Restricted series lift along an open surjection #
A continuous open surjection ฯ : A โ B of nonarchimedean rings whose source has countably
generated ๐ 0 induces a surjection of restricted power-series rings โ the trivial-weight
TauCeti.Huber.weightedRestrictedSubring, before completion โ coefficientwise.
Openness is the hypothesis that matters. Surjectivity of ฯ alone lifts each coefficient of a
restricted series separately, but the preimages so chosen need not tend to zero, and then the lift
is a power series and not a restricted one. Openness is what lets the preimages be drawn from a
shrinking family of neighbourhoods; TauCeti.exists_lift_tendsto_cofinite_nhds is where that
choice is made, and this file is its transcription into the weighted language, at the trivial
weight family where restrictedness is convergence to zero
(TauCeti.Huber.isWeightedRestricted_one_weight_iff).
Countable generation of ๐ (0 : A) is a hypothesis of both results here, not a background
assumption: it is what supplies the shrinking family the lifted coefficients are drawn from, and a
general NonarchimedeanRing need not have it. It is not restrictive in the intended application โ
a Huber ring satisfies it, by TauCeti.Huber.IsHuberRing.isCountablyGenerated_nhds_zero.
Main results #
TauCeti.Huber.weightedMap_one_weight_surjective, withIsOpenQuotientMap.weightedMap_one_weight_surjectivethe form taking the bundled structure โ in that namespace sohq.weightedMap_one_weight_surjectiveworks.
References #
- T. Wedhorn, Adic Spaces (arXiv:1910.05934v1), ยง5.6 and Proposition & Definition 6.36.
AINTLIB (github.com/CBirkbeck/AINTLIB, Apache-2.0) at commit
37bbdaeb9ad9e3bc9f0d660feadc2779e455a91c,
projects/AdicSpaces/Adic spaces/RestrictedModule.lean, has the corresponding statement for
modules in one variable, restrictedModule_map_surjective, with both sides assumed Hausdorff.
Nothing was copied.
A continuous surjection carrying neighbourhoods of zero onto neighbourhoods of zero stays
surjective on restricted series, provided ๐ (0 : A) is countably generated. Every restricted
series over B is then the image of one over A.
The hypothesis is stated as the filter inequality the proof consumes. Alongside the surjectivity
assumed here it is the filter-level formulation of IsOpenMap ฯ, not a weakening of it โ for a
surjective continuous additive map the two say the same thing, since openness of a group
homomorphism is decided at zero. IsOpenQuotientMap.weightedMap_one_weight_surjective is the form
for a caller holding the bundled open-quotient structure.
This is a step towards what Wedhorn's Proposition & Definition 6.36(ii) needs, not the whole of
it. TauCeti.Huber.IsStrictlyTopologicallyFiniteType asks for an open quotient map whose domain
is TauCeti.Huber.restrictedMvPowerSeriesCompletion k A โ the completion AโจXโ,โฆ,Xโโฉ at the
trivial weight โ whereas the surjection here is one level below, between the restricted subrings
themselves. Passing from this to an open quotient out of the completion is a separate step.
TauCeti.Huber.IsTopologicallyFiniteType is the weaker notion, and what weakens it is the
weight family, not completion: it allows an arbitrary finite family T in place of the trivial
one. Both notions take their quotient out of a completion.
The open-quotient form of TauCeti.Huber.weightedMap_one_weight_surjective, for a caller
holding the bundled IsOpenQuotientMap ฯ.
That is the interface finite-type presentations come in โ TauCeti.Huber
.IsStrictlyTopologicallyFiniteType produces one โ so it is the form a consumer actually has,
and it bundles exactly the continuity, surjectivity and openness the filter-level theorem needs.