Documentation

LeanPool.RegtsSevenster.RS.Classical.Deligne.FibreExact

Exactness of the fibre functor from a base-change section #

A short exact sequence of an abelian category does not split, so the splitting hypothesis of RS.fibreFun_shortExact is never available as stated. What Deligne's 2.10 supplies — in the form recorded by RS.Rappel210Statement — is weaker and is exactly what is needed: a section of the epimorphism after base change, that is, a morphism of module objects s : freeMod R S.X₃ ⟶ freeMod R S.X₂ splitting freeModMap R S.g.

This module derives short exactness of the realised sequence from that hypothesis alone. The route is:

The category of module objects carries no additive structure, so the transported splitting is assembled by hand rather than through ShortComplex.Splitting.map; RS.gammaModuleFunctor_map_add is the one additivity statement this needs, phrased on underlying morphisms.

Exactness of base change is assumed as an instance hypothesis on the ambient category; for Ind C it is supplied by RS.tensorLeft_ind_preservesFiniteLimits and its colimit counterpart.

Intertwining laws in raw tensor form #

theorem RS.lin_of_complement {D : Type u} [CategoryTheory.Category.{v, u} D] [CategoryTheory.MonoidalCategory D] (R : D) [CategoryTheory.MonObj R] [CategoryTheory.Preadditive D] [CategoryTheory.MonoidalPreadditive D] {V W : D} (f : CategoryTheory.MonoidalCategoryStruct.tensorObj R V ⟶ CategoryTheory.MonoidalCategoryStruct.tensorObj R W) [CategoryTheory.Mono f] (hf : CategoryTheory.CategoryStruct.comp (CategoryTheory.CategoryStruct.comp (CategoryTheory.MonoidalCategoryStruct.associator R R V).inv (CategoryTheory.MonoidalCategoryStruct.whiskerRight CategoryTheory.MonObj.mul V)) f = CategoryTheory.CategoryStruct.comp (CategoryTheory.MonoidalCategoryStruct.whiskerLeft R f) (CategoryTheory.CategoryStruct.comp (CategoryTheory.MonoidalCategoryStruct.associator R R W).inv (CategoryTheory.MonoidalCategoryStruct.whiskerRight CategoryTheory.MonObj.mul W))) (u : CategoryTheory.MonoidalCategoryStruct.tensorObj R W ⟶ CategoryTheory.MonoidalCategoryStruct.tensorObj R V) (c : CategoryTheory.MonoidalCategoryStruct.tensorObj R W ⟶ CategoryTheory.MonoidalCategoryStruct.tensorObj R W) (hc : CategoryTheory.CategoryStruct.comp (CategoryTheory.CategoryStruct.comp (CategoryTheory.MonoidalCategoryStruct.associator R R W).inv (CategoryTheory.MonoidalCategoryStruct.whiskerRight CategoryTheory.MonObj.mul W)) c = CategoryTheory.CategoryStruct.comp (CategoryTheory.MonoidalCategoryStruct.whiskerLeft R c) (CategoryTheory.CategoryStruct.comp (CategoryTheory.MonoidalCategoryStruct.associator R R W).inv (CategoryTheory.MonoidalCategoryStruct.whiskerRight CategoryTheory.MonObj.mul W))) (hid : CategoryTheory.CategoryStruct.comp u f + c = CategoryTheory.CategoryStruct.id (CategoryTheory.MonoidalCategoryStruct.tensorObj R W)) :

The complement of a module endomorphism is a module map. Let f be a monomorphism intertwining the free actions, let c be an endomorphism intertwining them, and let u satisfy u ≫ f + c = 𝟙. Then u intertwines the free actions as well: the law for u is checked after f, where it becomes the law for 𝟙 - c. This is the crux of the whole module.

The retraction produced by a base-change section #

Transport to the super modules #

Realisation turns a sum of module morphisms into a sum. The category of module objects has no additive structure, so the hypothesis is stated on underlying morphisms.

The splitting of the realised sequence. Its retraction and section are the realisations of the module retraction and of the given module section; the three identities are the images of the corresponding identities in the category of module objects.

Equations
Instances For

    The fibre functor carries a short exact sequence with a base-change section to a short exact sequence of super modules. No section in the ambient category is required: a section of the base-changed epimorphism as a map of module objects suffices.