Documentation

TauCeti.Algebra.Lie.Multiplicity

The multiplicity of an irreducible Lie module #

LieModule.isotypicMultiplicity R L M S is the finrank of the space S →ₗ⁅R,L⁆ M of morphisms from a Lie module S to a Lie module M. This file proves that when S is finite-dimensional and irreducible over an algebraically closed field, this number counts the summands equivalent to S in any finite decomposition of M into irreducible Lie submodules, and is therefore independent of which such decomposition is chosen: it is the multiplicity of S in M.

The argument #

The morphism space is additive over a direct sum in its target: a morphism S →ₗ⁅R,L⁆ ⨁ i, Pᵢ is the same thing as a family of morphisms S →ₗ⁅R,L⁆ Pᵢ, one for each i, which is TauCeti.LieModule.lieModuleHomDirectSumEquiv of TauCeti/Algebra/Lie/DirectSum.lean. Transporting along DirectSum.lieModuleEquivOfIsInternal of TauCeti/Algebra/Lie/Submodule/DirectSum.lean turns that into additivity over an internal decomposition of M by Lie submodules.

The per-summand contribution is Schur's lemma in the dimension form of TauCeti/Algebra/Lie/Schur.lean: for irreducible S and Nᵢ over an algebraically closed field, dim_K (S →ₗ⁅K,L⁆ Nᵢ) is 1 when Nᵢ ≃ S and 0 otherwise. Summing over the finite index type gives the count. Since the left-hand side never mentions the decomposition, two finite decompositions of the same module have the same number of summands equivalent to S; this is the uniqueness statement that makes "the multiplicity of S in M" well defined.

Nothing here needs complete reducibility: the counting theorem is stated for a finite decomposition it is handed. Complete reducibility is what produces such a decomposition, through TauCeti.exists_isInternal_isIrreducible of TauCeti/Algebra/Lie/Submodule/Decomposition.lean, which a consumer combines with the counting theorem below.

Implementation notes #

The multiplicity joins the Lie isotypy interface of TauCeti/Algebra/Lie/Isotypic.lean (LieModule.isotypicComponent, LieModule.IsIsotypicOfType), and like that interface it is stated without the universal enveloping algebra: this layer depends only on Lie submodules and Lie-module morphisms. The comparison with the ring-level multiplicity theory of TauCeti/RingTheory/Semisimple/Multiplicity.lean is therefore made where the rest of the enveloping-algebra dictionary lives, LieModule.isotypicMultiplicity_eq_natCard_of_linearEquiv_pi of TauCeti/Algebra/Lie/UniversalEnveloping/Multiplicity.lean, exactly as TauCeti/Algebra/Lie/UniversalEnveloping/Isotypic.lean does for the isotypic component.

Main definitions #

Main results #

Roadmap #

This is the isotypicMultiplicity target of the decomposition toolkit in Layer 6 of TauCetiRoadmap/RepresentationTheory/LieHighestWeight/README.md, together with the additivity of the morphism space over a direct-sum decomposition that TauCeti/Algebra/Lie/Schur.lean names as the missing ingredient of the multiplicity theorem. The multiplicity is stated for an arbitrary irreducible S rather than for the irreducible quotient L(λ), which the roadmap's own signature uses; the L(λ)-indexed form is this statement with S instantiated. The finrank of the isotypic component, dim (LieModule.isotypicComponent S M) = isotypicMultiplicity · dim S, is the next item of that milestone, "Isotypic components and multiplicities, through the enveloping-algebra dictionary"; it is proved through the dictionary, as LieModule.finrank_isotypicComponent of TauCeti/Algebra/Lie/UniversalEnveloping/Multiplicity.lean. The packaged decomposition M ≃ ⨁_λ L(λ)^{⊕ m_λ} that closes the milestone is TauCeti.nonempty_lieModuleEquiv_directSum_irreducibleQuotient of TauCeti/Algebra/Lie/HighestWeight/Decomposition.lean, which counts the summands of a decomposition with the theorems below. The toolkit's remaining milestone, the tensor multiplicities with the minuscule Pieri rule, is untouched here.

References #

The multiplicity of an irreducible module #

noncomputable def LieModule.isotypicMultiplicity (R : Type u) (L : Type v) (M : Type w) (S : Type w₁) [CommRing R] [LieRing L] [LieAlgebra R L] [AddCommGroup M] [Module R M] [LieRingModule L M] [LieModule R L M] [AddCommGroup S] [Module R S] [LieRingModule L S] :

The finrank of the space of morphisms S →ₗ⁅R,L⁆ M. For an irreducible, finite-dimensional S over an algebraically closed field this is the multiplicity of S in M: the number of summands equivalent to S in any finite decomposition of M into irreducibles (LieModule.isotypicMultiplicity_eq_ncard_of_isInternal), and stating it as the finrank of a morphism space is what makes it manifestly independent of the decomposition. In the generality of the definition it is not itself a multiplicity: for a finite-dimensional irreducible S over a field and a finite irreducible direct-sum decomposition of M, it is the multiplicity times dim End_L(S), a factor that Schur's lemma makes 1 over an algebraically closed field.

Equations
Instances For
    theorem LieModule.isotypicMultiplicity_def (R : Type u) (L : Type v) (M : Type w) (S : Type w₁) [CommRing R] [LieRing L] [LieAlgebra R L] [AddCommGroup M] [Module R M] [LieRingModule L M] [LieModule R L M] [AddCommGroup S] [Module R S] [LieRingModule L S] :

    The multiplicity is the finrank of the morphism space. The definition is not exposed, so this is how it is unfolded.

    @[simp]

    The zero module has multiplicity zero: the only morphism out of it is the zero morphism.

    @[simp]

    Everything occurs in the zero module with multiplicity zero: the only morphism into it is the zero morphism.

    The multiplicity depends only on the two equivalence classes #

    theorem LieModule.isotypicMultiplicity_eq_of_lieModuleEquiv {R : Type u} {L : Type v} {M : Type w} {M' : Type w₂} {S : Type w₁} [CommRing R] [LieRing L] [LieAlgebra R L] [AddCommGroup M] [Module R M] [LieRingModule L M] [LieModule R L M] [AddCommGroup M'] [Module R M'] [LieRingModule L M'] [LieModule R L M'] [AddCommGroup S] [Module R S] [LieRingModule L S] (e : M ≃ₗ⁅R,L⁆ M') :

    The multiplicity only depends on the equivalence class of the ambient module. Postcomposition with an equivalence identifies the two morphism spaces.

    theorem LieModule.isotypicMultiplicity_eq_of_lieModuleEquiv_type {R : Type u} {L : Type v} {M : Type w} {S : Type w₁} {S' : Type w₃} [CommRing R] [LieRing L] [LieAlgebra R L] [AddCommGroup M] [Module R M] [LieRingModule L M] [LieModule R L M] [AddCommGroup S] [Module R S] [LieRingModule L S] [AddCommGroup S'] [Module R S'] [LieRingModule L S'] (e : S ≃ₗ⁅R,L⁆ S') :

    The multiplicity only depends on the equivalence class of the module being counted. Precomposition with an equivalence identifies the two morphism spaces. This is what lets the multiplicity be read for an irreducible determined only up to equivalence, such as an irreducible highest-weight quotient.

    The multiplicity as a count of summands #

    theorem LieModule.isotypicMultiplicity_eq_sum_of_isInternal {K : Type u} {L : Type v} {M : Type w} {S : Type w₁} [Field K] [LieRing L] [LieAlgebra K L] [AddCommGroup M] [Module K M] [LieRingModule L M] [LieModule K L M] [AddCommGroup S] [Module K S] [LieRingModule L S] {ι : Type w₂} [DecidableEq ι] [Fintype ι] (N : ι → LieSubmodule K L M) (h : DirectSum.IsInternal fun (i : ι) => ↑(N i)) (hfin : ∀ (i : ι), FiniteDimensional K (S →ₗ⁅K,L⁆ ↥(N i))) :
    isotypicMultiplicity K L M S = ∑ i : ι, isotypicMultiplicity K L (↥(N i)) S

    Multiplicity is additive over an internal decomposition of its ambient module.

    @[simp]
    theorem LieModule.isotypicMultiplicity_self {K : Type u} {L : Type v} {S : Type w₁} [Field K] [LieRing L] [LieAlgebra K L] [AddCommGroup S] [Module K S] [LieRingModule L S] [LieModule K L S] [IsAlgClosed K] [FiniteDimensional K S] [IsIrreducible K L S] :

    An irreducible module occurs in itself with multiplicity one.

    theorem LieModule.isotypicMultiplicity_eq_ncard_of_isInternal {K : Type u} {L : Type v} {M : Type w} {S : Type w₁} [Field K] [LieRing L] [LieAlgebra K L] [AddCommGroup M] [Module K M] [LieRingModule L M] [LieModule K L M] [AddCommGroup S] [Module K S] [LieRingModule L S] [LieModule K L S] [IsAlgClosed K] [FiniteDimensional K S] [IsIrreducible K L S] {ι : Type w₂} [DecidableEq ι] [Finite ι] (N : ι → LieSubmodule K L M) (h : DirectSum.IsInternal fun (i : ι) => ↑(N i)) (hirr : ∀ (i : ι), IsIrreducible K L ↥(N i)) :
    isotypicMultiplicity K L M S = {i : ι | Nonempty (S ≃ₗ⁅K,L⁆ ↥(N i))}.ncard

    The multiplicity counts the summands equivalent to S. For a finite decomposition of a module into irreducible Lie submodules over an algebraically closed field, the multiplicity of a finite-dimensional irreducible S is the number of indices whose summand is equivalent to S: Schur's lemma makes each such summand contribute 1 to the morphism space and every other summand contribute 0.

    theorem LieModule.ncard_setOf_nonempty_lieModuleEquiv_eq {K : Type u} {L : Type v} {M : Type w} {S : Type w₁} [Field K] [LieRing L] [LieAlgebra K L] [AddCommGroup M] [Module K M] [LieRingModule L M] [LieModule K L M] [AddCommGroup S] [Module K S] [LieRingModule L S] [LieModule K L S] [IsAlgClosed K] [FiniteDimensional K S] [IsIrreducible K L S] {ι : Type w₂} [DecidableEq ι] [Finite ι] {ι' : Type w₃} [DecidableEq ι'] [Finite ι'] (N : ι → LieSubmodule K L M) (h : DirectSum.IsInternal fun (i : ι) => ↑(N i)) (hirr : ∀ (i : ι), IsIrreducible K L ↥(N i)) (N' : ι' → LieSubmodule K L M) (h' : DirectSum.IsInternal fun (j : ι') => ↑(N' j)) (hirr' : ∀ (j : ι'), IsIrreducible K L ↥(N' j)) :
    {i : ι | Nonempty (S ≃ₗ⁅K,L⁆ ↥(N i))}.ncard = {j : ι' | Nonempty (S ≃ₗ⁅K,L⁆ ↥(N' j))}.ncard

    The number of summands equivalent to a given irreducible does not depend on which finite decomposition is taken. Both counts compute the same multiplicity, which is defined without reference to any decomposition.