Next Article in Journal
An Improved Pity Beetle Algorithm for Solving Constrained Engineering Design Problems
Previous Article in Journal
A Novel Method for Remaining Useful Life Prediction of Bearing Based on Spectrum Image Similarity Measures
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Summation Formulae for Quintic q-Series

1
School of Mathematics and Statistics, Zhoukou Normal University, Zhoukou 466001, China
2
Department of Mathematics and Physics, University of Salento, P.O. Box 193, 73100 Lecce, Italy
Mathematics 2022, 10(13), 2210; https://doi.org/10.3390/math10132210
Submission received: 3 June 2022 / Revised: 21 June 2022 / Accepted: 22 June 2022 / Published: 24 June 2022

Abstract

:
By utilizing the modified Abel lemma on summation by parts, we examine a class of quintic q-series, that have close connections to the “twisted cubic q-series”. Several remarkable summation and transformation formulae are established. The related reversal series are also reviewed briefly.

1. Introduction and Motivation

Let N be the set of natural numbers with N 0 = { 0 } N . For an indeterminate x, the rising factorial with the base q is defined by ( x ; q ) 0 = 1 and
( x ; q ) n = ( 1 x ) ( 1 q x ) ( 1 q n 1 x ) for n N .
Its multiparameter forms will be abbreviated as follows:
A , B , , C ; q n = ( A ; q ) n ( B ; q ) n ( C ; q ) n , α , β , , γ A , B , , C | q n = α ; q n β ; q n γ ; q n A ; q n B ; q n C ; q n .
The basic hypergeometric series (shortly “q-series”, cf. [1,2,3]) playsplay an important role in pure and applied sciences [4,5,6,7,8,9] for their wide applications to number theory [2,10,11], theoretical physics [12,13] and combinatorial enumeration [12,14]. During the past few decades, there has been growing interest in finding new identities and transformation formulae. In particular, there was an intensive exploration to well–poised series [15,16,17,18], quadratic series [19,20,21,22], cubic series [19,21,23,24], and quartic series [19,20,25,26], mainly made by Gasper and Rahman (see also [2], §3.8) and the author of this paper, with his collaborators.
A quarter of a century ago, the author ([27], Equations 5.4d: ε = 1 , 2 ) found, by making use of inversion techniques, the following two summation formulae
k = 0 n 1 q 6 k + 1 / 2 [ q 5 n , q 5 n + 5 / 2 ; q 5 ] k [ q 1 5 n , q 5 n + 3 / 2 ; q ] k [ q 3 / 2 , q 1 / 2 ; q ] 2 k ( q 2 ; q ) 4 k q k = ( q 5 ; q 5 ) n ( q 1 2 ; q ) 5 n + 1 ( q 5 2 ; q 5 ) n ( q 2 ; q ) 5 n , k = 0 n 1 q 6 k + 7 / 2 [ q 5 n , q 5 n + 3 / 2 ; q 5 ] k [ q 3 5 n , q 5 n + 9 / 2 ; q ] k [ q 5 / 2 , q 1 / 2 ; q ] 2 k ( q 4 ; q ) 4 k q k = ( q 5 ; q 5 ) n ( q 5 2 ; q ) 5 n + 2 ( q 5 2 ; q 5 ) n + 1 ( q 4 ; q ) 5 n .
They are isolated examples of the “quintic series” defined by the partial sums
Λ n ( a , c ) = k = 0 n 1 q 6 k 1 a 3 [ a 2 c , a 3 / c ; q 5 ] k [ c , a / c ; q ] k [ q a , a / q ; q ] 2 k ( q a 2 ; q ) 4 k q k ,
λ n ( a , c ) = k = 0 n 1 q 6 k + 1 a 3 [ q c , q a / c ; q ] k [ q 5 a 2 c , q 5 a 3 / c ; q 5 ] k ( a 2 ; q ) 4 k [ a , q 2 a ; q ] 2 k q k ;
which satisfy obviously the following properties:
Λ n ( a , c ) = Λ n ( a , a / c ) and λ n ( a , c ) = λ n ( a , a / c ) .
These two series are called “quintic series” since the factorials of order k appearing in the numerator and the denominator can be paired off in such a manner that for each pair, the product of their bases is q 6 and the product of the corresponding parameters is almost equal to “ a 3 ” apart from(with a discrepancy of a small integer power of the base “q”).
Under the replacement k n k on the summation index, these two series are reversal of each other:
Λ n ( a , c ) = q 7 n 1 a 3 λ n ( q 2 n / a , q n / c ) q a , a / q ; q 2 n a 2 c , a 3 / c ; q 5 n q a 2 ; q 4 n c , a / c ; q n ,
λ n ( a , c ) = q 7 n + 1 a 3 Λ n ( q 2 n / a , q n / c ) a 2 ; q 4 n q c , q a / c ; q n a , q 2 a ; q 2 n q 5 a 2 c , q 5 a 3 / c ; q 5 n .
Observe that for n N 0 , the Λ n -series terminates above by the numerator parameters and below by the denominator parameters in the following eight different manners:
a 2 c = q 5 n with a = 1 , q , q 1 2 , q 3 2 ; a / c = q with a = ± q 1 5 n 2 , q 1 2 n , q 2 2 n .
By means of the modified Abel lemma on summation by parts, we shall show, in the next section, a reciprocal formula (see Lemma 1) between the Λ n -series and the following “twisted cubic series” introduced recently by the author [28]:
Ω n ( a , c ) = k = 0 n 1 q 4 k a ( a / c 2 ; q ) 2 k ( q c 2 ; q ) 2 k [ c , c , q c ; ; q ] k ( q a / c ; q ) 3 k ( q a c ; q 5 ) k ( 1 / c ; q 1 ) k q k .
When the Λ n -series terminates above by a 2 c = q 5 n , we shall derive four summation formulae (see Theorems 1 and 2), respectively, terminated below by a = 1 , q and a = q 1 / 2 , q 1 / 2 . When the Λ n -series terminates below by c = a / q , we find a nonterminating series identity (see Theorem 3) and establish five summation formulae for terminating series (see Theorems 4–6). Finally in Section 3, the corresponding identities for the reversal of λ n -series will be deduced.
In order to reduce lengthy expressions, we shall utilize the following notations. As usual, the logical function χ is defined by χ ( true ) = 1 and χ ( false ) = 0 . For a real number x, the greatest integer not exceeding x will be denoted by x . When m is a natural number, i m j signifies that “i is congruent to j modulo m”. To ensure the accuracy, we have checked numerically, throughout the paper, all the displayed equations and identities by appropriately devised Mathematica commands.

2. Identities for the Λ n -Series

The approach adopted in this paper will be the modified Abel’s lemma on summation by parts (cf. Chu et al. [16,24,29,30,31]). In order to make the paper self–contained, we record it as follows. For an arbitrary complex sequence { τ k } , the backward and forward difference operators ∇ and Δ · are defined, respectively, by
τ k = τ k τ k 1 and Δ · τ k = τ k τ k + 1 ,
where Δ · is designed for convenience in the present paper, which differs from the usual operator Δ only in the minus sign. Then, Abel’s lemma on summation by parts can be modified in the following way:
k = 0 n V k U k = U n V n + 1 U 1 V 0 + k = 0 n U k Δ · V k .
In fact, it is almost routine to check the following expression:
k = 0 n V k U k = k = 0 n V k U k U k 1 = k = 0 n U k V k k = 0 n U k 1 V k .
Replacing k with k + 1 for the last sum, we can reformulate the equation as follows
k = 0 n V k U k = U n V n + 1 U 1 V 0 + k = 0 n U k V k V k + 1 = U n V n + 1 U 1 V 0 + k = 0 n U k Δ · V k
which is exactly the equality highlighted in the modified Abel lemma.

2.1. Transformation from Λ n to Ω n

For the two sequences { U k , V k } given by
U k : = ( q a ; q 2 ) k ( q 2 a ; q 2 ) k ( q 3 a 2 ; q 4 ) k ( q 4 a 2 ; q 4 ) k ( q 2 a 2 / c ; q 3 ) k ( q 3 a c ; q 3 ) k ( q 5 a 2 c ; q 5 ) k ( a / c ; q ) k , V k : = ( a ; q 2 ) k ( q a ; q 2 ) k ( q a 2 ; q 4 ) k ( q 2 a 2 ; q 4 ) k ( q 3 a c ; q 3 ) k ( q 1 a 2 / c ; q 3 ) k ( a 3 / c ; q 5 ) k ( q 2 c ; q ) k ;
it is not difficult to compute their differences
U k : = ( a ; q 2 ) k ( a / q ; q 2 ) k ( q 3 a 2 ; q 4 ) k ( q 4 a 2 ; q 4 ) k ( q 1 a 2 / c ; q 3 ) k ( q 3 a c ; q 3 ) k ( a 2 c ; q 5 ) k ( a / c ; q ) k × q k c q 6 k 1 a 3 , q 2 k a , q k c , q k + 1 c a , q / a , a 2 c , q 1 a 2 / c | q 1 , Δ · V k : = ( a ; q 2 ) k ( q a ; q 2 ) k ( q 5 a 2 ; q 4 ) k ( q 6 a 2 ; q 4 ) k ( q 3 a c ; q 3 ) k ( q 2 a 2 / c ; q 3 ) k ( a 3 / c ; q 5 ) k ( q 3 c ; q ) k × q k q 2 + 6 k a 3 , q 1 + 2 k a , q k 1 a / c , q k 2 a / c q a 2 , q 2 a 2 , q 1 a 2 / c , q 2 / c | q 1 ;
and the two boundary values
U n V n + 1 = [ q 5 a 2 c , q 5 a 3 / c ; q 5 ] n [ q 3 c , a / c ; q ] n [ q a , q 2 a ; q ] 2 n ( q 3 a 2 ; q ) 4 n × q 3 + 3 n a c , a , q a , a 3 / c q a 2 , q 2 a 2 , q 2 c , q 1 a 2 / c | q 1 , U 1 V 0 = a c , a 2 , a 2 / q , q 1 a / c a , a / q , a 2 c , q 1 a 2 / c | q 1 .
Observe that the Λ n -sum can be expressed as
Λ n ( a , c ) = k = 0 n 1 q 6 k 1 a 3 [ a 2 c , a 3 / c ; q 5 ] k [ c , a / c ; q ] k [ q a , a / q ; q ] 2 k ( q a 2 ; q ) 4 k q k = q / a , a 2 c , a 2 / c q q c , 1 / c | q 1 × k = 0 n V k U k .
Then by making use of the modified Abel lemma on summation parts, we can manipulate it as follows:
Λ n ( a , c ) = q / a , a 2 c , a 2 / q c q c , 1 / c | q 1 × U n V n + 1 U 1 V 0 + k = 0 n U k Δ · V k .
After substituting the related terms into the last expression, we can restate it as the recurrence relation below:
Λ n ( a , c ) = Λ n ( q a , q 3 c ) q a , a , a / q , a 2 c a 1 , c , q c , q 2 c | q 1 ( q c / a ; q ) 2 ( q a 2 ; q ) 2 + a c , a 2 , a 2 / q , q c / a a , c , q c | q 1 + q c a [ q 5 a 2 c , q 5 a 3 / c ; q 5 ] n [ q 3 c , a / c ; q ] n × q 3 + 3 n a c , q a , a , a / q , a 2 c , a 3 / c c , q c , q 2 c , q a 2 , q 2 a 2 | q 1 [ q a , q 2 a ; q ] 2 n ( q 3 a 2 ; q ) 4 n .
By iterating the relation m-times and then simplifying the resulting equation, we obtain the following transformation formula.
Lemma 1
(Reciprocal relation). Between the quintic Λ n -series (1) and the cubic Ω m -series (5), there holds
Λ n ( a , c ) = Λ n ( q m a , q 3 m c ) ( q c / a ; q ) 2 m [ q a , a , a / q ; q ] m ( a 2 c ; q 5 ) m ( q a 2 ; q ) 2 m ( c ; q ) 3 m ( a 1 ; q 1 ) m + t Ω m 1 ( a c , a / q ) a 2 , a 2 / q , q c / a a , c , q c | q 1 + Ω m 1 ( q 3 + 3 n a c , q 1 + 2 n a ) × q c a ( q a ; q ) 2 n + 1 ( a / q ; q ) 2 n + 2 a 2 c , a 3 / c ; q 5 n + 1 ( a / c ; q ) n ( c ; q ) n + 3 ( q a 2 ; q ) 4 n + 2 .

2.2. Case a 2 c = q 5 n

When m = n + 1 and a 2 c = q 5 n in Lemma 1, the first term and the third term on the right vanish. We find, in this case, the following simplified transformation formula.
Proposition 1
( a 2 c = q 5 n with n N 0 ).
Λ n ( a , c ) = Ω n ( a c , a / q ) a 2 , a 2 / q , q c / a a , c , q c | q 1 .
Letting a = q and a = q 3 2 in Proposition 1, we obtain two reduced relations
Λ n ( q , q 2 5 n ) = Ω n ( q 1 5 n , 1 ) q , q 2 , q 2 5 n q , q 1 5 n , q 2 5 n | q 1 , Λ n ( q 3 2 , q 3 5 n ) = Ω n ( q 3 2 5 n , q 1 2 ) q 2 , q 3 2 , q 7 2 5 n q 2 5 n , q 3 5 n | q 1 .
By combining them with the two summation formulae for the twisted cubic series (Chu [28], Corollary 3)
Ω n ( q 1 5 n , 1 ) = q 1 5 n ( q ; q ) 5 n + 1 ( q 5 ; q 5 ) n ( q ; q ) 5 n ( q 5 ; q 5 ) n , = k = 0 n 1 + q 4 k 5 n 1 ( q 5 n ; q 5 ) k ( q ; q ) 2 k ( q 1 5 n ; q ) 2 k ( q 5 n ; q ) 3 k 1 , 1 , q ; q k ( 1 ; q 1 ) k q k ,
Ω n ( q 3 2 5 n , q 1 2 ) = 1 q 5 n + 3 / 2 ( q 5 ; q 5 ) n ( q 3 / 2 ; q ) 5 n + 1 ( q 5 / 2 ; q 5 ) n ( q ; q ) 5 n + 1 , = k = 0 n 1 q 4 k 5 n 3 2 ( q 5 n ; q 5 ) k ( q ; q ) 2 k + 1 ( q 5 n 5 2 ; q ) 2 k ( q 5 n 1 ; q ) 3 k q 1 2 , q 1 2 , q 3 2 ; q k ( q 1 2 ; q 1 ) k q k .
we find the following explicit formulae.
Theorem 1
((Chu [27], Equation 5.4d: ε = 2 ) for the second identity).
( a ) Λ n ( q , q 2 5 n ) = k = 0 n 1 + q 6 k + 2 [ q 5 n , q 5 n + 5 ; q 5 ] k [ q 2 5 n , q 3 + 5 n ; q ] k [ q 2 , 1 ; q ] 2 k ( q ; q ) 4 k + 2 q k = ( q 5 ; q 5 ) n ( q 2 ; q ) 5 n + 1 ( q 5 ; q 5 ) n ( q ; q ) 5 n + 2 , ( b ) Λ n ( q 3 2 , q 3 5 n ) = k = 0 n 1 q 6 k + 7 2 [ q 5 n , q 15 2 + 5 n ; q 5 ] k [ q 3 5 n , q 9 2 + 5 n ; q ] k [ q 1 2 , q 5 2 ; q ] 2 k ( q ; q ) 4 k + 3 q k = ( q 5 ; q 5 ) n ( q 5 2 ; q ) 5 n + 2 ( q 5 2 ; q 5 ) n + 1 ( q ; q ) 5 n + 3 .
When a 1 and a q 1 2 , the factorial ( a 2 / q ; q ) 2 k becomes a zero denominator in Ω n ( a c , a / q ) . Pulling out it by shifting forward the summation index k k + 1 , we can evaluate the following limits:
Λ n ( 1 , q 5 n ) = lim a 1 Ω n ( q 5 n / a , a / q ) a 2 , a 2 / q , q 1 5 n / a 3 a , q 5 n / a 2 , q 1 5 n / a 2 | q 1 = lim a 1 Ω n 1 ( q 4 5 n / a , a ) a 2 ( 1 q / a ) ( 1 q 5 n ) ( q 1 5 n / a 3 ; q ) 3 q ( q 5 n / a 2 ; q ) 5 = Ω n 1 ( q 4 5 n , 1 ) × ( 1 + q 1 ) ( q 1 5 n ; q ) 3 ( q 1 5 n ; q ) 4 ; Λ n ( q 1 2 , q 1 5 n ) = lim a q 1 2 Ω n ( q 5 n / a , a / q ) a 2 , a 2 / q , q 1 5 n / a 3 a , q 5 n / a 2 , q 1 5 n / a 2 | q 1 = lim a q 1 2 Ω n 1 ( q 4 5 n / a , a ) a 2 ( 1 q / a ) ( 1 q 5 n ) ( q 1 5 n / a 3 ; q ) 3 q ( q 5 n / a 2 ; q ) 5 = Ω n 1 ( q 7 2 5 n , q 1 2 ) × ( 1 q 1 2 ) ( q 1 2 5 n ; q ) 3 ( 1 q 1 5 n ) ( q 1 5 n ; q ) 3 .
By making use of (7) and (8) under the replacement n n 1 , we derive two further summation formulae.
Theorem 2
((Chu [27], Equation 5.4d: ε = 1 ) for the second identity).
( a ) Λ n ( 1 , q 5 n ) = k = 0 n 1 + q 6 k 1 [ q 5 n , q 5 n ; q 5 ] k [ q 5 n , q 5 n ; q ] k [ q , 1 / q ; q ] 2 k ( q ; q ) 4 k q k , = ( q 5 ; q 5 ) n ( q ; q ) 5 n ( q 5 ; q 5 ) n ( q ; q ) 5 n ( 1 + q 1 ) , ( b ) Λ n ( q 1 2 , q 1 5 n ) = k = 0 n 1 q 6 k + 1 2 [ q 5 n , q 5 2 + 5 n ; q 5 ] k [ q 1 5 n , q 3 2 + 5 n ; q ] k [ q 3 2 , q 1 2 ; q ] 2 k ( q ; q ) 4 k + 1 q k = ( q 5 ; q 5 ) n ( q 1 2 ; q ) 5 n + 1 ( q 5 2 ; q 5 ) n ( q ; q ) 5 n + 1 .

2.3. Case a / c = q

When m = 1 + n 2 and c = a / q in Lemma 1, the first term and the second term on the right will be annihilated. We obtain, in this case, the following simplified reciprocal relation.
Proposition 2
(Transformation formula). Between the quintic Λ n -series (3) and the cubic Ω m -series (5), there holds
Λ n ( a , q ) = Ω n 2 ( q 2 + 3 n a 2 , q 1 + 2 n a ) × ( q a ; q ) 2 n + 1 ( a / q ; q ) 2 n + 2 q a 2 , a 3 / q ; q 5 n + 1 ( q ; q ) n ( a / q ; q ) n + 3 ( q a 2 ; q ) 4 n + 2 .
Letting n in Proposition 2 and then invoking the Weierstrass M-test on uniformly convergent series (cf. Stromberg ([32], §3.106)), we find the following beautiful infinite series identity.
Theorem 3
(Nonterminating series identity).
Λ ( a , q ) = k = 0 1 q 6 k 1 a 3 [ q a 2 , a 3 / q ; q 5 ] k [ q , a / q ; q ] k [ q a , a / q ; q ] 2 k ( q a 2 ; q ) 4 k q k = q a q , q a 2 | q q a 2 , a 3 / q ; q 5 n = 0 q 5 n 2 + 3 n ( a ) n .
When a = q 2 , the last sum can be evaluated in closed form (cf. Bailey ([33], §8.6))
n = 0 q 5 n + 1 2 = ( q 10 ; q 10 ) ( q 5 ; q 10 ) .
This leads us to a very strange infinite series identity.
Corollary 1
(Nonterminating series identity).
Λ ( q 2 , q ) = k = 0 1 + q 6 k + 5 ( q 10 ; q 10 ) k ( q 2 ; q 2 ) k ( q ; q ) 2 k ( q ; q ) 2 k + 2 ( q ; q ) 4 k + 4 q k = ( q ; q ) ( q 10 ; q 10 ) 2 ( q ; q ) 2 ( q 5 ; q 10 ) .
Letting a = q 1 5 n 3 in Proposition 2, we deduce immediately the limiting result.
Theorem 4
(Limiting relation).
Λ n ( a , q ) 0 a s a q 1 5 n 3 .
When a ± q 1 5 n 2 in Proposition 2, the factorial quotient on the right contains a zero factor, which cancels all the terms in Ω n 2 ( q 2 + 3 n a 2 , q 1 + 2 n a ) , except for the ultimate one. Consequently, we have the identities.
Theorem 5
(Summation formulae for terminating series).
( a ) If n 2 0 , then Λ n ( ± q 1 5 n 2 , q ) = k = 0 n 1 q 6 k 5 + 15 n 2 [ q 5 n , ± q 5 + 15 n 2 ; q 5 ] k [ q , ± q 3 + 5 n 2 ; q ] k [ ± q 1 5 n 2 , ± q 3 + 5 n 2 ; q ] 2 k ( q 5 n ; q ) 4 k q k = ( 1 ) n + 2 2 ( q 5 ; q 5 ) n ( q ; q ) 5 n ( ± q 1 / 2 ; q ) 5 n 2 ( ± q 5 / 2 ; q 5 ) 2 + 3 n 2 q ( 5 n 2 60 n 20 ) / 8 . ( b ) If n 2 1 , then Λ n ( ± q 1 5 n 2 , q ) k = 0 n 1 q 6 k 5 + 15 n 2 [ q 5 n , ± q 5 + 15 n 2 ; q 5 ] k [ q , ± q 3 + 5 n 2 ; q ] k [ ± q 1 5 n 2 , ± q 3 + 5 n 2 ; q ] 2 k ( q 5 n ; q ) 4 k q k = ( 1 1 ) ( q 5 ; q 5 ) n ( q ; q ) 5 n ( q ; q ) 5 n 1 2 ( q 5 ; q 5 ) 1 + 3 n 2 q ( 5 n 2 60 n 17 ) / 8 .
The two remaining terminating series Λ n ( q 1 2 n , q ) and Λ n ( q 2 2 n , q ) are much harder to evaluate, whose closed expressions are given as follows.
Theorem 6
(Summation formulae for terminating series).
( a ) Λ n ( q 1 2 n , q ) = k = 0 n 1 q 6 k 6 n 4 [ q 1 4 n , q 4 6 n ; q 5 ] k [ q , q 2 2 n ; q ] k [ q 2 n , q 2 2 n ; q ] 2 k ( q 1 4 n ; q ) 4 k q k = χ ( n 5 0 , 2 ) ( q 1 4 n ; q 5 ) 2 n + 2 ( q 2 n + 1 ; q ) 2 n + 1 q 2 n 2 3 n 3 . ( b ) Λ n ( q 2 2 n , q ) = k = 0 n 1 q 6 k 6 n 7 [ q 3 4 n , q 7 6 n ; q 5 ] k [ q , q 3 2 n ; q ] k [ q 1 2 n , q 3 2 n ; q ] 2 k ( q 3 4 n ; q ) 4 k q k = χ ( n 5 0 , 1 ) ( q 3 4 n ; q 5 ) 2 n + 3 ( q 2 n + 2 ; q ) 2 n + 2 q 2 n 2 n 4 .
Proof. 
The difficulty to prove the theorem lies in the fact that there are zero factors appearing in both numerators and denominators of the cubic series Ω n 2 in Proposition 2. In order to overcome this obstacle, we first reformulate, under the replacement a = q 2 2 n x , the equality in Proposition 2 as
Λ n ( q 2 2 n x , q ) = P n ( x ) × Q n ( x )
where P n ( x ) and Q n ( x ) = Ω n 2 ( q 2 n x 2 , x ) are given explicitly by
P n ( x ) = ( q 1 2 n x ; q ) 2 n + 1 ( q 3 2 n x ; q ) 2 n + 2 q 3 4 n x 2 , q 7 6 n x 3 ; q 5 n + 1 ( q ; q ) n ( q 3 2 n x ; q ) n + 3 ( q 3 4 n x 2 ; q ) 4 n + 2 , Q n ( x ) = k = 0 n 2 1 q 4 k 2 n x 2 ( q n ; q ) 2 k ( x 2 / q ; q ) 2 k x , x / q , x / q ; q k ( q 2 n x 3 ; q 5 ) k ( q / x ; q 1 ) k ( q n x ; q ) 3 k q k .
Then it is not hard to evaluate the limits:
lim x q P n ( x ) = ( 1 ) n q 1 n , q 4 + n ; q 5 n + 1 2 ( 1 q n ) ( q 2 n + 1 ; q ) 2 n + 1 q n 2 9 n 8 2 , lim x 1 P n ( x ) = ( 1 ) n q 3 n , q 7 + n ; q 5 n + 1 ( q 2 n + 2 ; q ) 2 n + 2 q n 2 9 n 10 2 .
When x q , the only nonzero terms in Q n ( x ) are those with their summation index k = 0 and n + 2 3 k n 2 . Writing the initial term separately and then shifting forward the summation index by k k + n + 2 3 , we have the expression
lim x q Q n ( x ) = ( 1 q n ) + Ω n 2 n + 2 3 ( q 4 n + 2 3 n , q n + 2 3 ) × q n + 2 3 lim x q ( q n ; q ) 2 n + 2 3 ( x 2 / q ; q ) 2 n + 2 3 x , x / q , x / q ; q n + 2 3 ( q 2 n x 3 ; q 5 ) n + 2 3 ( q / x ; q 1 ) n + 2 3 ( q n x ; q ) 3 n + 2 3 = ( 1 q n ) + ρ ( n ) × Ω n 2 n + 2 3 ( q 4 n + 2 3 n , q n + 2 3 ) ,
where ρ ( n ) is given by
ρ ( n ) = ( 1 q n ) ( q 1 n ; q 5 ) n + 2 3 ( 1 q ) ( q n + 5 3 ; q ) n + 2 3 × 1 q 1 q n 3 q n 2 6 n 9 , n 3 0 ; 1 q n 1 3 1 q 2 q n 2 8 n + 7 9 , n 3 1 ; q n 2 7 n + 1 9 , n 3 2 .
Analogously for x 1 , the nonzero terms in Q n ( x ) are those with k = 0 , 1 and n + 3 3 k n 2 . Putting the first two terms together and then shifting forward the summation index by k k + n + 3 3 , we have another expression
lim x 1 Q n ( x ) = 1 q 2 n 2 + Ω n 2 n + 3 3 ( q 2 + 4 n 3 n , q n 3 ) × q n + 3 3 lim x 1 ( q n ; q ) 2 n + 3 3 ( x 2 / q ; q ) 2 n + 3 3 x , x / q , x / q ; q n + 3 3 ( q 2 n x 3 ; q 5 ) n + 3 3 ( q / x ; q 1 ) n + 3 3 ( q n x ; q ) 3 n + 3 3 = 1 q 2 n 2 + ϱ ( n ) × Ω n 2 n + 3 3 ( q 2 + 4 n 3 n , q n 3 ) ,
where ϱ ( n ) is given by
ϱ ( n ) = ( q 2 n ; q 5 ) n + 3 3 2 ( q n + 3 3 ; q ) n 3 × ( 1 q n 3 3 ) q n 2 3 n + 9 9 ( 1 q ) ( 1 q 2 ) , n 3 0 ; q n 2 2 n + 1 9 1 q , n 3 1 ; q n 2 n 2 9 1 q n 2 3 , n 3 2 .
Consequently, we have established the following reduction formulae:
Λ n ( q 1 2 n , q ) = lim x q P n ( x ) × Q n ( x ) = ( 1 ) n q 1 n , q 4 + n ; q 5 n + 1 2 ( 1 q n ) ( q 2 n + 1 ; q ) 2 n + 1 q n 2 9 n 8 2 , × ( 1 q n ) + ρ ( n ) Ω n 2 n + 2 3 q 4 n + 2 3 n , q n + 2 3 ,
Λ n ( q 2 2 n , q ) = lim x 1 P n ( x ) × Q n ( x ) = ( 1 ) n q 3 n , q 7 + n ; q 5 n + 1 ( q 2 n + 2 ; q ) 2 n + 2 q n 2 9 n 10 2 × 1 q 2 n 2 + ϱ ( n ) Ω n 2 n + 3 3 q 2 + 4 n 3 n , q n 3 .
Classifying n according to its residues moduolo 6, we can tabulate the afore-displayed two cubic series with their explicit parameters:
n
Ω n 2 n + 2 3 q 4 n + 2 3 n , q n + 2 3
Ω n 2 n + 3 3 q 2 + 4 n 3 n , q n 3
6 m
Ω m ( q 2 m , q 2 m )
Ω m 1 ( q 2 + 2 m , q 2 m )
6 m + 1
Ω m 1 ( q 3 + 2 m , q 1 + 2 m )
Ω m 1 ( q 1 + 2 m , q 2 m )
6 m + 2
Ω m ( q 2 + 2 m , q 1 + 2 m )
Ω m ( q 2 m , q 2 m )
6 m + 3
Ω m ( q 1 + 2 m , q 1 + 2 m )
Ω m 1 ( q 3 + 2 m , q 1 + 2 m )
6 m + 4
Ω m ( q 4 + 2 m , q 2 + 2 m )
Ω m ( q 2 + 2 m , q 1 + 2 m )
6 m + 5
Ω m ( q 3 + 2 m , q 2 + 2 m )
Ω m ( q 1 + 2 m , q 1 + 2 m )
All these cubic series can be evaluated in closed forms by the three known formulae due to the author ([28], Theorem 10):
Ω ( c , c ) = ( c ; q ) q 5 , q 3 c , q 2 / c , q c 2 , q 4 c 3 ; q 5 q , q c 2 ; q = k = 0 1 q 4 k c c , c , q c ; q k ( q ; q ) 3 k ( 1 / c ; q ) 2 k ( q c 2 ; q ) 2 k ( q c 2 ; q 5 ) k ( 1 / c ; q 1 ) k q k , Ω ( q c , c ) = ( q c ; q ) q 5 , q c , q 4 / c , q 2 c 2 , q 3 c 3 ; q 5 q , q c 2 ; q = k = 0 1 q 1 + 4 k c c , c , q c ; q k ( q ; q ) 3 k + 1 ( q / c ; q ) 2 k ( q c 2 ; q ) 2 k ( q 2 c 2 ; q 5 ) k ( 1 / c ; q 1 ) k q k , Ω ( q 2 c , c ) = ( q c ; q ) q 5 , q 4 c , q 6 / c , q 3 c 2 , q 2 c 3 ; q 5 q , q c 2 ; q = k = 0 1 q 2 + 4 k c c , c , q c ; q k ( q ; q ) 3 k + 2 ( q 2 / c ; q ) 2 k ( q c 2 ; q ) 2 k ( q 3 c 2 ; q 5 ) k ( 1 / c ; q 1 ) k q k .
After long and tedious computations, we obtain the simplified expressions
ρ ( n ) × Ω n 2 n + 2 3 ( q 4 n + 2 3 n , q n + 2 3 ) = ( 1 q n ) × + 1 , n 5 0 , 2 ; 0 , n 5 1 ; 1 , n 5 3 , 4 . ϱ ( n ) × Ω n 2 n + 3 3 ( q 2 + 4 n 3 n , q n 3 = 1 q 2 n 2 ) × + 1 , n 5 0 , 1 ; 0 , n 5 3 ; 1 , n 5 2 , 4 .
Substituting them into (9) and (10) and then making some routine simplifications, we confirm the two summation formulae in Theorem 6. □

3. Closed Formulae for the λ n -Series

For the terminating series identities for Λ n established in the last section, their counterparts for the series λ m can analogously be deduced in view of the reciprocal relations (3) and (4). We illustrate the following example to show how this approach works. For m N 0 with m 3 2 , we have, according to (4), the transformation
λ m q 1 m 3 , q 1 m = ( q 6 m ) × lim x 1 Λ m q 1 5 m 3 x , q × ( q 2 2 m 3 ; q ) 4 m q m , q 5 + 2 m 3 ; q m q 1 m 3 , q 5 m 3 ; q 2 m q 5 , q 10 5 m 3 ; q 5 m .
Because of m 3 2 , the above factorial quotient does not have zero factors in its denominator. In view of Theorem 4, we establish the following summation formula.
Corollary 2
(Reversal series from Theorem 4: m N 0 with m 3 2 ).
λ m q 1 m 3 , q 1 m = k = 0 m 1 q 6 k m [ q m , q 5 + 2 m 3 ; q ] k [ q 5 , q 10 5 m 3 ; q 5 ] k ( q 2 + 2 m 3 ; q ) 4 k [ q 5 m 3 , q 1 + m 3 ; q ] 2 k q k = 0 .
By examining the reversal series for those given in Theorems 1, 2, 5 and 6, we can further deduce the following summation formulae.
Corollary 3
(Reversal series from Theorem 1).
( a ) λ m q 1 2 m , q 2 + 4 m = k = 0 m 1 + q 6 k 6 m 2 [ q 3 + 4 m , q 2 6 m ; q ] k [ q 5 , q 10 m ; q 5 ] k ( q 2 4 m ; q ) 4 k [ q 1 2 m , q 1 2 m ; q ] 2 k q k = q 2 7 m 2 m 2 ( q ; q ) 6 m + 2 ( q ; q ) 2 m + 1 ( 1 ; q ) 2 m ( q 5 ; q 5 ) 2 m . ( b ) λ m q 3 2 2 m , q 3 + 4 m = k = 0 m 1 q 6 k 6 m 7 2 [ q 4 + 4 m , q 7 2 6 m ; q ] k [ q 5 , q 5 2 10 m ; q 5 ] k ( q 3 4 m ; q ) 4 k [ q 1 2 2 m , q 3 2 2 m ; q ] 2 k q k = q 4 8 m 2 m 2 ( q 1 2 ; q ) 6 m + 4 ( q 3 2 ; q ) 2 m + 1 ( q 1 2 ; q ) 2 m + 1 ( q 5 2 ; q 5 ) 2 m + 1 .
Corollary 4
(Reversal series from Theorem 2).
( a ) λ m q 2 m , q 4 m = k = 0 m 1 + q 6 k 6 m + 1 [ q 1 + 4 m , q 1 6 m ; q ] k [ q 5 , q 5 10 m ; q 5 ] k ( q 4 m ; q ) 4 k [ q 2 2 m , q 2 m ; q ] 2 k q k = q 1 5 m 2 m 2 ( q ; q ) 6 m 1 ( q ; q ) 2 m ( 1 ; q ) 2 m 1 ( q 5 ; q 5 ) 2 m 1 . ( b ) λ m q 1 2 2 m , q 1 + 4 m = k = 0 m 1 q 6 k 6 m 1 2 [ q 2 + 4 m , q 1 2 6 m ; q ] k [ q 5 , q 5 2 10 m ; q 5 ] k ( q 1 4 m ; q ) 4 k [ q 3 2 2 m , q 1 2 2 m ; q ] 2 k q k = q 6 m 2 m 2 ( q 1 2 ; q ) 6 m + 1 ( q 1 2 ; q ) 2 m + 1 ( q 1 2 ; q ) 2 m 1 ( q 5 2 ; q 5 ) 2 m .
Corollary 5
(Reversal series from Theorem 5).
( a ) I f m 2 0 , t h e n λ m ± q 1 + m 2 , q 1 m = k = 0 m 1 q 6 k + 5 + 3 m 2 [ q m , ± q 5 + 3 m 2 ; q ] k [ q 5 , ± q 15 + 5 m 2 ; q 5 ] k ( q 1 + m ; q ) 4 k [ ± q 1 + m 2 , ± q 5 + m 2 ; q ] 2 k q k = ( 1 ) m 2 ( ± q 1 2 ; q ) m 2 ( ± q 5 2 ; q 5 ) m + 2 2 ( ± q 5 + m 2 ; q ) m q m 2 + 4 m 8 . ( b ) I f m 2 1 , t h e n λ m ± q 1 + m 2 , q 1 m = k = 0 m 1 q 6 k + 5 + 3 m 2 [ q m , ± q 5 + 3 m 2 ; q ] k [ q 5 , ± q 15 + 5 m 2 ; q 5 ] k ( q 1 + m ; q ) 4 k [ ± q 1 + m 2 , ± q 5 + m 2 ; q ] 2 k q k = ( 1 1 ) ( q ; q ) m 1 2 ( q 5 ; q 5 ) m + 1 2 ( q 5 + m 2 ; q ) m q m 2 + 4 m + 3 8 .
Corollary 6
(Reversal series from Theorem 6).
( a ) I f m 5 1 , t h e n λ m q , q 1 m = k = 0 m 1 q 6 k + 4 [ q m , q 3 + m ; q ] k [ q 6 m , q 9 + m ; q 5 ] k ( q 2 ; q ) 4 k [ q , q 3 ; q ] 2 k q k = q m ( 1 q 2 ) ( 1 q 1 m ) ( 1 q 4 + m ) ( 1 q 1 + m ) ( 1 q 2 + m ) χ ( m 5 0 , 2 ) . ( b ) I f m 5 3 , t h e n λ m q 2 , q 1 m = k = 0 m 1 q 6 k + 7 [ q m , q 4 + m ; q ] k [ q 8 m , q 12 + m ; q 5 ] k ( q 4 ; q ) 4 k [ q 2 , q 4 ; q ] 2 k q k = q m ( 1 q ) ( 1 q 3 m ) ( 1 q 7 + m ) ( 1 q 1 + m ) ( 1 q 3 + m ) χ ( m 5 0 , 1 ) .

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Acknowledgments

The author expresses his sincere gratitude to the three anonymous referees for their careful reading and valuable comments that contribute significantly to improving the manuscript during the revision.

Conflicts of Interest

The author declares no conflict of interest.

References

  1. Chu, W. Basic hypergeometric identities: An introductory revisiting through the Carlitz inversions. Forum Math. 1995, 7, 117–129. [Google Scholar]
  2. Gasper, G.; Rahman, M. Basic Hypergeometric Series, 2nd ed.; Cambridge University Press: Cambridge, UK, 2004. [Google Scholar]
  3. Gessel, I.; Stanton, D. Applications of q-Lagrange inversion to basic hypergeometric series. Trans. Am. Math. Soc. 1983, 277, 173–201. [Google Scholar]
  4. Bressoud, D.M. Some identities for terminating q-series. Math. Proc. Cambridge Philos. Soc. 1981, 89, 211–223. [Google Scholar] [CrossRef]
  5. Bressoud, D.M. An easy proof of the Rogers–Ramanujan identities. J. Combin. Theory Ser. A 1983, 16, 235–241. [Google Scholar] [CrossRef] [Green Version]
  6. Chan, H.H. Triple product identity, quintuple product identity and Ramanujan’s differential equations for the classical Eisenstein series. Proc. Am. Math. Soc. 2007, 135, 1987–1992. [Google Scholar] [CrossRef] [Green Version]
  7. Chan, H.H.; Liu, Z.G.; Ng, S.T. Elliptic functions and the quintuple, Hirschhorn and Winquist product identities. Int. J. Number. Theory 2005, 1, 33–43. [Google Scholar] [CrossRef] [Green Version]
  8. Gessel, I.; Xin, G.C. A short proof of the Zeilberger–Bressoud q-Dyson theorem. Proc. Am. Math. Soc. 2006, 134, 2179–2187. [Google Scholar] [CrossRef]
  9. Stanton, D. q-Analogues of Euler’s Odd = Distinct theorem. Ramanujan J. 2009, 19, 107–113. [Google Scholar] [CrossRef]
  10. Andrews, G.E. The Theory of Partitions; Cambridge University Press: Cambridge, UK, 1976. [Google Scholar]
  11. Hardy, G.H.; Wright, E.M. An Introduction to the Theory of Numbers, 5th ed.; Oxford Science Publications: Oxford, UK, 1980. [Google Scholar]
  12. Andrews, G.E. q-series: Their Development and Application in Analysis, Number Theory, Combinatorics, Physics and Computer Algebra; CBMS Regional Conference Series in Mathematics 66; American Mathematical Society: Providence, RI, USA, 1986. [Google Scholar]
  13. Baxter, R.J. Exactly Solved Models in Statistical Mechanics; Academic Press: New York, NY, USA, 1982. [Google Scholar]
  14. Berndt, B.C.; Ono, K. q-series with Applications to Combinatorics, Number Theory, and Physics; American Mathematical Society: Providence, RI, USA, 2001. [Google Scholar]
  15. Chu, W. Abel’s lemma on summation by parts and Ramanujan’s 1ψ1-series identity. Aequationes Math. 2006, 72, 172–176. [Google Scholar] [CrossRef]
  16. Chu, W. Bailey’s very well–poised 6ψ6-series identity. J. Combin. Theory Ser. A 2006, 113, 966–979. [Google Scholar] [CrossRef] [Green Version]
  17. Chu, W. Abel’s method on summation by parts and bilateral well–poised 5ψ5-series identities. Port. Math. 2009, 66, 275–302. [Google Scholar] [CrossRef]
  18. Chu, W.; Jia, C. Abel’s method on summation by parts and terminating well–poised q-series identities. J. Comput. Appl. Math. 2007, 207, 360–370. [Google Scholar] [CrossRef] [Green Version]
  19. Gasper, G. Summation, transformation, and expansion formulas for bibasic series. Trans. Am. Math. Soc. 1989, 312, 257–277. [Google Scholar] [CrossRef]
  20. Gasper, G.; Rahman, M. An indefinite bibasic summation formula and some quadratic, cubic and quartic summation and transformation formulas. Canad. J. Math. 1990, 42, 1–27. [Google Scholar] [CrossRef]
  21. Rahman, M. Some quadratic and cubic summation formulas for basic hypergeometric series. Canad. J. Math. 1993, 45, 394–411. [Google Scholar] [CrossRef]
  22. Rahman, M.; Verma, A. Quadratic transformation formulas for basic hypergeometric series. Trans. Am. Math. Soc. 1993, 335, 277–302. [Google Scholar] [CrossRef]
  23. Rahman, M. Some cubic summation formulas for basic hypergeometric series. Utilitas Math. 1989, 36, 161–172. [Google Scholar] [CrossRef]
  24. Chu, W.; Wang, C. Abel’s Lemma on summation by parts and partial q-series transformations. Sci. China Math. (Engl.) 2009, 52, 720–748, Sci. China Math. (Chin.)2009, 39, 405–432. [Google Scholar] [CrossRef]
  25. Chu, W.; Jia, C. Quartic theta hypergeometric series. Ramanujan J. 2013, 32, 23–81. [Google Scholar] [CrossRef]
  26. Chu, W.; Wang, C. Partial sums of two quartic q-series SIGMA: Symmetry Integrability. Geom. Methods Appl. 2009, 5, 19. [Google Scholar]
  27. Chu, W. Inversion techniques and combinatorial identities: Jackson’s q-analogue of the Dougall–Dixon theorem and the dual formulae. Compos. Math. 1995, 95, 43–68. [Google Scholar]
  28. Chu, W. Summation formulae for twisted cubic q-series. Math. Methods Appl. Sci. 2019, 42, 1831–1843. [Google Scholar] [CrossRef]
  29. Chu, W. Abel’s lemma on summation by parts and basic hypergeometric series. Adv. Appl. Math. 2007, 39, 490–514. [Google Scholar] [CrossRef] [Green Version]
  30. Chu, W.; Wang, X. Abel’s lemma on summation by parts and terminating q-series identities. Numer. Alg. 2008, 49, 105–128. [Google Scholar] [CrossRef]
  31. Chu, W.; Zhang, W. Abel’s method on summation by parts and nonterminating q-series identities. Collect. Math. 2009, 60, 193–211. [Google Scholar] [CrossRef]
  32. Stromberg, K.R. An Introduction to Classical Real Analysis; Wadsworth, INC.: Belmont, CA, USA, 1981. [Google Scholar]
  33. Bailey, W.N. Generalized Hypergeometric Series; Cambridge University Press: Cambridge, UK, 1935. [Google Scholar]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Chu, W. Summation Formulae for Quintic q-Series. Mathematics 2022, 10, 2210. https://doi.org/10.3390/math10132210

AMA Style

Chu W. Summation Formulae for Quintic q-Series. Mathematics. 2022; 10(13):2210. https://doi.org/10.3390/math10132210

Chicago/Turabian Style

Chu, Wenchang. 2022. "Summation Formulae for Quintic q-Series" Mathematics 10, no. 13: 2210. https://doi.org/10.3390/math10132210

APA Style

Chu, W. (2022). Summation Formulae for Quintic q-Series. Mathematics, 10(13), 2210. https://doi.org/10.3390/math10132210

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