Estimates near an isolated zero #
A lower bound on a circle around a point at whose centre the function has its only zero, and two ways of concluding that the analytic order at a point is finite.
The first two are the standing hypotheses of every Rouché comparison made on such a disc: both
TauCeti.Analysis.Complex.Conformal.LocalDegree and
TauCeti.Analysis.Complex.Conformal.Hurwitz set up such a comparison, and each needs the same
two facts about the disc before Rouché can be applied. The third has no Rouché content at all: it
draws the same finiteness from the global hypothesis a zero count comes with, that the zeros in an
open set lie in a finite set, and serves the arbitrary-cycle argument principle
(TauCeti.Contour.argumentPrinciple_nullHomologous_of_analyticOnNhd). No result here mentions
Rouché's theorem, so this module does not import it:
Main results #
TauCeti.exists_pos_le_norm_of_mem_sphere: on the bounding circle the function is bounded below by a positive constant — this is what a competitor has to beat.TauCeti.analyticOrderAt_ne_top_of_forall_ne_zero: the analytic order at the centre is finite, so the count Rouché produces is a natural number rather than⊤.TauCeti.analyticOrderAt_ne_top_of_zeros_subset: the same finiteness under the hypothesis a global zero count comes with, that the zeros in an open set lie in a finite set.
They are separated here only because several files need them.
A continuous zero-free function on a sphere is bounded below there by a positive constant.
No hypothesis on the radius. At ρ = 0 the sphere is the set of points at zero distance from
a — the single point a when E is a metric space, but not in general, since the domain is
only assumed pseudometric. For ρ < 0 it is empty, nothing is attained, and the bound holds
vacuously.
A function with no zero off the centre does not vanish identically there. If f is
nonzero at every point of an open ball of positive radius other than its centre a, then f
has finite analytic order at a.
Nothing is assumed about f a, which may or may not be zero, nor about analyticity of f.
A function whose zeros are confined to a finite set has finite order. If every zero of f
in an open set U lies in a finite set S, then f does not vanish identically near any point of
U, so analyticOrderAt f z ≠ ⊤ there.
The point z itself is allowed to be a zero, indeed to lie in S. This is the finite-set form
of TauCeti.analyticOrderAt_ne_top_of_forall_ne_zero.