The Hecke slash sum vanishes, and is bounded, at the cusps #
heckeSlashSum is a finite sum of slashes, so its behaviour at a cusp follows from that of its
summands. A slash is zero at c exactly when the original function is zero at g • c
(OnePoint.IsZeroAt.smul_iff), and g • c is again a cusp because the representatives are
rational (IsCusp.smul_map_ratCast). So a function vanishing at every cusp has a slash sum
vanishing at every cusp, whatever representatives were chosen.
This is the step the Layer 2 statement "Tₙ preserves S_k" rests on.
Both are stated for an arbitrary arithmetic Γ, which is the hypothesis a CuspForm Γ k
supplies through zero_at_cusps'. IsCusp.smul_map_ratCast reduces to 𝒮ℒ internally via
Subgroup.IsArithmetic.isCusp_iff_isCusp_SL2Z, so no comparison is left for the caller.
Main results #
HeckeRing.GL2.isZeroAt_heckeSlashSum: the slash sum of a function vanishing at every cusp vanishes at every cusp.HeckeRing.GL2.isBoundedAt_heckeSlashSum: the same for boundedness.
Provenance #
The shape is AINTLIB's heckeT_p_ut_zero_at_cusps
(LeanModularForms/HeckeRIngs/GL2/AdjointTheory.lean
lines 62-70, commit 2baa76f742bdb4fb8ee323fabba41203bd390e08, Apache-2.0, Chris Birkbeck),
which runs that argument by hand with Finset.sum_induction over its own representatives, once
per call site. Here it is factored: the general statement for an arbitrary finite family of
rational matrices lives in ModularForms/Cusps/Rat/Slash.lean, and this module only specialises it
to heckeSlashSum.
The slash sum vanishes at every cusp when the function does.
The slash sum is bounded at every cusp when the function is.