Next Article in Journal
Modified Relaxed CQ Iterative Algorithms for the Split Feasibility Problem
Next Article in Special Issue
A New Representation of the k-Gamma Functions
Previous Article in Journal
Existence and Stability Analysis for Fractional Differential Equations with Mixed Nonlocal Conditions
Previous Article in Special Issue
Further Extension of the Generalized Hurwitz-Lerch Zeta Function of Two Variables
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Further Extension for Ramanujan’s Beta Integral and Applications

1
College of Mathematics and Physics, Chongqing University of Science and Technology, Chongqing Higher Education Mega Center, Huxi Campus, Chongqing 401331, China
2
Department of Mathematics, Chongqing Normal University, Chongqing Higher Education Mega Center, Huxi Campus, Chongqing 401331, China
*
Author to whom correspondence should be addressed.
Mathematics 2019, 7(2), 118; https://doi.org/10.3390/math7020118
Submission received: 26 December 2018 / Revised: 19 January 2019 / Accepted: 21 January 2019 / Published: 23 January 2019
(This article belongs to the Special Issue Special Functions and Applications)

Abstract

:
In 1915, Ramanujan stated the following formula 0 t x 1 ( a t ; q ) ( t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) , where 0 < q < 1 , x > 0 , and 0 < a < q x . The above formula is called Ramanujan’s beta integral. In this paper, by using q-exponential operator, we further extend Ramanujan’s beta integral. As some applications, we obtain some new integral formulas of Ramanujan and also show some new representation with gamma functions and q-gamma functions.

1. Introduction, Preliminaries and Main Results

The gamma function is the most natural extension of the factorial
n ! = 1 · 2 n .
Euler’s original definition is
Γ ( x + 1 ) = k = 1 k k + x k + 1 k x .
The integral representation of the gamma function is the following form
Γ ( x ) = 0 t x 1 e t d t , ( x ) > 0 .
The q-shifted factorials are defined by
( a ; q ) 0 = 1 , ( a ; q ) n = k = 0 n 1 ( 1 a q k ) ,
( a ; q ) = lim n k = 0 n 1 ( 1 a q k ) = k = 0 ( 1 a q k ) , n 1 .
Clearly,
( a ; q ) n = ( a ; q ) ( a q n ; q ) .
Analogously with Γ ( x ) , F. H. Jackson [1] defined Γ q ( x ) by
Γ q ( x ) = ( q ; q ) ( q x ; q ) ( 1 q ) 1 x , 0 < q < 1 .
Γ q ( x ) is called the q-gamma function.
The functional equation for Γ ( x ) ,
Γ ( x + 1 ) = x Γ ( x ) ,
becomes
Γ q ( x + 1 ) = 1 q x 1 q Γ q ( x )
for the q-gamma function. In the future, we will always take 0 < q < 1 .
We also adopt the following compact notations for the multiple q-shifted factorials:
( a 1 , a 2 , , a m ; q ) n = ( a 1 ; q ) n ( a 2 ; q ) n ( a m ; q ) n , ( a 1 , a 2 , , a m ; q ) = ( a 1 ; q ) ( a 2 ; q ) ( a m ; q ) .
The basic hypergeometric series, or q-series r ϕ s is usually defined by
r ϕ s a 1 , a 2 , , a r b 1 , , b s ; q , z = n = 0 ( a 1 ; q ) n ( a 2 ; q ) n ( a r ; q ) n ( q , b 1 ; q ) n ( b 2 ; q ) n ( b s ; q ) n ( 1 ) n q n 2 1 + s r z n ,
with n 2 = n ( n 1 ) / 2 , where q 0 , when r > s + 1 . Clearly, we have
r + 1 ϕ r a 1 , a 2 , , a r + 1 b 1 , b 2 , , , b r ; q , z = n = 0 ( a 1 , a 2 , , a r + 1 ; q ) n ( q , b 1 , b 2 , , b r ; q ) n z n .
The usual q-differential operator, or q-derivative operator D q is defined by (see ([2], p. 177, (2.1)) or [1,3,4,5]).
D q { f ( a ) } = f ( a ) f ( a q ) a ,
D q n { f ( a ) } = D q D q ( n 1 ) { f ( a ) } .
The q-shift operator η is (see ([6], p. 112)):
η { f ( a ) } = f ( a q ) ,
η 1 { f ( a ) } = f ( a q 1 ) .
The operator θ is (see [7]):
θ = η 1 D q .
The q-exponential operator E ( b θ ) is defined by (see ([6], p. 112))
E ( b θ ) = n = 0 ( b θ ) n q n 2 ( q ; q ) n .
Recently, Fang further generalized the q-exponential operator E ( b θ ) in the following form (see [8], or ([9], p. 1394, Equation (5))):
1 ϕ 0 b ; q , c θ = n = 0 ( b ; q ) n ( c θ ) n ( q ; q ) n
and obtained two q-operator identities as follows:
1 ϕ 0 b ; q , c θ { ( a s ; q ) } = ( a s , b c s ; q ) ( c s ; q ) ,
1 ϕ 0 b ; q , c θ ( a s ; q ) ( a ω ; q ) = ( a s ; q ) ( a ω ; q ) 2 ϕ 1 b , s / ω q / a ω ; q , q c / a .
In 1915, Ramanujan stated the following formula in [10,11]:
0 t x 1 ( a t ; q ) ( t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) ,
where 0 < q < 1 , x > 0 , and 0 < a < q x . The right-hand side must be interpreted using a limit when x is an integer. The above formula is called Ramanujan’s beta integral.
Hardy gave the first proof of (20) in [12]. He closed this paper with the evaluation of “another curious integral”, which is another important integral. Hardy gave a nice treatment of Ramanujan’s method of evaluating integrals of this type in his book on Ramanujan [13]. Rahman and Suslov gave a simple proof of (20) in ([14], pp. 109–110) by Ramanujan’s sum formula 1 ψ 1 . Askey ([15], p. 349) gave an elementary proof of (20) and obtained the following formula when a = q x + y in (20):
0 t x 1 ( t q x + y ; q ) ( t ; q ) d t = Γ q ( y ) Γ ( x ) Γ ( 1 x ) Γ q ( x + y ) Γ q ( 1 x )
in terms of the q-gamma function and the ordinary gamma function. When q 1 , this reduces to
0 t x 1 ( 1 + t ) x + y d t = Γ ( x ) Γ ( y ) Γ ( x + y ) = B ( x , y ) ,
where B ( x , y ) denotes the beta function defined by
B ( x , y ) = 0 1 t x 1 ( 1 t ) y 1 d t .
Recently, Chen and Liu ([6], p. 123. Equation (7.3)) gave an extension of (20) by the method of the operator as follows:
0 t x 1 ( a t , b t ; q ) ( t , a b q x 1 t ; q ) d t = π sin π x ( q 1 x , a , b ; q ) ( q , a q x , b q x ; q ) .
The aim of the present paper is to further generalize Ramanujan’s beta integral by the operator 1 ϕ 0 b ; q , c θ and to give some new formulas of Ramanujan’s beta integral. We also show the connections with gamma functions and q-gamma functions.
We now state our result as follows:
Theorem 1.
If 0 < q < 1 , x > 0 , 0 < a , a 1 , a 2 , , a 2 r + 2 < q x , | q a 2 / a | < 1 and | q a 2 j + 2 / a 2 j 1 | < 1 ( j = 1 , 2 , , r ) ; a 1 , a 2 a , a l + 2 , a l + 3 a l ( l = 1 , 3 , 5 , , 2 r 1 ) , then we have
0 t x 1 ( a t , a 1 t , a 3 t , , a 2 r + 1 t ; q ) ( t , a 2 t , a 4 t , , a 2 r + 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x , a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k × k 1 + k 2 + + k r = k 0 k r k r 1 k 2 k 1 k j = 1 r ( q k j 1 , a 2 j + 1 / a 2 j + 2 ; q ) k j ( q a 2 j + 2 / a 2 j 1 ) k j ( q , q 1 k j 1 a 2 j / a 2 j 1 ; q ) k j .

2. Proof of Theorem 1

Proof of Theorem 1.
Firstly, we write Ramanujan’s formula as follows:
0 t x 1 ( a t ; q ) ( t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) .
Applying the operator
1 ϕ 0 a 1 ; q , a 2 θ
on both sides of (26) with respect to variable a, we obtain
0 t x 1 1 ( t ; q ) 1 ϕ 0 a 1 ; q , a 2 θ { ( a t ; q ) } d t = π sin π x ( q 1 x ; q ) ( q ; q ) 1 ϕ 0 a 1 ; q , a 2 θ ( a ; q ) ( a q x ; q ) .
By (18) and (19), we get that
0 t x 1 ( a t , a 1 a 2 t ; q ) ( t , a 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) 2 ϕ 1 a 1 , q x q 1 + x / a ; q , q a 2 / a .
By (5), we rewrite the formula (28) in the following form:
0 t x 1 ( a t ; q ) ( t , a 2 t ; q ) ( a 1 a 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k ( a 1 ; q ) ( a 1 q k ; q ) .
Next, by applying the operator
1 ϕ 0 a 3 ; q , a 4 θ
on both sides of the Equation (29) with respect to variable a 1 , we arrive at
0 t x 1 ( a t , a 1 a 2 t , a 2 a 3 a 4 t ; q ) ( t , a 2 t , a 2 a 4 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x , a 1 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k k 1 = 0 k ( q k , a 3 ; q ) k 1 ( q a 4 / a 1 ) k 1 ( q , q 1 k / a 1 ; q ) k 1 .
We rewrite (30) in the following form:
0 t x 1 ( a t , a 1 a 2 t ; q ) ( t , a 2 t , a 2 a 4 t ; q ) ( a 2 a 3 a 4 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x , a 1 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k k 1 = 0 k ( q k ; q ) k 1 ( q a 4 / a 1 ) k 1 ( q , q 1 k / a 1 ; q ) k 1 ( a 3 ; q ) ( a 3 q k 1 ; q ) .
Applying the operator
1 ϕ 0 a 5 ; q , a 6 θ
on both sides of the Equation (31) with respect to variable a 3 , we have
0 t x 1 ( a t , a 1 a 2 t , a 2 a 3 a 4 t , a 2 a 4 a 5 a 6 t ; q ) ( t , a 2 t , a 2 a 4 t , a 2 a 4 a 6 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) × k = 0 ( q x , a 1 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k k 1 = 0 k ( q k , a 3 ; q ) k 1 ( q a 4 / a 1 ) k 1 ( q , q 1 k / a 1 ; q ) k 1 k 2 = 0 k 1 ( q k 1 , a 5 ; q ) k 2 ( q a 6 / a 3 ) k 2 ( q , q 1 k 1 / a 3 ; q ) k 2 .
By the mathematical induction, iterating r + 1 times, and applying the operator
1 ϕ 0 a 2 r + 1 ; q , a 2 r + 2 θ
and noting that (18) and (19), we obtain
0 t x 1 ( a t , a 1 a 2 t , a 2 a 3 a 4 t , a 2 a 4 a 5 a 6 t , , a 2 a 4 a 6 a 8 a 2 r 2 a 2 r a 2 r + 1 a 2 r + 2 t ; q ) ( t , a 2 t , a 2 a 4 t , a 2 a 4 a 6 t , , a 2 a 4 a 6 a 8 a 2 r 2 a 2 r a 2 r + 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x , a 1 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k k 1 = 0 k ( q k , a 3 ; q ) k 1 ( q a 4 / a 1 ) k 1 ( q , q 1 k / a 1 ; q ) k 1 × k 2 = 0 k 1 ( q k 1 , a 5 ; q ) k 2 ( q a 6 / a 3 ) k 2 ( q , q 1 k 1 / a 3 ; q ) k 2 k r = 0 k r 1 ( q k r 1 , a 2 r + 1 ; q ) k r ( q a 2 r + 2 / a 2 r 1 ) k r ( q , q 1 k r 1 / a 2 r 1 ; q ) k r .
Letting a 2 j 1 a 2 j 1 / a 2 j and a 2 j + 2 a 2 j + 2 / a 2 j ( j = 1 , 2 , r ) in (33), we show that
0 t x 1 ( a t , a 1 t , a 3 t , , a 2 r + 1 t ; q ) ( t , a 2 t , a 4 t , , a 2 r + 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x , a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k × k 1 + k 2 + + k r = k 0 k r k r 1 k 2 k 1 k j = 1 r ( q k j 1 , a 2 j + 1 / a 2 j + 2 ; q ) k j ( q a 2 j + 2 / a 2 j 1 ) k j ( q , q 1 k j 1 a 2 j / a 2 j 1 ; q ) k j .
The proof of Theorem 1 is complete.

3. Some Applications

In this section, we will obtain the corresponding new integral formulas from (25).
Taking r = 0 in (25) and defining the empty sum equal to 1, we obtain the following integral formula:
Corollary 1.
For 0 < q < 1 , x > 0 , 0 < a , a 1 , a 2 < q x and | q a 2 / a | < 1 ; a 1 , a 2 a , we have
0 t x 1 ( a t , a 1 t ; q ) ( t , a 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) 2 ϕ 1 q x , a 1 / a 2 q 1 + x / a ; q , q a 2 / a .
Remark 1.
If setting a 2 = a a 1 q x 1 in (35), we get
0 t x 1 ( a t , a 1 t ; q ) ( t , a a 1 q x 1 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) 1 ϕ 0 q x ; q , a 1 q x .
Applying the q-binomial theorem (see ([16], p. 8. (1.3.2)))
1 ϕ 0 a ; q , z = ( a z ; q ) ( z ; q )
in (36), we obtain
0 t x 1 ( a t , a 1 t ; q ) ( t , a a 1 q x 1 t ; q ) d t = π sin π x ( q 1 x , a , a 1 ; q ) ( q , a q x , a 1 q x ; q ) .
Setting a 1 = b in (38), we obtain (24) immediately. Hence, we say that the formula (35) is an extension of result (24) of Chen and Liu.
Corollary 2.
For 0 < q < 1 , x > 0 , 0 < a , b < q x ; a b , we have
0 t x 1 ( a t ; q ) ( b t ; q ) d t = π sin π x ( q 1 x , a , q / a , b q 1 + x / a ; q ) ( q , a q x , q b / a , q 1 + x / a ; q ) .
Proof. 
Setting a 1 1 , a 2 b in (35) and applying q-Gauss sum formula ([16], p. 14, Equation (1.5.1)):
2 ϕ 1 a , b c ; q , c / a b = ( c / a , c / b ; q ) ( c , c / a b ; q ) , | c / a b | < 1 ,
we obtain (39).
Remark 2.
If letting b 1 in (39), we obtain (20) immediately. Hence, the formula (39) is also an extension of (20):
Taking r = 1 in (25), we have
Corollary 3.
For 0 < q < 1 , x > 0 , 0 < a , a 1 , a 2 , a 3 , a 4 < q x , | q a 2 / a | < 1 and | q a 4 / a 1 | < 1 ; a 1 , a 2 a , a 3 , a 4 a 1 , we have
0 t x 1 ( a t , a 1 t , a 3 t ; q ) ( t , a 2 t , a 4 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) k = 0 ( q x , a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q , q 1 + x / a ; q ) k 2 ϕ 1 q k , a 3 / a 4 a 2 q 1 k / a 1 ; q , q a 4 / a 1 .
Theorem 2.
If 0 < q < 1 , 0 < a , a 1 , a 2 , , a 2 r + 2 < q n ( n = 1 , 2 , ) , | q a 2 / a | < 1 and | q a 2 j + 2 / a 2 j 1 | < 1 , ( j = 1 , 2 , , r ) ; a 1 , a 2 a , a l + 2 , a l + 3 a l ( l = 1 , 3 , 5 , , 2 r 1 ) , then we have
0 t n 1 ( a t , a 1 t , a 3 t , , a 2 r + 1 t ; q ) ( t , a 2 t , a 4 t , , a 2 r + 2 t ; q ) d t = ( 1 ) n 1 ( q ; q ) n q n log q ( a 1 q ; q ) n a n ( 1 q n ) k = 0 ( q n , a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q , q 1 + n / a ; q ) k × k 1 + k 2 + + k r = k 0 k r k r 1 k 2 k 1 k j = 1 r ( q k j 1 , a 2 j + 1 / a 2 j + 2 ; q ) k j ( q a 2 j + 2 / a 2 j 1 ) k j ( q , q 1 k j 1 a 2 j / a 2 j 1 ; q ) k j .
Proof. 
By (4), we easily obtain
( q 1 x ; q ) = ( 1 q 1 x ) ( 1 q 2 x ) ( 1 q n 1 x ) ( 1 q n x ) ( 1 q n + 1 x ) ( 1 q n + 2 x ) ,
( a q x ; q ) = ( 1 a q x ) ( 1 a q 1 x ) ( 1 a q 2 x ) ( 1 a q n 1 x ) ( 1 a q n x ) ( 1 a q n + 1 x ) ( 1 a q n + 2 x ) .
Noting (42), (43) and using the L’Hospital rule, via some simple calculation, we get
lim x n π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) = ( a ; q ) ( q ; q ) lim x n π sin π x ( q 1 x ; q ) ( a q x ; q ) = ( a ; q ) ( q ; q ) ( 1 ) n 1 q 1 2 ( n 1 ) ( 1 q ) ( 1 q 2 ) ( 1 q n 1 ) ( 1 q ) ( 1 q 2 ) ( 1 ) n q 1 2 n ( a q ) ( a q 2 ) ( a q n ) ( 1 a ) ( 1 a q ) ( 1 a q 2 ) lim x n π ( 1 q n x ) sin π x = ( a ; q ) ( q ; q ) ( 1 q ) ( 1 q 2 ) ( 1 q n 1 ) ( q ; q ) q n ( a q ) ( a q 2 ) ( a q n ) ( a ; q ) log q cos n π = ( 1 ) n 1 ( q ; q ) n q n log q ( a 1 q ; q ) n a n ( 1 q n ) .
Letting x n on both sides of (25) and using the above limit, we obtain (41) immediately.
Taking r = 0 in (41) and defining the empty sum equal to 1, we obtain the following integral formula:
Corollary 4.
For 0 < q < 1 , 0 < a , a 1 , a 2 < q n ( n = 1 , 2 , ) and | q a 2 / a | < 1 ; a 1 , a 2 a , we have
0 t n 1 ( a t , a 1 t ; q ) ( t , a 2 t ; q ) d t = ( 1 ) n 1 ( q ; q ) n q n log q ( a 1 q ; q ) n a n ( 1 q n ) k = 0 ( q n , a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q , q 1 + n / a ; q ) k .
Remark 3.
Taking a 1 = a 2 in (44), we deduce
0 t n 1 ( a t ; q ) ( t ; q ) d t = ( 1 ) n 1 ( q ; q ) n q n log q ( a 1 q ; q ) n a n ( 1 q n ) = ( q ; q ) n 1 q n log q ( q a ) ( q 2 a ) ( q n a ) ,
which is exactly Askey’s result in ([15], p. 349, (2.9)).
Taking r = 1 in (41), we get
Corollary 5.
For 0 < q < 1 , 0 < a , a 1 , a 2 , a 3 , a 4 < q n ( n = 1 , 2 , ) , | q a 2 / a | < 1 and | q a 4 / a 1 | < 1 ; a 1 , a 2 a , a 3 , a 4 a 1 , we have
0 t n 1 ( a t , a 1 t , a 3 t ; q ) ( t , a 2 t , a 4 t ; q ) d t = ( q ; q ) n q n log q ( a q ) ( a q 2 ) ( a q n ) ( 1 q n ) k = 0 ( q n , a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q , q 1 + n / a ; q ) k 2 ϕ 1 q k , a 3 / a 4 a 2 q 1 k / a 1 ; q , q a 4 / a 1 .
Theorem 3.
If 0 < q < 1 , 0 < a , a 1 , a 2 , , a 2 r + 2 < q , | q a 2 / a | < 1 and | q a 2 j + 2 / a 2 j 1 | < 1 ( j = 1 , 2 , , r ) ; a 1 , a 2 a , a l + 2 , a l + 3 a l ( l = 1 , 3 , 5 , , 2 r 1 ) , then we have
0 ( a t , a 1 t , a 3 t , , a 2 r + 1 t ; q ) ( t , a 2 t , a 4 t , , a 2 r + 2 t ; q ) d t = q log q a q k = 0 ( a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q 2 / a ; q ) k × k 1 + k 2 + + k r = k 0 k r k r 1 k 2 k 1 k j = 1 r ( q k j 1 , a 2 j + 1 / a 2 j + 2 ; q ) k j ( q a 2 j + 2 / a 2 j 1 ) k j ( q , q 1 k j 1 a 2 j / a 2 j 1 ; q ) k j .
Proof. 
Taking n = 1 in (41), we obtain (47) immediately.
Taking r = 0 in (47) and defining the empty sum equal to 1, we obtain the following integral formula:
Corollary 6.
For 0 < q < 1 , 0 < a , a 1 , a 2 < q and | q a 2 / a | < 1 ; a 1 , a 2 a , we have
0 ( a t , a 1 t ; q ) ( t , a 2 t ; q ) d t = q log q a q 2 ϕ 1 q , a 1 / a 2 q 2 / a ; q , q a 2 / a .
Taking r = 1 in (47), we deduce
Corollary 7.
For 0 < q < 1 , 0 < a , a 1 , a 2 , a 3 , a 4 < q , | q a 2 / a | < 1 and | q a 4 / a 1 | < 1 ; a 1 , a 2 a , a 3 , a 4 a 1 , we have
0 ( a t , a 1 t , a 3 t ; q ) ( t , a 2 t , a 4 t ; q ) d t = q log q a q k = 0 ( a 1 / a 2 ; q ) k ( q a 2 / a ) k ( q 2 / a ; q ) k 2 ϕ 1 q k , a 3 / a 4 a 2 q 1 k / a 1 ; q , q a 4 / a 1 .

4. Connections with the q-Gamma Function

In this section, we give the corresponding formulas with the q-gamma function from (25).
Theorem 4.
If 0 < q < 1 , x > 0 , 0 < a 1 , a 2 , , a 2 r + 2 < q x , | a 2 / q x + y 1 | < 1 and | q a 2 j + 2 / a 2 j 1 | < 1 , ( j = 1 , 2 , , r ) ; a 1 , a 2 q x + y , a l + 2 , a l + 3 a l ( l = 1 , 3 , 5 , , 2 r 1 ) , then we have
0 t x 1 ( t q x + y , a 1 t , a 3 t , , a 2 r + 1 t ; q ) ( t , a 2 t , a 4 t , , a 2 r + 2 t ; q ) d t = Γ q ( y ) Γ q ( 1 y ) Γ ( x ) Γ ( 1 x ) Γ q ( x + y ) Γ q ( x ) Γ q ( 1 x ) k = 0 Γ q ( k + x ) ( a 1 / a 2 ; q ) k Γ q ( k + 1 y ) ( q ; q ) k ( a 2 / q x + y 1 ) k × k 1 + k 2 + + k r = k 0 k r k r 1 k 2 k 1 k j = 1 r ( q k j 1 , a 2 j + 1 / a 2 j + 2 ; q ) k j ( q a 2 j + 2 / a 2 j 1 ) k j ( q , q 1 k j 1 a 2 j / a 2 j 1 ; q ) k j .
Proof. 
Taking a = q x + y in (25) and noting that
Γ q ( x ) = ( q ; q ) ( q x ; q ) ( 1 q ) 1 x a n d Γ ( x ) Γ ( 1 x ) = π sin π x ,
we easily obtain (50).
Taking r = 0 in (50) and defining the empty sum equal to 1, we obtain the following integral formula.
Corollary 8.
For 0 < q < 1 , x > 0 , 0 < a 1 , a 2 < q x and | a 2 / q x + y 1 | < 1 ; a 1 , a 2 q x + y , we have
0 t x 1 ( t q x + y , a 1 t ; q ) ( t , a 2 t ; q ) d t = Γ q ( y ) Γ q ( 1 y ) Γ ( x ) Γ ( 1 x ) Γ q ( x + y ) Γ q ( x ) Γ q ( 1 x ) k = 0 Γ q ( k + x ) ( a 1 / a 2 ; q ) k Γ q ( k + 1 y ) ( q ; q ) k ( a 2 / q x + y 1 ) k .
Taking a 1 = a 2 in (51), we deduce the result of Askey as follows:
Corollary 9
([15], p. 350, Equation (2.10)). For 0 < q < 1 , x > 0 , we have
0 t x 1 ( t q x + y ; q ) ( t ; q ) d t = Γ q ( y ) Γ ( x ) Γ ( 1 x ) Γ q ( x + y ) Γ q ( 1 x ) .
Remark 4.
Letting a 0 and b 1 in (39), we have
0 t x 1 ( t ; q ) d t = π sin π x ( q 1 x ; q ) ( q ; q ) .
Applying
Γ q ( x ) = ( q ; q ) ( q x ; q ) ( 1 q ) 1 x a n d Γ ( x ) Γ ( 1 x ) = π sin π x ,
we obtain
0 t x 1 ( t ; q ) d t = Γ ( x ) Γ ( 1 x ) ( 1 q ) x Γ q ( 1 x ) ,
which is exactly the result of Askey (see ([15], p. 353, Equation (4.2))).

5. Conclusions

In this paper, by applying q-exponential operator
1 ϕ 0 b ; q , c θ = n = 0 ( b ; q ) n ( c θ ) n ( q ; q ) n ,
we further extend the following Ramanujan’s beta integral [10]
0 t x 1 ( a t ; q ) ( t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) .
Especially, we obtain two new integral formulas
0 t x 1 ( a t , a 1 t ; q ) ( t , a 2 t ; q ) d t = π sin π x ( q 1 x , a ; q ) ( q , a q x ; q ) 2 ϕ 1 q x , a 1 / a 2 q 1 + x / a ; q , q a 2 / a
and
0 t x 1 ( a t ; q ) ( b t ; q ) d t = π sin π x ( q 1 x , a , q / a , b q 1 + x / a ; q ) ( q , a q x , q b / a , q 1 + x / a ; q ) .
We also show that Ramanujan’s beta integral can be represented with q-gamma functions [15].

Author Contributions

Both authors contributed equally to this work. In addition, both authors read and approved the final manuscript.

Funding

This research received no external funding.

Acknowledgments

The authors would like to thank two referees for the helpful comments and suggestions which improved this paper.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Jackson, F.H. On q-definite integrals. Q. J. Math. 1910, 50, 101–112. [Google Scholar]
  2. Chen, W.Y.C.; Liu, Z.-G. Parameter augmentation for basic hypergeometric series II. J. Comb. Theory Ser. A 1997, 80, 175–195. [Google Scholar] [CrossRef]
  3. Bowman, D. q-Difference operators, orthogonal polynomials, and symmetric expansions. Mem. Am. Math. Soc. 2002, 159. [Google Scholar] [CrossRef]
  4. Rogers, L.J. On the expansion of some infinite products. Proc. Lond. Math. Soc. 1892, 24, 337–352. [Google Scholar] [CrossRef]
  5. Watson, G.N. The continuations of functions defined by generalized hypergeometric functions. Trans. Camb. Philos. Soc. 1910, 21, 281–299. [Google Scholar]
  6. Chen, W.Y.C.; Liu, Z.-G. Parameter augmentation for basic hypergeometric series I. In Mathematical Essays in Honor of Gian-Carlo Rota; Sagan, B.E., Stanley, R.P., Eds.; Birkhauser: Basel, Switzerland, 1998; pp. 111–129. [Google Scholar]
  7. Roman, S. More on the umbral caculus, with emphasis on the q-umbral calculs. J. Math. Soc. 1985, 107, 222–254. [Google Scholar]
  8. Chen, V.Y.B.; Gu, N.S.S. The Cauchy Operator for basic hypergeometric series. Adv. Appl. Math. 2008, 41, 177–196. [Google Scholar] [CrossRef]
  9. Fang, J.-P. q-differential operator identities and applications. J. Math. Anal. Appl. 2007, 333, 1393–1407. [Google Scholar] [CrossRef]
  10. Ramanujan, S. Some definite integrals. Messenger Math. 1915, 44, 10–18, reprinted in: Hardy, G.H., Seshu Aiyer, P.V.,Wilson, B.M., Eds. [Google Scholar]
  11. Ramanujan, S. Collected Papers of Srinivasa Ramanujan; Hardy, G.H., Aiyar, P.V.S., Wilson, B.M., Eds.; Cambridge University Press: Cambridge, UK, 1927; reprinted by Chelsea, New York, 1962. [Google Scholar]
  12. Hardy, G.H. Proof of a formula of Mr. Ramanujan. Messenger of Math. 1915, 44, 18–21, reprinted in Collected Papers of G.H. Hardy, vol. 5, Oxford, 1972, 594–597. [Google Scholar]
  13. Hardy, G.H. Ramanujan; Cambridge University Press: Cambridge, UK, 1940; reprinted by Chelsea, New York, 1959. [Google Scholar]
  14. Rahman, M.; Suslov, S.K. A unified approach to the summation and integration formulas for q-hypergeometric functions I. J. Stat. Plan. Inference 1996, 54, 101–118. [Google Scholar] [CrossRef]
  15. Askey, R. Ramanujan’s extension of the gamma and beta functions. Am. Math. Mon. 1980, 87, 346–359. [Google Scholar] [CrossRef]
  16. Gasper, G.; Rahma, M. Basic Hypergeometric Series; Cambridge Unversity Press: Cambridge, UK, 2004. [Google Scholar]

Share and Cite

MDPI and ACS Style

Xi, G.-W.; Luo, Q.-M. A Further Extension for Ramanujan’s Beta Integral and Applications. Mathematics 2019, 7, 118. https://doi.org/10.3390/math7020118

AMA Style

Xi G-W, Luo Q-M. A Further Extension for Ramanujan’s Beta Integral and Applications. Mathematics. 2019; 7(2):118. https://doi.org/10.3390/math7020118

Chicago/Turabian Style

Xi, Gao-Wen, and Qiu-Ming Luo. 2019. "A Further Extension for Ramanujan’s Beta Integral and Applications" Mathematics 7, no. 2: 118. https://doi.org/10.3390/math7020118

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop