Dual bases for the Killing form #
For a Lie algebra with nondegenerate Killing form, every basis has a Killing-dual basis. This file develops its coordinate equations and the canonical basis-independent contraction of a bilinear map against a basis and its Killing dual.
Main definitions #
Ado.killingDualBasis: the basis dual to a given one under the Killing form.
Main results #
Ado.sum_killingForm_killingDualBasis_eq_trace: contraction against a Killing-dual basis computes the trace.Ado.sum_apply_killingDualBasis_eq: contraction of a bilinear map against a basis and its Killing dual is independent of the basis.Ado.sum_apply_killingDualBasis_of_isAdjointPair: a Killing-adjoint pair may be moved between the two slots of the contraction.Ado.sum_apply_lie_killingDualBasis_add_eq_zero: the canonical contraction is invariant under the adjoint action.
The Killing-dual basis #
The basis of L dual to b under the Killing form: κ (b i) (killingDualBasis b j) is
1 when i = j and 0 otherwise (Ado.killingForm_killingDualBasis).
Equations
- Ado.killingDualBasis b = (killingForm K L).dualBasis ⋯ b
Instances For
The defining biorthogonality of the Killing-dual basis.
The symmetric orientation of the defining biorthogonality of the Killing-dual basis.
Coordinates in the Killing-dual basis are Killing pairings against b.
Conjugating twice returns the original basis: the Killing-dual basis of the Killing-dual
basis of b is b.
Coordinates in b are Killing pairings against the Killing-dual basis.
Expansion in b, with the coefficients read off by the Killing-dual basis.
Expansion in the Killing-dual basis, with the coefficients read off by b.
The sum ∑ᵢ κ (p xᵢ) yᵢ along a basis and its Killing-dual basis is the trace of p. The
Killing-dual basis reads off the coordinates in b (Ado.repr_eq_killingForm), so the sum is
the sum of the diagonal entries of the matrix of p.
The canonical invariant element ∑ᵢ xᵢ ⊗ yᵢ #
The sum ∑ᵢ f xᵢ yᵢ of a bilinear map along a basis and its Killing-dual basis does not
depend on the basis. Expanding each c j in the basis b produces exactly the coefficients
that expand killingDualBasis b i in killingDualBasis c.
A Killing-adjoint pair of operators may be moved from one slot to the other. If κ (p x) y
is κ x (q y) then applying p to the basis vectors and applying q to their Killing duals give
the same sum, because both expand to ∑ᵢ ∑ⱼ κ (p xᵢ) yⱼ • f xⱼ yᵢ.
The element ∑ᵢ xᵢ ⊗ yᵢ is invariant under the adjoint action. Read through a bilinear
map f, the Leibniz expansion of the adjoint action of z vanishes, because the two coefficient
families it produces are negatives of each other by the invariance of the Killing form.