Lacunarity of Han-Nekrasov-Okounkov q-series

From MaRDI portal
Publication:2660055



Abstract: A power series is called lacunary if `almost all' of its coefficients are zero. Integer partitions have motivated the classification of lacunary specializations of Han's extension of the Nekrasov-Okounkov formula. More precisely, we consider the modular forms [F_{a,b,c}(z) := frac{eta(24az)^a eta(24acz)^{b-a}}{eta(24z)},] defined in terms of the Dedekind eta-function, for integers a,cgeq1 where bgeq1 is odd throughout. Serre determined the lacunarity of the series when a=c=1. Later, Clader, Kemper, and Wage extended this result by allowing a to be general, and completely classified the Fa,b,1(z) which are lacunary. Here, we consider all c and show that for ain1,2,3, there are infinite families of lacunary series. However, for ageq4, we show that there are finitely many triples (a,b,c) such that Fa,b,c(z) is lacunary. In particular, if ageq4, bgeq7, and cgeq2, then Fa,b,c(z) is not lacunary. Underlying this result is the proof the t-core partition conjecture proved by Granville and Ono.


Recall that a series \(\sum_{n = 0}^\infty a(n) q^n\) is \textit{lacunary} if \[ \lim_{x \to \infty} \frac{\#\{0 \leq n < x \ : \ a(n) = 0\}}{x} = 1. \] The authors begin their paper by surveying some results relating partitions of integers to certain generating functions. They begin with Euler's identity \(\prod_{n = 1}^\infty (1 - q^n) = \sum_{m = - \infty}^\infty (-1)^m q^{\frac{3m^2 + m}{2}}\), and culminate with an identity by Han which relates the infinite product \(\prod_{n = 1}^\infty \frac{(1 - q^{an})^a (1 - (y q^a)^n)^{b-a}}{1 - q^n}\) to a combinatorial sum over partitions. A small change of variables yields the following function, which is the main object of interest in this paper: \[ F_{a, b, c}(z) := \frac{\eta(24 a z)^a \eta(24 a c z)^{b - a}}{\eta(24 z)} = q^{abc + a^2 - a^2 c - 1} \prod_{n = 1}^\infty \frac{(1 - q^{24 an})^a (1 - q^{24 a c n})^{b-a}}{1 - q^{24n}}. \] Here, \(\eta(z) := q^{1/24} \prod_{n = 1}^\infty (1 - q^n)\) is the Dedekind eta function, and \(q := e^{2 \pi i z}\) as is typical in the study of modular forms. Throughout, the authors assume that \(a, b, c \geq 1\) are integers and \(b\) is odd. Under these hypotheses, \(F_{a, b, c}\) is a weakly holomorphic modular form of weight \(\frac{b - 1}{2}\). The authors prove that \(F_{3, a, c}\) is lacunary whenever \(a \in \{1, 2, 3\}\) and \(F_{5, a, c}\) is lacunary whenever \(a \in \{1, 2\}\). This is almost immediate from Deligne's and Serre's result that a weight-one holomorphic modular form on a congruence group is lacunary, though they also use Serre's result that positive-definine binary quadratic forms represent 0\% of the integers. On the other hand, the authors prove that if \(a \geq 4\), \(b \geq 1\) is odd, and \(c \geq 2\), then \(F_{a, b, c}\) is not lacunary except possibly if \((a, b, c) \in \{(4, 5, 3), (4, 5, 5), (4, 5, 11)\}\). This proof is more involved: the authors first reduce their task to verifying certain \(F_{a, b, c}\) are not linear combinations of cusp forms with complex multiplication, and then to verifying that these \(F_{a, b, c}\) are not annihilated by the application of particular Hecke operators, and then use the Polya-Vindograv inequality to preclude annihilation in this manner. The authors conjecture that \(F_{a, b, c}\) is lacunary even if \((a, b, c) \in \{(4, 5, 3), (4, 5, 5), (4, 5, 11)\}\), and suggest a computational method for verifying this conjecture, but say they lacked the computational power to perform this verification themselves. They prove The reader is advised that there is a typo in equation 1.5: the expression \((1 - q)^n\) in the denominator of the righthand side of the equation ought to be \(1 - q^n\). In addition, there is a line missing from the table at the top of page 626: when \(a = 3\), the function \(F_{a, b, 1}\) is lacunary exactly if \(b \in \{3, 5, 9\}\).











This page was built for publication: Lacunarity of Han-Nekrasov-Okounkov \(q\)-series

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2660055)