Documentation

Hironaka.Resolution.Algebraic.Wlo05.EmbeddedRemaining

The remaining members after an isolation #

In the isolation passage of the proof of [Wlo05, Theorem 4.7.1], after the isolation the modified run restarts on (X_n, I_n : I_Γ, E_n) with the remaining components. This module shows that the invariant InvCE of the loop (Hironaka.Resolution.Algebraic.Wlo05.EmbeddedHypotheses) holds again for the strict transforms of the members not absorbed at the absorbing stage: each has a generic point η' over the member's generic point (the proof of [Kol07, Corollary 22]: the blow-ups are local isomorphisms at the generic point before the first containing centre), lying on no member of the boundary, off the absorbed strict transforms, so that the colon by I_Γ changes nothing at η' and the isolated ideal agrees there with the strict transform, which is the reduced ideal of the closure of η'; and η' is a generic point of the support of the isolated ideal. The remaining strict transforms are disjoint from the absorbed ones.

The disjointness from the absorbed strict transforms is a hypothesis of the general form exists_genericPoint_isolatedMarkedIdeal_of_not_absorbed_of_disjoint, discharged here from ClaimKC (a CONDITIONAL lemma, since ClaimKC is false) and unconditionally from CP1 in Hironaka.Resolution.Algebraic.Wlo05.EmbeddedCP1Remaining. Used in Hironaka.Resolution.Algebraic.Wlo05.EmbeddedIsolatedState, Hironaka.Resolution.Algebraic.Wlo05.EmbeddedCP1Remaining, Hironaka.Resolution.Algebraic.Wlo05.EmbeddedCP2Entry and Hironaka.Resolution.Algebraic.Wlo05.EmbeddedCP1Absorbed.

The data of a member not absorbed at the stage of the loop: the invariant of the loop gives its generic point η, off the boundary, at which T.I agrees with it; the stop rule gives no absorption before the stage of the loop.

theorem Hironaka.Resolution.disjoint_strictTransformSeq_of_not_absorbed {k : Type u} [Field k] [CharZero k] {T : MarkedTriple k} {hm : T.m = 1} {C : Finset T.X.left.IdealSheafData} {h : ∃ (n : ℕ), HasAbsorptionAt T hm C n} (hinv : InvCE T.I T.E C) (hKC : ClaimKC k) {c' : T.X.left.IdealSheafData} (hc' : c' ∈ C) (hnot : ¬CenterContains (bmoOneRun T hm) c' (Nat.find h)) {c : T.X.left.IdealSheafData} (hc : c ∈ absorbed T hm C h) :

A member not absorbed at the stage of the loop has its strict transform disjoint from every absorbed one (the disjointness of [Wlo05, Theorem 4.7.1] for the loop; conditional on ClaimKC).

The invariant is restored for a remaining member, with the disjointness from the absorbed strict transforms as a hypothesis (the proof of [Kol07, Corollary 22]): the strict transform of a member not absorbed at the stage of the loop is the reduced ideal of the closure of a generic point η' of the support of the isolated ideal, lying on no member of the boundary, at which the isolated ideal agrees with it.

The invariant is restored for a remaining member (the proof of [Kol07, Corollary 22]; conditional on ClaimKC, which supplies the disjointness): the strict transform of a member not absorbed at the stage of the loop is the reduced ideal of the closure of a generic point η' of the support of the isolated ideal, lying on no member of the boundary, at which the isolated ideal agrees with it.

theorem Hironaka.Resolution.eq_of_strictTransformSeq_eq {k : Type u} [Field k] [CharZero k] {T : MarkedTriple k} {hm : T.m = 1} {C : Finset T.X.left.IdealSheafData} {h : ∃ (n : ℕ), HasAbsorptionAt T hm C n} (hinv : InvCE T.I T.E C) {c₁ c₂ : T.X.left.IdealSheafData} (hc₁ : c₁ ∈ C) (hc₂ : c₂ ∈ C) (heq : (bmoOneRun T hm).strictTransformSeq c₁ (absorbIndex T hm C h).castSucc = (bmoOneRun T hm).strictTransformSeq c₂ (absorbIndex T hm C h).castSucc) :
c₁ = c₂

The strict transform at the stage of the loop determines the member (the proof of [Kol07, Corollary 22]): two members with the same strict transform have the same generic point, the unique point over it being the generic point of the strict transform.