On series identities of Gosper and integrals of Ramanujan theta function $\psi(q)$
Mohamed El Bachraoui

TL;DR
This paper proves new Lambert series identities originally stated by Gosper, and applies these results to evaluate integrals involving Ramanujan's theta function, as well as related squared expressions inspired by Ramanujan's identities.
Contribution
It introduces new Lambert series identities related to Ramanujan theta functions and evaluates complex integrals and squared expressions based on these identities.
Findings
Proved Lambert series identities originally stated by Gosper.
Evaluated integrals involving Ramanujan theta function $\psi(q)$.
Expressed squares of certain Ramanujan theta function combinations in terms of Lambert series.
Abstract
We prove some Lambert series which were stated by Gosper without proof or reference. As an application, we shall evaluate integrals involving Ramanujan theta function . Furthermore, motivated by Ramanujan's identities for and , we shall evaluate the squares of and in terms of Lambert series.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Mathematical Identities · Analytic Number Theory Research · History and Theory of Mathematics
On series identities of Gosper and integrals of Ramanujan theta function
Mohamed El Bachraoui
Department of Mathematical Sciences, United Arab Emirates University, PO Box 15551, Al-Ain, UAE
Abstract.
We prove some Lambert series which were stated by Gosper without proof or reference. As an application we shall evaluate integrals involving Ramanujan theta function . Furthermore, motivated by Ramanujan’s identities for and , we shall evaluate the squares of and in terms of Lambert series.
Key words and phrases:
Ramanujan theta functions; Lambert series; integrals; -trigonometric functions.
1991 Mathematics Subject Classification:
33E05, 11F11, 11F12
1. Introduction
Throughout the paper let with guaranteeing that . As usual, the -shifted factorials for a complex number are given by
[TABLE]
Gosper [10] introduced the following -analogues of and
[TABLE]
along with the following -constant
[TABLE]
for which one has . The Ramanujan theta functions , , and are given by
[TABLE]
but our focus in this work is mainly on . We refer the reader to Berndt [3] and Cooper [6] for a survey on Ramanujan theta functions, the connections between them, and their applications. Our primary goal in this paper is to confirm the following formulas which were stated without proof or reference by Gosper [10, p. 102] and to give some of their applications.
Theorem 1**.**
We have
[TABLE]
To prove some of our identities we shall also need the following result which apparently is new.
Theorem 2**.**
We have
[TABLE]
Remark 1**.**
Recall for the divisor function the basic fact that
[TABLE]
Now letting
[TABLE]
we clearly have and . More importantly, we can verify that
[TABLE]
Therefore, with the exception of part (f), each one of the series in Theorems 1 and 2 can be written in terms of .
Remark 2**.**
Note that Theorem 1(e) is essentially equivalent to the Ramanujan’s identity
[TABLE]
see Berndt [3, p. 223]. We include it here as our proof is different.
Our primary application is motivated by Ramanujan’s integral identities of the theta functions which are found in [15, p. 207]. Ramanujan recorded six integrals involving the functions , , and . Son [17] gave complete proofs for five of these integrals and a partial proof for the sixth integral stating that
[TABLE]
Berndt and Zaharescu [5] found a complete proof for (4) and established Ramanujan’s constant . Extensions for the integral identity (4) along with a method to compute related constants are obtained in Ahlgren et al. [2]. Besides, among the six integral formulas of Ramanujan we find
[TABLE]
Observe that if we consider indefinite integrals rather than definite integrals in the two identities of (5), then we get Lambert series involved as values. Indeed, as to the first formula in (5), on taking logarithms and expanding as powers series, we get
[TABLE]
Likewise the second formula in (5) is equivalent to
[TABLE]
It is also worth to notice that can be obtained by using Ramanujan’s famous formula for (see Berndt [4, p. 138]) as follows
[TABLE]
We have in this direction the following results including a different formula for .
Theorem 3**.**
We have
[TABLE]
Our secondary application is related to the following Lambert series representations for and which are due to Ramanujan’s (see Berndt [3, p. 223, 226]),
[TABLE]
For other related identities involving Lambert series using we refer the reader to Shen [16] and Liu [13]. It seems that not much is known about series representations of higher powers of and . We have the following contribution concerning the squares of and .
Theorem 4**.**
We have
[TABLE]
[TABLE]
[TABLE]
The paper is organised as follows. As the main tool in this note is the -trigonometry of Gosper [10], the needed ingredients about this will be presented in Section 2. In Section 3 we will prove Theorem 1, in Section 4 we shall prove Theorem 2, Section 5 is devoted to the proof of Theorem 4, and in Section 6 we will prove Theorem 3.
2. Facts on Gosper’s -trigonometry
In this section we collect the properties of Gosper’s functions and , as defined in (1), which we need to derive our results. Just as the classical functions and , it is easy to see that \cos_{q}(z)=\sin_{q}\Big{(}\frac{\pi}{2}-z\Big{)}, the function is odd and the function is even and therefore we have for any nonnegative integer
[TABLE]
where derivatives are with respect to . Gosper gave several -trigonometric identities based on a computer algebra facility called MACSYMA. For our current purposes, we shall need the following formulas which are marked with the same labels as in Gosper [10].
[TABLE]
[TABLE]
[TABLE]
[TABLE]
[TABLE]
[TABLE]
and
[TABLE]
Gosper by combining (2) and (2) obtained
[TABLE]
In addition to the previous formulas which are in one variable , Gosper [10, p. 101] also stated many multi-variable identities. For instance he conjectured that
[TABLE]
A proof for (2) can be found in [7, 14], a proof for (2) is given in [7], proofs for (2), (2), and (2) are obtained in [8], and a proof for (2) is found in [9] . In addition, it was observed in Gosper [10, p. 89-93] that (2) is equivalent to (2) and that (2) can be deduced by a combination of (2), (2), and (2). Furthermore, we shall need some explicit derivates with respect to the variable of the functions and . To evaluate these derivatives, we make an appeal to series expansions for and in powers of which were stated without proof by Gosper [10, p. 98-99]. We first confirm these series expansions as in the following theorem.
Theorem 5**.**
If is a complex number such that , then
[TABLE]
Proof.
Clearly,
[TABLE]
Writing
[TABLE]
we easily see that
[TABLE]
But
[TABLE]
This proves part (a). Part (b) follows similarly. ∎
From these expansions one can straightforwardly deduce the following explicit derivatives with respect to .
Corollary 1**.**
We have
[TABLE]
We note that the derivatives and are also given in Gosper [10, p. 101].
3. Proof of Theorem 1
(a) Observe first the elementary fact that
[TABLE]
where is defined as in Remark 1. It follows that
[TABLE]
On the other hand, by an identity of Ramanujan, see Berndt [3, p. 139],
[TABLE]
Now combine the above with the generating function (3) to arrive at
[TABLE]
Thus with the help of (9) the following three formulas are equivalent:
[TABLE]
[TABLE]
[TABLE]
[TABLE]
The foregoing identity after simplification and rearrangement means
[TABLE]
or, equivalently
[TABLE]
But the previous identity holds true since by (9) we have
[TABLE]
This completes the proof of part (a). We postpone the proof of (b) as we need parts (d, e).
As to part (c), square both sides of (2), differentiate three times with respect to , and then let to obtain,
[TABLE]
which by appealing to Corollary 1, simplifying, and rearranging reduces to
[TABLE]
or equivalently,
[TABLE]
which is equivalent to the desired formula.
(d) Differentiating (2) three times with respect to yields
[TABLE]
which at boils down to
[TABLE]
Now by an appeal to the relations in Corollary 1 the foregoing identity implies
[TABLE]
[TABLE]
which after straightforward simplifications becomes
[TABLE]
Then dividing both sides by we arrive at the desired formula.
(e) Letting in the formula (2) we deduce
[TABLE]
Differentiating twice with respect to and letting yield
[TABLE]
which simplifies to
[TABLE]
Then by virtue of Corollary 1 we obtain
[TABLE]
or equivalently,
[TABLE]
as desired.
We now prove part (b). By an appeal to parts (d, e) and Theorem 2 with replaced by and elementary simplifications we get
[TABLE]
which proves part (b).
(f) Taking the fourth derivative at in (2) we find
[TABLE]
Then using the derivatives in Corollary 1 we deduce that
[TABLE]
or, equivalently
[TABLE]
which is clearly equivalent to the desired identity.
4. Proof of Theorem 2
Differentiating (2) three times with respect to , taking , and simplifying yield
[TABLE]
Now make an appeal to Corollary 1 and simplify to derive
[TABLE]
The foregoing identity reduces to
[TABLE]
which completes the proof.
5. Proof of Theorem 4
(a) On the one hand, by Theorem 1(d) and Theorem 2 we get
[TABLE]
On the other hand, by virtue of (8) with replaced by we obtain
[TABLE]
Now compare the above to complete the proof.
(b) By Theorem 1(f) with replaced by we derive
[TABLE]
Now combine the previous formula with (9) to conclude the proof.
6. Proof of Theorem 3
(a) From the fact and Theorem 1(f) we have
[TABLE]
Moreover, we have the following elementary identities
[TABLE]
and
[TABLE]
It follows from the above that
[TABLE]
as desired.
(b) By Theorem 1(c) and the definition of we have
[TABLE]
implying that
[TABLE]
Thus after simplification we deduce
[TABLE]
which proves the first identity of part (b). As to the second formula, by Theorem 1(a) we obtain
[TABLE]
which with the help of (6) implies that
[TABLE]
[TABLE]
Then
[TABLE]
as desired.
(c) From Theorem 1(c), we deduce that
[TABLE]
It follows that
[TABLE]
that is,
[TABLE]
Now use part (b) and simplify to arrive at the result.
(d) By an appeal to Theorem 2 we deduce that
[TABLE]
and therefore
[TABLE]
which yields the desired formula.
(e) By virtue of Theorem 1(d) and easy simplification we deduce
[TABLE]
Now proceed as in the previous parts and use part (d) to conclude the proof.
(f) Using Theorem 1(b), we get
[TABLE]
implying that
[TABLE]
Now combine part (d) with (7) and simplify to establish the desired integral.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] S. Abo Touk, Z. Al Houchan, and M. El Bachraoui, Proofs for two q 𝑞 q -trigonometric identities of Gosper , J. Math. Anal. Appl. 456 (2017), 662-670.
- 2[2] S. Ahlgren, B. C. Berndt, A. J. Yee, and A. Zaharescu, Integrals of Eisenstein series and derivatives of L-functions , Internat. Math. Res. Not. 32 (2002) 1723-1738.
- 3[3] B. C. Berndt, Ramanujan’s notebook. III , Springer-Verlag, 1991.
- 4[4] B. C. Berndt, Number theory in the spirit of Ramanujan , American Mathematical Society, Providence, Rhode Island, 2006.
- 5[5] B. C. Berndt and A. Zaharescu, An integral of Dedekind eta-functions in Ramanujan?s lost notebook , J. Reine Angew. Math. 551 (2002) 33-39.
- 6[6] S. Cooper, Ramanujan’s theta functions , Springer, 2017.
- 7[7] M. El Bachraoui, Confirming a q 𝑞 q -trigonometric conjecture of Gosper , Proc. Amer. Math. Soc. 146:4 (2018), 1619–1625.
- 8[8] M. El Bachraoui, Proving some identities of Gosper on q-trigonometric functions Proc. Amer. Math. Soc. In press, DOI: https://doi.org/10.1090/proc/14084.
