𝑛 𝑛−1
𝑥𝑃𝑛+1 (𝑥) = 2𝑛+3 𝑃𝑛+2 (𝑥) + 2𝑛+3 𝑃𝑛 (𝑥) and 𝑥𝑃𝑛−1 (𝑥) = 2𝑛−1 𝑃𝑛 (𝑥) + 2𝑛−1 𝑃𝑛−2 (𝑥)

So,
1 1 𝑛+2 𝑛+1 𝑛 𝑛−1
∫−1 𝑥 2 𝑃𝑛+1 (𝑥)𝑃𝑛−1 (𝑥)𝑑𝑥 = ∫−1 (2𝑛+3 𝑃𝑛+2 (𝑥) + 2𝑛+3 𝑃𝑛 (𝑥)) (2𝑛+1 𝑃𝑛 (𝑥) + 2𝑛+1 𝑃𝑛−2 (𝑥)) 𝑑𝑥
𝑛+1 𝑛 1 𝑛+1 𝑛 2 2𝑛(𝑛+1)
= 2𝑛+3 × 2𝑛+1 ∫−1 𝑃𝑛 (𝑥)𝑃𝑛 (𝑥)𝑑𝑥 = 2𝑛+3 × 2𝑛−1 × 2𝑛+1 = (4𝑛2 −1)(2𝑛+3)

Sheet 4
Legendre Polynomials

1. Expand the following functions in series of Legendre polynomials

(i) 𝑓(𝑥) = 6𝑥 2 + 9𝑥

(ii) 𝑓(𝑥) = 50 + 11𝑥 + 𝑥 2

(iii) 𝑓(𝑥) = 𝑥 4 − 3𝑥 2 + 𝑥

2. Find the general solution of the following differential equations in terms of 𝑷𝒏 (𝒙) and 𝑸𝒏 (𝒙)

(i) (1 − 𝑥 2 )𝑦 ′′ − 2𝑥𝑦 ′ = 0

(ii) (1 − 𝑥 2 )𝑦 ′′ − 2𝑥𝑦 ′ + 42𝑦 = 0

(iii) (1 − 𝑥 2 )𝑦 ′′ − 2𝑥𝑦 ′ + 30𝑦 = 0

3. Find the value of 𝒂 and 𝒃 such that 𝒚 = 𝒂𝒙𝟑 + 𝒃𝒙 is a solution of the Legendre differential
equation (𝟏 − 𝒙𝟐 )𝒚′′ − 𝟐𝒙𝒚′ + 𝟏𝟐𝒚 = 𝟎.

4. Evaluate the following integrals

1 1 1
(i) ∫−1 𝑃2 (𝑥)𝑃4 (𝑥)𝑑𝑥 (ii) ∫−1 𝑃22 (𝑥)𝑑𝑥 (iii) ∫−1 𝑥𝑃5 (𝑥)𝑑𝑥

1 1 1
(iv) ∫−1(2𝑥 2 − 𝑥 + 3)𝑃1 (𝑥) 𝑑𝑥 (v) ∫−1(3𝑥 2 − 2𝑥 + 1)𝑃2 (𝑥) 𝑑𝑥 (vi) ∫−1 𝑥 2 𝑃5 (𝑥)𝑃7 (𝑥) 𝑑𝑥

1 1 1 𝑥2
(vii) ∫−1 𝑥 2 𝑃4 (𝑥)𝑃6 (𝑥) 𝑑𝑥 (vii) ∫−1 𝑥 2 𝑃6 (𝑥)𝑃8 (𝑥) 𝑑𝑥 (viii) ∫−1 𝑑𝑥
√5−4𝑥
1
1 1 𝑃 (𝑥) 𝑃5 (𝑥) −5−7𝑥−4𝑥 2
(ix) ∫−1(𝑥 2 + 1) 𝑃4 (𝑥) 𝑃6 (𝑥)𝑑𝑥 (x) ∫−1 1√2−2𝑥 𝑑𝑥 (xi) ∫ 𝑑𝑥
√53−28𝑥
−1
1 1 1 1−𝑥 3
(xii) ∫−1 𝑃12 (𝑥)𝑃5 (𝑥)𝑃7 (𝑥)𝑑𝑥 (xiii) ∫−1 𝑥 3 𝑃1 (𝑥)𝑃4 (𝑥)𝑑𝑥 (xiv) ∫−1 3 𝑑𝑥
(1−𝑥)2

- 32 -
Chapter 2 Special Functions

5. Prove that
1
2𝑛(𝑛 + 1)
(i) ∫ 𝑥 2 𝑃𝑛+1 (𝑥)𝑃𝑛−1 (𝑥)𝑑𝑥 =
−1 (4𝑛2 − 1)(2𝑛 + 3)
1
2𝑛
(ii) ∫ 𝑥𝑃𝑛 (𝑥)𝑃𝑛−1 (𝑥)𝑑𝑥 =
−1 (4𝑛2 − 1)
1
1 3 1
(iii) ∫ 𝑥 2 𝑃𝑛2 (𝑥)𝑑𝑥 = + +
−1 8(2𝑛 − 1) 4(2𝑛 + 1) 8(2𝑛 + 3)

- 33 -