Morphisms surjective on stalks #
We define the class of morphisms between schemes that are surjective on stalks. We show that this class is stable under composition and base change.
We also show that (AlgebraicGeometry.SurjectiveOnStalks.isEmbedding_pullback)
if Y ⟶ S is surjective on stalks, then for every X ⟶ S, X ×ₛ Y is a subset of
X × Y (Cartesian product as topological spaces) with the induced topology.
The class of morphisms f : X ⟶ Y between schemes such that
𝒪_{Y, f x} ⟶ 𝒪_{X, x} is surjective for all x : X.
- stalkMap_surjective (x : ↥X) : Function.Surjective ⇑(CategoryTheory.ConcreteCategory.hom (Scheme.Hom.stalkMap f x))
Instances
theorem AlgebraicGeometry.surjectiveOnStalks_iff {X Y : Scheme} (f : X ⟶ Y) :
SurjectiveOnStalks f ↔ ∀ (x : ↥X), Function.Surjective ⇑(CategoryTheory.ConcreteCategory.hom (Scheme.Hom.stalkMap f x))
theorem AlgebraicGeometry.Scheme.Hom.stalkMap_surjective {X Y : Scheme} (f : X ⟶ Y) [self : SurjectiveOnStalks f] (x : ↥X) :
Alias of AlgebraicGeometry.SurjectiveOnStalks.stalkMap_surjective.
@[instance 900]
instance AlgebraicGeometry.SurjectiveOnStalks.instOfIsOpenImmersion {X Y : Scheme} (f : X ⟶ Y) [IsOpenImmersion f] :
instance AlgebraicGeometry.SurjectiveOnStalks.comp {X Y Z : Scheme} (f : X ⟶ Y) (g : Y ⟶ Z) [SurjectiveOnStalks f] [SurjectiveOnStalks g] :
theorem AlgebraicGeometry.SurjectiveOnStalks.eq_stalkwise :
@SurjectiveOnStalks = stalkwise fun {R S : Type u_1} [CommRing R] [CommRing S] (x : R →+* S) => Function.Surjective ⇑x
instance AlgebraicGeometry.SurjectiveOnStalks.instHasRingHomPropertySurjectiveOnStalks :
HasRingHomProperty @SurjectiveOnStalks fun {R S : Type u_1} [CommRing R] [CommRing S] => RingHom.SurjectiveOnStalks
theorem AlgebraicGeometry.SurjectiveOnStalks.iff_of_isAffine {X Y : Scheme} {f : X ⟶ Y} [IsAffine X] [IsAffine Y] :
theorem AlgebraicGeometry.SurjectiveOnStalks.of_comp {X Y Z : Scheme} (f : X ⟶ Y) (g : Y ⟶ Z) [SurjectiveOnStalks (CategoryTheory.CategoryStruct.comp f g)] :
theorem AlgebraicGeometry.SurjectiveOnStalks.mono_of_injective {X Y : Scheme} {f : X ⟶ Y} [SurjectiveOnStalks f] (hf : Function.Injective ⇑f) :
theorem AlgebraicGeometry.SurjectiveOnStalks.isEmbedding_pullback {X Y S : Scheme} (f : X ⟶ S) (g : Y ⟶ S) [SurjectiveOnStalks g] :
Topology.IsEmbedding fun (x : ↥(CategoryTheory.Limits.pullback f g)) =>
((CategoryTheory.Limits.pullback.fst f g) x, (CategoryTheory.Limits.pullback.snd f g) x)
If Y ⟶ S is surjective on stalks, then for every X ⟶ S, X ×ₛ Y is a subset of
X × Y (Cartesian product as topological spaces) with the induced topology.