Documentation

Hironaka.Resolution.Analytic.OrderReduction.Modified.NonmonomialIndiff

The nonmonomial part is indifferent to empty boundary members #

The counterpart, for boundary members, of the empty blow-up convention [Kol07, 32], for the decomposition I = M(I) · N(I) of [Kol07, Definition–Lemma 110] in its fine form: deleting empty members of the boundary changes neither the fine monomial part M(I) nor the nonmonomial part N(I) = I : M(I). Two boundary families G, G' with simple normal crossings which differ by an empty extension along an order embedding e : G.ι ↪o G'.ι (HypersurfaceFamily.IsEmptyExtension: the members correspond along e, the members outside its range are empty) have the same connected components: an empty member has none, and the components of corresponding members are the same subsets of M (emptyExtIdx, a bijection of the component indices which preserves the component sets and their chosen representatives). So the stalks of M(I), finite products of the component factors (stalkIdeal_monomialPart), agree (Finset.prod_bij), hence M(I) and N(I) agree (IdealSheaf.ofStalks_congr). The nonmonomial triple of a triple with an empty extension of the boundary is therefore the nonmonomial triple of the original with the boundary replaced (nonmonomialTriple_eq_of_isEmptyExtension): the input of the indifference property of the rounds on the nonmonomial part (the value of bo d read at the nonmonomial triple), and of the transport of the invariant of the descent through the deletion of empty blow-ups.

Connected components along an equality of subsets #

Transport of connected components along an equality of subsets (a cast).

Equations
Instances For
    theorem Hironaka.Manifold.BMOmod.ccCast_ccCast {M : Type u} [TopologicalSpace M] {s t : Set M} (h : s = t) (C : ConnectedComponents ↑s) :
    ccCast ⋯ (ccCast h C) = C
    theorem Hironaka.Manifold.BMOmod.out_val_ccCast {M : Type u} [TopologicalSpace M] {s t : Set M} (h : s = t) (C : ConnectedComponents ↑s) :
    ↑(Quotient.out (ccCast h C)) = ↑(Quotient.out C)

    The chosen representative of a transported component is the representative of the component (the transport is a cast, so the two Quotient.out are the same term after subst).

    The underlying subset of a transported component is the underlying subset of the component.

    The components of an empty extension #

    The component of G' corresponding to a component ⟨j, C⟩ of G under an empty extension along e: the member e j (whose set is G.hyp j) with the component C transported.

    Equations
    Instances For

      Every component of G' comes from a component of G: a member outside the range of e is empty (he.2), so it has no connected component.

      The nonempty members of G are finite when those of G' are (they inject along e).

      The nonempty members of G' are finite when those of G are (every nonempty member of G' is the image of a member of G).

      The fine monomial part and the nonmonomial part #

      Corresponding components have the same reduced ideal sheaf (the ideal sheaf of a closed submanifold depends only on the subset, IsClosedSubmanifold.idealSheaf_congr).

      Corresponding components have the same exponent ord_D I (the same ideal sheaf, read at the same representative).

      Corresponding components have the same fine monomial factor 𝓘_D ^ (ord_D I).

      The stalk products which define M(I) agree: the components through x correspond along emptyExtIdx (same component sets), with the same factors.

      theorem Manifold.IdealSheaf.ofStalks_congr {X : TopCat} {𝒪 : TopCat.Sheaf CommRingCat X} {f g : (x : ↑X) → Ideal ↑(𝒪.presheaf.stalk x)} (h : f = g) (hf : HasLocalGenerators f) (hg : HasLocalGenerators g) :
      ofStalks 𝒪 f hf = ofStalks 𝒪 g hg

      Two ideal sheaves prescribed by the same stalk ideals are equal (the generator proofs are irrelevant).

      The fine monomial part is indifferent to empty members: M(I) for G' is M(I) for G under an empty extension G ↪ G' (the counterpart, for boundary members, of [Kol07, 32], for the decomposition of [Kol07, Definition–Lemma 110]).

      The nonmonomial part is indifferent to empty members: N(I) = I : M(I) for G' is N(I) for G under an empty extension G ↪ G' (the counterpart, for boundary members, of [Kol07, 32], for the decomposition of [Kol07, Definition–Lemma 110]).

      The nonmonomial triple #

      theorem Hironaka.Manifold.BMOmod.nonmonomialTriple_eq_of_isEmptyExtension {𝕜 : Type} [RCLike 𝕜] {E : Type u_1} [NormedAddCommGroup E] [NormedSpace 𝕜 E] {n : ℕ} {ψ₀ : E ≃L[𝕜] Fin n → 𝕜} {X : AnalyticManifold 𝕜 E} (T : Manifold.AnalyticTriple ψ₀ X) (F' : Manifold.HypersurfaceFamily X.carrier) (hsnc' : Manifold.HypersurfaceFamily.IsSnc ψ₀ F') {e : F'.ι ↪o T.F.ι} (he : Manifold.HypersurfaceFamily.IsEmptyExtension e) :
      BMO.nonmonomialTriple { I := T.I, isNonzeroEverywhere := ⋯, F := F', isSnc := hsnc' } = { I := (BMO.nonmonomialTriple T).I, isNonzeroEverywhere := ⋯, F := F', isSnc := hsnc' }

      The nonmonomial triple of a triple with an empty extension of the boundary is the nonmonomial triple of the original with the boundary replaced: the ideal N(𝓘) is the same (nonmonomialPart_eq_of_isEmptyExtension), the boundary is the replaced one. This is the input of the indifference property of the rounds on the nonmonomial part, the value of bo d being read at the nonmonomial triple (the counterpart, for boundary members, of [Kol07, 32]).