Theorem Positive reflection parameters precede negative parameters
✓ theorem proved

Positive reflection parameters precede negative parameters

Fix a plane basis \(B=(a_L,a_R)\) in a finite-generator real Coxeter system, with induced dihedral group \(D\) and geometric slope \(\sigma\). Assume intrinsic lengths in \(D\) are unbounded: \[ \forall N\in\mathbb N,\ \exists v\in D,\quad N\le\ell_D(v). \] Let \(e:x\to_Dy\) and \(f:z\to_Dt\) be actual edges with \[ yx^{-1}=w_{2k+1},\qquad tz^{-1}=w_{-(2l+1)} \qquad(k,l\in\mathbb N). \] Write \(\operatorname{Inv}_W(g,a)\) when \(g^{-1}a\) is negative, and denote the actual plane edge roots of \(e,f\) by \(\alpha_e,\alpha_f\). They satisfy \[ \sigma(\alpha_e)<\sigma(\alpha_f). \]

Relation to the OpenAI notes

Ambient inversion cuts in place of a root-coordinate recurrence.

OpenAI notes: The notes set \(\kappa=-(a_L,a_R)/2\), where the parentheses denote the ambient Tits form. The midpoint-order lemma writes prefix roots using \(p_0=0\), \(p_1=1\) and \(p_{n+1}=2\kappa p_n-p_{n-1}\). It obtains \(\kappa=\cos(\pi/m)\) in finite type and \(\kappa\ge1\) in infinite type, then compares the resulting coordinate ratios. The functional ratio gives that angular order or its reverse. dih:midpoint-order.

Lean (primary): PlaneBasis chooses its endpoint roots in increasing geometric slope. Convexity of inversion signs gives a middle-descent test, which orders the two actual ambient prefix-root sequences. A separating element \(w_{k+1}\) compares the two rays. Unbounded lengths give the signed Before order; a maximum length gives the odd residue order modulo 2m. DyerPaths:923, DyerPaths:1106, DihedralIntervals:832, DihedralIntervals:844, DihedralIntervals:1101, DihedralIntervals:1143, DihedralIntervals:1251.

Role in the proof: The ordering conclusion is developed along Lean’s inversion and length proof. The additional classification of \(\kappa\) in the notes is not a conclusion of these Lean angular-order declarations.

Proof

Step 1. Choose the element separating the two inversion signs

Take \(u=w_{k+1}\). Unbounded lengths and the signed-length theorem give \(\ell_D(w_i)=|i|\) for every integer \(i\): for a specified \(i\), choose a witness of length at least \(|i|\). In particular \(\ell_D(u)=k+1\).

For every odd \(q\), the signed reflection-action formula is \(w_qw_i=w_{q-i}\). Apply it to each edge reflection at \(u\).

Step 2. The positive-parameter root is inverted

The first reflection gives \[ w_{2k+1}u=w_{(2k+1)-(k+1)}=w_k. \] Its intrinsic length is \(k<k+1=\ell_D(u)\). The inversion-descent bridge for edge roots therefore says \[ \operatorname{Inv}_W(u,\alpha_e). \]

Step 3. The negative-parameter root is noninverted

For the other reflection, \[ w_{-(2l+1)}u=w_{-(2l+1)-(k+1)} =w_{-(2l+k+2)}. \] Its length is \(2l+k+2>k+1\). The same bridge shows that \(\operatorname{Inv}_W(u,\alpha_f)\) is false.

Step 4. The left endpoint is inverted

The positive alternating word \(u=w_{k+1}\) begins with \(s\) and is reduced. Cancelling its first letter gives a left descent at \(u\). The simple-root part of the inversion-descent bridge, with representative \(1\), gives \(\operatorname{Inv}_W(u,a_L)\).

The inversion-cut theorem applies to the two actual edge roots in \(P\): the left endpoint and \(\alpha_e\) are inverted, whereas \(\alpha_f\) is noninverted. Its left-cut inequality is exactly \(\sigma(\alpha_e)<\sigma(\alpha_f)\).

References

The local source declarations below give the construction and proof.

Lean formalization

OAI.KLInvariance.TitsSpace.PlaneBasis.edgeRoot_positive_lt_negative: DihedralIntervals.lean:1101.

Used by