Definition Reflection block coordinates and evaluation matrices
✓ definition formalized

Reflection block coordinates and evaluation matrices

Let \(A\) be a commutative ring, \(\tau:A\to A\) a ring automorphism, \(\alpha\in A\), and \(J\) a finite index set. For a matrix \(E=(E_{jk})_{j,k\in J}\), write \[ (\tau E)_{jk}=\tau(E_{jk}),\qquad (Ev)_j=\sum_{k\in J}E_{jk}v_k. \] The automorphism acts on the matrix entries in \(\tau E\).

For a pair of vectors \(v_0,v_1\in A^J\), define the reflection wall-coordinate map \[ \mathcal W_E(v_0,v_1) =\bigl(E(v_0+\alpha v_1),\,(\tau E)(v_0-\alpha v_1)\bigr). \] Its block matrix is \[ \begin{pmatrix} E&\alpha E\\ \tau E&-\alpha\,\tau E \end{pmatrix}. \] Multiplying this matrix by \((v_0,v_1)\) gives the two displayed coordinate vectors by distributivity of the finite sums. The source names the map wallCoordinate and the matrix wallMatrix.

The evaluation matrix of a word

For a finite-generator Coxeter system, now take \(A\) to be the symmetric algebra of its real extended realization. For its tensor word module, let \((b_\gamma)_{\gamma\in J_{\mathbf t}}\) be its standard word basis. Define \[ (E_{\mathbf t})_{\eta\gamma} =E_{\mathbf t,\eta}(b_\gamma). \] Rows record evaluations; columns record basis vectors. If \(m=\sum_\gamma v_\gamma b_\gamma\), linearity gives \[ E_{\mathbf t,\eta}(m) =\sum_\gamma (E_{\mathbf t})_{\eta\gamma}v_\gamma. \] Thus full evaluation is multiplication by this matrix after taking basis coordinates.

Prepending one reflection

Let \(s\) act by \(\tau\), with root polynomial \(\alpha_s\). The new basis vectors are \[ b_{(0,\gamma)}=1\otimes b_\gamma,\qquad b_{(1,\gamma)}=1\otimes\alpha_s b_\gamma. \] The evaluation recursion gives all four blocks: \[ \begin{array}{c|cc} &b_{(0,\gamma)}&b_{(1,\gamma)}\\ (0,\eta)&(E_{\mathbf t})_{\eta\gamma} &\alpha_s(E_{\mathbf t})_{\eta\gamma}\\ (1,\eta)&\tau((E_{\mathbf t})_{\eta\gamma}) &-\alpha_s\tau((E_{\mathbf t})_{\eta\gamma}) \end{array} \] The last entry uses \(\tau(\alpha_s)=-\alpha_s\). Therefore the new evaluation matrix is exactly the reflection block matrix for \(E_{\mathbf t}\).

The empty-word matrix is \((1)\). For one letter this construction gives \[ E_s=\begin{pmatrix}1&\alpha_s\\1&-\alpha_s\end{pmatrix}. \]

References

The local source declarations below give the construction and proof.

Lean formalization

OAI.KLInvariance.ReflectionFrobenius.wallCoordinate: Frobenius.lean:1137.

OAI.KLInvariance.ReflectionFrobenius.wallMatrix: Frobenius.lean:1141.

OAI.KLInvariance.ReflectionFrobenius.wallMatrix_mulVec: Frobenius.lean:1146.

OAI.KLInvariance.TitsSpace.bottEvaluationMatrix: Frobenius.lean:1213.

OAI.KLInvariance.TitsSpace.bottEvaluation_expand: Frobenius.lean:1216.

OAI.KLInvariance.TitsSpace.bottEvaluationMatrix_cons: Frobenius.lean:1225.

OAI.KLInvariance.TitsSpace.bottEvaluationMatrix_mulVec_cons: Frobenius.lean:1247.

Used by