Recognising a semidirect sum from an ideal and a complementary subalgebra #
Mathlib's LieAlgebra.SemiDirectSum K L ψ, written K ⋊⁅ψ⁆ L, is the external semidirect sum
of two Lie algebras twisted by a Lie homomorphism ψ : L →ₗ⁅R⁆ LieDerivation R K K. A Lie
algebra L is presented internally as a semidirect sum by an ideal S, a Lie subalgebra H,
and the requirement that the two underlying submodules be complementary. This file connects the
two descriptions.
The twisting homomorphism is the adjoint action. An ideal S of L is stable under ⁅x, -⁆ for
every x : L, and the Jacobi identity says exactly that the resulting endomorphism of S is a Lie
derivation; the assignment is itself a homomorphism of Lie algebras, so it gives
LieIdeal.ad S : L →ₗ⁅R⁆ LieDerivation R S S. Restricting it along the inclusion of a
Lie subalgebra H produces the ψ that a semidirect sum needs, and when S and H are
complementary as submodules, (s, h) ↦ s + h is an isomorphism ↥S ⋊⁅ψ⁆ ↥H ≃ₗ⁅R⁆ L.
In the converse direction, an external semidirect sum K ⋊⁅ψ⁆ L carries such internal data
tautologically: the kernel of the projection to L is an ideal, the range of the inclusion of L
is a Lie subalgebra, and the two are complementary. So neither presentation is more general than
the other, and a theorem may be stated against the external form without loss.
The file also records two closure properties of the external semidirect sum. It is
module-finite when both factors are, by transport along the linear equivalence toProdl with the
product, and solvable when both factors are, because it is an extension of L by K: the range of
inl is the kernel of the surjection projr.
Main definitions #
LieIdeal.ad: the adjoint action ofLon an idealS, as a Lie homomorphismL →ₗ⁅R⁆ LieDerivation R S S. Theψattached to a Lie subalgebraHis the composite(LieIdeal.ad S).comp H.incl.LieIdeal.semiDirectSumHom: the Lie homomorphism↥S ⋊⁅ψ⁆ ↥H →ₗ⁅R⁆ Lgiven by(s, h) ↦ s + h.LieIdeal.semiDirectSumEquiv: that homomorphism as an isomorphism, when the underlying submodules ofSandHare complementary.
Main statements #
LieIdeal.nonempty_lieEquiv_semiDirectSumandLieIdeal.exists_lieEquiv_semiDirectSum: the recognition theorem, in theNonempty (L ≃ₗ⁅R⁆ ↥S ⋊⁅ψ⁆ ↥H)form that consumers of a splitting hypothesis take as input.LieIdeal.semiDirectSumEquiv_symm_apply_right_eq_zero_iffandLieIdeal.semiDirectSumEquiv_symm_apply_left_eq_zero_iff: the recognition isomorphism carriesSandHonto the two factors, which is what lets a statement about the left factor be read back as a statement about the ideal.LieAlgebra.SemiDirectSum.isCompl_ker_projr_range_inr: the converse, that an external semidirect sum is internally presented by the kernel ofprojrand the range ofinr, withLieAlgebra.SemiDirectSum.exists_lieEquiv_semiDirectSum_ker_projrthe resulting reconstruction.- Instances
Module.Finite R (K ⋊⁅ψ⁆ L)andIsSolvable (K ⋊⁅ψ⁆ L): an external semidirect sum of module-finite, resp. solvable, Lie algebras is module-finite, resp. solvable.
References #
- [W. Fulton and J. Harris, Representation Theory: A First Course][fulton-harris1991], Appendix E, §E.2, where Ado's theorem extends a representation of a solvable ideal across a complementary subalgebra.
The adjoint action of a Lie algebra on one of its ideals, as a homomorphism into the Lie
algebra of Lie derivations of that ideal. The ideal is stable under ⁅x, -⁆, and the Jacobi
identity is both the Leibniz rule for each ⁅x, -⁆ and the statement that x ↦ ⁅x, -⁆ preserves
brackets.
Equations
Instances For
The canonical map from the semidirect sum of an ideal S and a Lie subalgebra H of L,
twisted by the adjoint action of H on S, back to L: it adds the two components. It is a
homomorphism of Lie algebras because the twist is the adjoint action.
Equations
Instances For
Internal recognition of a semidirect sum. An ideal S and a Lie subalgebra H of L
whose underlying submodules are complementary exhibit L as the semidirect sum of S and H,
twisted by the adjoint action of H on S.
Equations
- S.semiDirectSumEquiv H h = LieEquiv.ofBijective (S.semiDirectSumHom H) ⋯
Instances For
The inverse of the recognition isomorphism is Mathlib's decomposition of an element of L
along the complementary pair S.toSubmodule, H.toSubmodule, read as an element of the
semidirect sum.
The recognition isomorphism identifies the ideal S with the left factor: an element of L
lies in S exactly when its preimage has vanishing right component.
The recognition isomorphism identifies the Lie subalgebra H with the right factor: an element
of L lies in H exactly when its preimage has vanishing left component.
The recognition theorem in the external form a splitting hypothesis is stated in: an ideal and
a complementary Lie subalgebra make L isomorphic to a semidirect sum.
The recognition theorem with the twisting homomorphism existentially quantified, the form in which a Levi-style decomposition theorem delivers its conclusion.
A semidirect sum of module-finite Lie algebras is module-finite.
A semidirect sum of solvable Lie algebras is solvable.
The range of the inclusion of the right factor consists of the elements whose left component
vanishes. This is not a simp lemma: Mathlib's LieHom.mem_range is already @[simp] and
rewrites the left-hand side to ∃ y, inr ψ y = z.
An external semidirect sum carries the internal data that recognises it: the kernel of the projection onto the right factor is an ideal, the range of the inclusion of the right factor is a Lie subalgebra, and their underlying submodules are complementary.
Every external semidirect sum is reconstructed by the internal recognition theorem applied to
the kernel of projr and the range of inr. Nothing is lost by stating a splitting hypothesis in
the external form.