The simply connected root datum of type E₆ #
This file constructs the pinned integral root datum of type E₆ on the character and cocharacter
lattices Fin 6 → ℤ. The character lattice is written in the fundamental-weight basis and the
cocharacter lattice in the simple-coroot basis, so that the i-th simple root is the i-th row of
the Bourbaki-numbered Cartan matrix CartanMatrix.E 6 and the i-th simple coroot is the i-th
standard basis vector.
The coordinates #
A coroot is stored by its coordinates in the simple coroots, and the corresponding root by its
pairings against the simple coroots. Because E₆ is simply laced, a root and its coroot have the
same simple-root coordinates, so a single stored vector determines both: if a coroot has
coordinates c, then the root has coordinates ⟨β, αⱼ^∨⟩ = ∑ i, cᵢ ⟨αᵢ, αⱼ^∨⟩, that is, the
Cartan-matrix image of c. That relation is TauCeti.DynkinType.e6Root_eq_mulVec, and it is what
makes every later step structural rather than a second table: the root reflections are the image of
the coroot reflections under a linear map, and the pairing is symmetric because CartanMatrix.E 6
is.
The seventy-two roots — the count fixed by TauCeti.DynkinType.numRoots — are ordered with the six
simple roots first, in Bourbaki order, then the remaining thirty positive roots by height and then
lexicographically in their simple-root coordinates, and finally the thirty-six negative roots in
the matching order. The last positive root is therefore the highest root, whose simple-root
coordinates are the Bourbaki marks (1, 2, 2, 3, 2, 1).
Only the coroots are asked to span their lattice, and only that half is recorded, in
TauCeti.DynkinType.corootSpan_e6SimplyConnectedRootDatum_eq_top. The roots span the root lattice,
which sits inside the weight lattice with index 3 — the determinant of CartanMatrix.E 6 — so the
datum carries no RootPairing.IsRootSystem instance. That asymmetry is what "simply connected"
means here.
Main definitions #
TauCeti.DynkinType.e6CorootandTauCeti.DynkinType.e6Root: the coroot coordinate embedding and the root embedding derived from it by the Cartan-matrix map.TauCeti.DynkinType.e6SimplyConnectedRootDatum: the pinned root datum of typeE₆.TauCeti.DynkinType.e6SimpleIndex: the first six root indices, the Bourbaki-numbered simple roots.TauCeti.DynkinType.e6SimplyConnectedBase: the base they form.
Main results #
TauCeti.DynkinType.e6Root_eq_mulVec: the root coordinates are the Cartan-matrix image of the coroot coordinates.TauCeti.DynkinType.root_e6SimpleIndexandTauCeti.DynkinType.coroot_e6SimpleIndex: thei-th simple root is thei-th row ofCartanMatrix.E 6and thei-th simple coroot isPi.single i 1, which is what pins the two lattices as the weight and coroot lattices.TauCeti.DynkinType.hasCartanType_e6SimplyConnectedRootDatum: the pinned base has Cartan typeE6.TauCeti.DynkinType.corootSpan_e6SimplyConnectedRootDatum_eq_top: the coroots span the cocharacter lattice, the simply connected condition.
References #
The coordinates and the node numbering follow Bourbaki, Lie Groups and Lie Algebras, Chapters
4--6, Plate V, and Humphreys, Introduction to Lie Algebras and Representation Theory, section
12.1. This is the E₆ branch of the target "a named datum per valid type" in Layer 6 of
TauCetiRoadmap/RepresentationTheory/RootSystems/README.md.
The coordinate data #
The coroots of type E₆ in the simple-coroot basis. The first six are the standard basis
vectors, the first thirty-six are the positive coroots, and the last thirty-six are their
negatives.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The roots of type E₆ in the fundamental-weight basis, ordered compatibly with
TauCeti.DynkinType.e6Coroot. They are derived from the coroot coordinates by the Cartan-matrix
map, and the first six are the rows of CartanMatrix.E 6.
Equations
- TauCeti.DynkinType.e6Root = { toFun := fun (i : Fin 72) => (CartanMatrix.E 6).mulVec (TauCeti.DynkinType.e6Coroot i), inj' := TauCeti.DynkinType.e6Root._proof_1✝ }
Instances For
The three blocks of root indices #
The i-th simple root of type E₆ sits at root index i, the Bourbaki node i + 1.
Equations
Instances For
The i-th positive root of type E₆ sits at root index i.
Equations
Instances For
The negative of the i-th positive root of type E₆ sits at root index i + 36.
Equations
Instances For
The Cartan matrix as the bridge between the two embeddings #
The root coordinates are the Cartan-matrix image of the coroot coordinates.
Writing a coroot as β^∨ = ∑ i, cᵢ αᵢ^∨ in the simple coroots, the pairings of β against the
simple coroots are ⟨β, αⱼ^∨⟩ = ∑ i, cᵢ ⟨αᵢ, αⱼ^∨⟩, which is the j-th entry of the
Cartan-matrix image of c because CartanMatrix.E 6 is symmetric.
This is deliberately not a simp lemma: unfolding a root into the Cartan-matrix image of its
coroot is a change of representation, not a normal form, and it would take the left-hand sides of
the lemmas below out of normal form.
The simple coroots are the standard basis. This is what pins the cocharacter lattice as the coroot lattice, so that the datum is the simply connected one.
The simple roots are the rows of the Cartan matrix. In the fundamental-weight basis the
i-th simple root of the pinned type E₆ datum is the i-th row of CartanMatrix.E 6, which is
what pins the character lattice as the weight lattice.
The pairing of the pinned tables is symmetric. This is the simply-laced feature of E₆: the
pairing is the symmetric bilinear form attached to CartanMatrix.E 6, read on the shared
simple-root coordinates of a root and its coroot.
The second half of the table lists the negatives of the coroots in the first half.
The second half of the table lists the negatives of the roots in the first half.
The reflections #
The pinned datum #
The pinned simply connected root datum of type E₆.
Both lattices are Fin 6 → ℤ: the character lattice in the fundamental-weight basis and the
cocharacter lattice in the simple-coroot basis. Root indices 0 through 5 are the Bourbaki
simple roots; see TauCeti.DynkinType.root_e6SimpleIndex.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The root embedding of the pinned E₆ datum is the explicit table e6Root.
The coroot embedding of the pinned E₆ datum is the explicit table e6Coroot.
The perfect pairing of the pinned E₆ datum is the dot product of coordinate vectors, the
fundamental-weight and simple-coroot bases being dual to one another.
Pairing a pinned E₆ root with a coroot computes as their coordinate dot product.
The coroots of the pinned type E₆ datum span the cocharacter lattice. This is the simply
connected lattice condition required by the pinned Chevalley--Demazure construction. Its
counterpart for the roots is deliberately absent: they span the root lattice, which sits inside the
weight lattice with index 3 (Bourbaki, Plate V).
The pinned base #
The Bourbaki-numbered base of the pinned simply connected root datum of type E₆. Its support
is the set of the first six root indices, carrying the simple roots in Bourbaki order.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Membership in the pinned base support is exactly membership among the first six root indices.
The Cartan integers at the first six root indices are Mathlib's Bourbaki-numbered E₆
matrix. This pins the node order independently of the existential relabelling in
TauCeti.HasCartanType.
The pinned datum of type E₆ has Cartan type E6. Its Bourbaki-numbered base realizes the
standard Cartan matrix CartanMatrix.E 6, with the node numbering of TauCeti.DynkinType.