A fundamental system of a root system is a subset which is a basis of and for which every has an expansionwhose coefficients are either all nonnegative or all nonpositive. Its associated positive system of a root system isThus , and the elements of are the simple roots.
The Weyl reflection in satisfiesNow take and expand it in the basis . At least one coefficient belonging to a simple root other than is positive. Sincethe reflection changes only the coefficient of . Every root has simple-root coefficients of one sign, so the unchanged positive coefficient prevents from being negative. Hence . Because is an involution, it permutes , while it exchanges and . Therefore
Let be the fundamental chamber of a root system. The chambers and are adjacent across the reflecting hyperplane orthogonal to . The chamber lies on the side on which is positive. If , then lies on that same side, so crossing this wall moves one step farther from ; if , it moves one step nearer. The gallery distance from to is the Coxeter length , and adjacent chamber distances differ by one. Consequently
Write for the Inversion set of a Weyl-group element. Part b shows that permutes . It follows that right multiplication by changes the size of the inversion set byIndeed, all roots other than are merely relabelled, while . Part c gives exactly the same recursion for the Coxeter length. Both quantities vanish at the identity, so induction along any word in the simple reflections gives
By the assumed transitivity on fundamental systems, some sends to . It therefore sends the entire positive system of a root system to . Part d then givesFor every , its inversion set is contained in , so and has maximal length.
If also has maximal length, then , so . Hence preserves and has no inversions. Part d makes its Coxeter length zero, so it is the identity. Thus , proving that the Longest element of a finite Coxeter group is unique and has length .
Let be the real vector space with basis . The Coxeter Gram matrix defines the symmetric bilinear formIts diagonal entries are . The Geometric representation of a Coxeter group is generated by the reflectionsEach has square one, and on the product has order . The reflections therefore satisfy the Coxeter relations and define a group representation .
The Coxeter graph has vertex set . Distinct vertices are joined precisely when , and the edge is labelled when ; the customary unlabelled edge therefore means . The Coxeter system is an Irreducible Coxeter system precisely when this graph is connected.
PutAfter ordering before , the only nonzero off-diagonal entries between the two blocks occur at and , where they equal . Thuswhere the coordinate vectors in the two blocks are understood. Expanding the determinant according to whether neither or both cross-block entries are selected givesThe minus sign is the sign of the transposition pairing the two cross-block entries. This formula remains valid when either diagonal block is singular, so no inverse or Schur complement is needed.
All edges in the Type E(p,q,k) Coxeter graph are unlabelled, so the factor in part c is one. Separate the arm of length from the central vertex. The remaining two arms form a type chain, while deleting the central vertex leaves the disjoint type and type chains. Using in the formula from part c yieldsThe associated bilinear form is degenerate exactly whenThe positive-integer solutions of , up to permutation, are , , and . Consequently the degenerate arm-length triples are the permutations ofFor every other allowed the determinant is nonzero, so the form is nondegenerate.
The Generic Hecke algebra of a Coxeter system is the free module with basis over the polynomial ring in parameters , subject to whenever and are conjugate, and with multiplicationEquivalently, its generators satisfy the Coxeter braid relations andIn type , all simple generators are conjugate, so there is one parameter.
A BN-pair consists of subgroups for which , the subgroup is normal in , the quotient is generated by a distinguished set of involutions, and the Bruhat multiplication and nondegeneracy axioms hold. The quotient is the associated Weyl group, and the axioms give the Bruhat decomposition of a BN-pair
The Iwahori-Hecke algebra of a BN-pair may be defined, up to the usual opposite-algebra convention, byIts standard basis is indexed by the Bruhat double cosets. For a simple generator represented by , setThe double-coset multiplication rule is the generic rule from part a with specialized to . Thus is a specialization of the generic algebra.
Every Hecke parameter of a BN-pair divides and, because is a -group, is a power of . If , then in the field of characteristic . The specialized quadratic relation becomeswhile the braid relations are unchanged. These are the defining relations of the Coxeter group , so induces a surjective homomorphismBoth algebras have bases indexed by , hence the homomorphism is an isomorphism of group algebras.
There is a missing hypothesis in the printed claim: it is false when every irreducible component of has type . The intended statement holds as soon as has an irreducible component of rank at least two, which we now assume.
Since , every Hecke parameter of a BN-pair vanishes in , and is the 0-Hecke algebra withLet be the Longest element of a finite Coxeter group. Choose a simple generator in a component of rank at least two, putThe element is again a simple generator. The identities and giveIf is simple, then is a left descent of both and : using , one gets . HenceThe one-dimensional subspace is therefore a left ideal. It is nonzero because and are distinct basis elements.
Every reduced expression in a Coxeter group for contains : in an irreducible finite component of rank at least two, deleting one final generator from does not remove any vertex from its support. A reduced expression for contains as well. Since , associativity now givesIf were a semisimple algebra, the left ideal would be a direct summand of the regular module. The corresponding projection would produce a nonzero idempotent in , impossible because . Thus is not semisimple.
For completeness, if , thenwhich is semisimple. This is the counterexample showing why the omitted rank condition is necessary.
The Coxeter number is the order of a Coxeter element. In the stated families the values areHere is the dihedral group of order , is the symmetric group, is the signed symmetric group, and is the even signed symmetric group.
SetIn the Geometric representation of a Coxeter group,including , since . As , telescoping givesSumming from to yieldsor equivalentlySumming the same telescoping identity all the way to and substituting this first formula giveswhich is the second required identity.
Let be the matrix whose th column consists of the coordinates of in the basis . Define the upper-triangular matrix and lower-triangular matrix byThe first identity in part b says , so . The second says that the matrix of is . Since has diagonal entries one, , and thereforeThus is the characteristic polynomial of the Coxeter element in its geometric representation.
At , the diagonal entries of are , while every off-diagonal entry isHence is exactly the Coxeter Gram matrix , and part c givesFor a Finite Coxeter group the Gram matrix is positive definite, and for a Hyperbolic Coxeter group it is nondegenerate with Lorentzian signature. In either case , so is not an eigenvalue of and the Coxeter element fixes no nonzero vector.
For an Affine Coxeter group, the Gram form has a nonzero radical. If , then for every , and every generating reflection satisfiesTheir product therefore fixes . Thus every affine Coxeter element has a nonzero fixed vector.
Articles by others on the same topic
There are currently no matching articles.