Rademacher's Gamma quotient estimate
OpenLL_Decomp530a.leaf_rademacher_gamma_ratioFor every real and every complex with , one has , where the right side is a real power of a nonnegative real number.
import Mathlib open Classical Complex Set Filter ZMod open scoped Topology
namespace LL_Decomp530a
theorem leaf_rademacher_gamma_ratio :
∀ Q : ℝ, 0 ≤ Q → ∀ s : ℂ, -(1 / 2 : ℝ) ≤ s.re → s.re ≤ 1 / 2 →
‖Complex.Gamma ((Q : ℂ) / 2 + (1 - s) / 2) /
Complex.Gamma ((Q : ℂ) / 2 + s / 2)‖ ≤
(‖(Q : ℂ) + 1 + s‖ / 2) ^ (1 / 2 - s.re) := by sorry
end LL_Decomp530aRead-back
What the Lean code literally says, in plain math · claude-haiku
The declaration is named LL_Decomp530a.leaf_rademacher_gamma_ratio and is a single theorem with no hypotheses beyond those in its statement. It quantifies over a real number and a complex number (with ). The hypotheses are and ; is unrestricted. The conclusion is the inequality
where is the Euler Gamma function (as Complex.Gamma), is the complex modulus (‖·‖), and the exponent is a real number, so the power is the real power. Every quantifier is universal: the statement says that for all such and the inequality holds. The hypotheses are satisfiable (for example , ), so the statement is not vacuous. Two edge cases need attention. First, Lean's Complex.Gamma is total: it takes the value at the nonpositive integers, and Lean's complex division returns when the divisor is . The denominator is a genuine pole exactly when , that is with real and (here is forced). At those points the left side is by junk-value convention rather than by a mathematical limit, so the inequality holds trivially there and says nothing about the actual ratio, which is undefined. Everywhere else the denominator's real part is and it avoids every nonpositive integer, so the ratio is the honest Gamma quotient. Second, the numerator has real part at least , so it never hits a pole. The base is never zero, since its real part is at least , so the real power has no or ambiguity. The bound is the expected shape of a Stirling-type estimate: for large the Gamma ratio behaves like , whose modulus is , which matches the right side up to the shift . So the statement is a nontrivial, mathematically plausible bound, with the only trivial cases being the junk-value points on the real segment.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.