4
$\begingroup$

Suppose I have a holomorphic modular form $f \in M_k(\Gamma_0(N), \chi)$ with $k \in \mathbb Z^+$ and $f = \sum_{n=1}^\infty a(n)q^n$. Considering the $q$-expansion, one may suspicious that $f$ is actually a cusp form. Then, it remains to check that $f$ vanishes at the cusps of $\Gamma_0(N)$ (say $\Gamma_0(p)$, so we only have the two cusps).

It is clear from the $q$-expansion of $f$ that it vanishes at the cusp at $\infty$, but how exactly do you show that it vanishes at the cusp at $0$?

I have heard that it has something to do with finding $\gamma = \left( \begin{array}{c c} a & b \\ c & d \end{array} \right) \in GL_2(\mathbb Z)^+$ that takes $\infty$ to $0$ under the transformation $\gamma z = \frac{az+b}{cz+d}$. Then, somehow wrangle $f |_k \gamma = (cz+d)^{-k} f(\gamma z)$ back into a $q$-expansion at $\infty$. If this has no constant term, then $f$ vanishes at the cusp at $0$.

If this is indeed the procedure for verifying vanishing at the cusps, then it is unclear to me what sort of "wrangling" needs to occur. I think the Fricke involution $W_N = \left( \begin{array}{c c} 0 & -1 \\ N & 0 \end{array} \right)$ would work for $\gamma$, but the $q$-expansion of $f$ is an unintelligible mess after hitting it with the $|_k$ operator.

$\endgroup$
3
$\begingroup$

As far as I can tell, the method listed in the question has little practical value. Mangling the $q$-expansion with the $|_k$ operator makes it very difficult to get back to a $q_N$ expansion.

A more hopeful approach would be to attempt to create a basis-like set of modular forms for $S_k(\Gamma_0(N), \chi)$ and then to write $f$ as a linear combination of these forms.

For example, in Arithmetic of the 13-regular partition function modulo 3, Webb uses Dedekind's eta function $$ \eta(z) := q^{1/24} \prod_{n=1}^\infty (1-q^n) $$ to create a pseudo-basis for $M_{12}(\Gamma_0(312), \chi_{13})$ modulo 3. Many of the forms in his "basis" are actually cusp forms, and could potentially be used create the desired cusp form (modulo 3) in $M_{12}(\Gamma_0(312), \chi_{13})$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.