Hankel-type determinants and Drinfeld quasi-modular forms (Q1932377): Difference between revisions

Let \(A = \mathbb F_q[\theta]\) be the polynomial ring over a finite field with fraction field \(K\). The theory of (Drinfeld) modular forms over \(A\) is largely similar to the theory of classical elliptic modular forms over the base ring \(\mathbb Z\). \medskip The \(K\)-algebra of modular forms is generated by two distingished modular forms \(g\) (similar to classical \(g_2, g_3\)) and \(h\), a form with no immediate \(\mathbb Z\)-counterpart, but whose \(q-1\)-th power up to sign agrees with the Drinfeld discriminant function \(\Delta\). In order to have a differential algebra, one is forced to generalize Drinfeld ``modular form'' to Drinfeld ``quasi-modular form''. It is easy to see that the algebra of Drinfeld quasi-modular forms is generated over \(K\) by the three algebraically independent forms \(g,h\) and \(E\), where \(E\) is the logarithmic derivative of \(h\) (and is similar to the classical quasi-modular form \(E_2\)). Besides their weight and type, such forms have a depth (its degree in E\()\), and the \(K\)-algebra \(\widetilde{M} \)is bigraded through weight and type, and filtered through the depth function. Moreover, there are natural \(A\)-structures on spaces of such forms, given by the integrality of the \(u\)-series expansion coefficients (corresponding to classical \(q\)-expansions). \medskip As the authors discuss in their excellent introduction, it is a major problem to find a minimal set of generators of the \(A\)-algebra \(\widetilde{M}_A\) of \(A\)-integral quasi-modular forms. They achieve important partial results toward a solution of this problem. Viz, they define (by an insightful construction involving Hankel determinants) a double series \(E_{j,k}\) (\(j \in Z\), \(k\geq 1\)) of quasi-modular forms with remarkable integrality properties. For example, for \(j\geq 0\), the vanishing order of \(E_{j,k}\) w.r.t. the \(u\)-expansion is determined, it is \(A\)-integral and normalized, and if moreover \(k = 1\) or 2 then it is extremal, that is, has maximal vanishing order among all forms of the same weight, type and depth.