Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Canonical affine compression elements and their recursions

Open
QAlgorithms.Papers.Independent.Intakenext100.P34.theorem3_6

by Gohan · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let CCC be a commutative unital ring, with units q,z,Q=q−q−1q,z,Q=q-q^{-1}q,z,Q=q−q−1 and scalars Z0(ℓ)∈CZ_0^{(\ell)}\in CZ0(ℓ)​∈C for every integer ℓ\ellℓ. Assume Z0(0)=1+(z−z−1)/QZ_0^{(0)}=1+(z-z^{-1})/QZ0(0)​=1+(z−z−1)/Q. Write WkW_kWk​ for the quotient of the affine braid group algebra by (2.42)–(2.43), with Y1=zXϵ1Y_1=zX^{\epsilon_1}Y1​=zXϵ1​, Yi+1=TiYiTiY_{i+1}=T_iY_iT_iYi+1​=Ti​Yi​Ti​, and Ei=1−(Ti−Ti−1)/QE_i=1-(T_i-T_i^{-1})/QEi​=1−(Ti​−Ti−1​)/Q. Both left and right versions of its defining relations are imposed.

There is a compatible family of central Laurent-polynomial representatives Zn(ℓ)∈WnZ_n^{(\ell)}\in W_nZn(ℓ)​∈Wn​ (n≥0n\ge0n≥0, ℓ∈Z\ell\in\mathbb Zℓ∈Z), starting at the supplied scalars, with

EiYiℓEi=Zi−1(ℓ)Ei.E_iY_i^\ell E_i=Z_{i-1}^{(\ell)}E_i.Ei​Yiℓ​Ei​=Zi−1(ℓ)​Ei​.

Define Zn±(u)=∑ℓ≥0Zn(±ℓ)u−ℓZ_n^\pm(u)=\sum_{\ell\ge0}Z_n^{(\pm\ell)}u^{-\ell}Zn±​(u)=∑ℓ≥0​Zn(±ℓ)​u−ℓ, so both have zeroth coefficient Zn(0)Z_n^{(0)}Zn(0)​. Put H=u2/(u2−1)H=u^2/(u^2-1)H=u2/(u2−1) and

R(Y,u)=(u−Y)2(u−q−2Y−1)(u−q2Y−1)(u−Y−1)2(u−q2Y)(u−q−2Y).R(Y,u)=\frac{(u-Y)^2(u-q^{-2}Y^{-1})(u-q^2Y^{-1})}{(u-Y^{-1})^2(u-q^2Y)(u-q^{-2}Y)}.R(Y,u)=(u−Y−1)2(u−q2Y)(u−q−2Y)(u−Y)2(u−q−2Y−1)(u−q2Y−1)​.

For i≥1i\ge1i≥1, the identity in level i+1i+1i+1 is

(Zi−1−−z/Q−H)(Zi−1++z−1/Q−H)Ei=−(u2−q2)(u2−q−2)Q2(u2−1)2Ei.(Z_{i-1}^- -z/Q-H)(Z_{i-1}^+ +z^{-1}/Q-H)E_i =-\frac{(u^2-q^2)(u^2-q^{-2})}{Q^2(u^2-1)^2}E_i.(Zi−1−​−z/Q−H)(Zi−1+​+z−1/Q−H)Ei​=−Q2(u2−1)2(u2−q2)(u2−q−2)​Ei​.

In level i+2i+2i+2,

(Zi++z−1/Q−H)Ei+1=(Zi−1++z−1/Q−H)R(Yi,u)Ei+1.(Z_i^+ +z^{-1}/Q-H)E_{i+1}=(Z_{i-1}^+ +z^{-1}/Q-H)R(Y_i,u)E_{i+1}.(Zi+​+z−1/Q−H)Ei+1​=(Zi−1+​+z−1/Q−H)R(Yi​,u)Ei+1​.

For k≥1k\ge1k≥1, in level k+1k+1k+1,

(Zk−1++z−1/Q−H)Ek=(Z0++z−1/Q−H)∏j=1k−1R(Yj,u)Ek.(Z_{k-1}^+ +z^{-1}/Q-H)E_k=(Z_0^+ +z^{-1}/Q-H)\prod_{j=1}^{k-1}R(Y_j,u)E_k.(Zk−1+​+z−1/Q−H)Ek​=(Z0+​+z−1/Q−H)j=1∏k−1​R(Yj​,u)Ek​.

The positive representatives are constructed coefficientwise from this product. At positive levels, the negative representatives are obtained from the first identity using the unit constant z/Qz/Qz/Q of the positive factor; at level zero the prescribed bilateral scalars are retained. Compression remains a conclusion and is never cancelled.

This supplies both recursions and the bilateral moment interface without assuming admissibility or tower injectivity.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P34
Formal statement
namespace QAlgorithms.Papers.Independent.Intakenext100.P34

open QAlgorithms.Papers.Independent.Support21.P08 (Central)

noncomputable section

universe u

variable (C : Type u) [CommRing C]

/-- Theorem 3.6, PDF pp.16--18: centrality and compression for the
constructed representatives, followed by (3.26)--(3.28). Both positive and
negative moments have the same zeroth term. The squared denominator in
(3.26) is retained, and both stale E_(i+1) in (3.28) are corrected to E_k
as forced by its induction. No admissibility hypothesis is imposed. -/
theorem theorem3_6 (P : AParams C) :
    (∀ (n : ℕ) (ell : ℤ), 0 < n → Central (aMomentAt C P n n ell (le_refl n))) ∧
    (∀ (n : ℕ) (ell : ℤ),
      let j : Adj (n+2) := ⟨n, by omega⟩
      let m := awMap C P (n+2)
      m (affineContractionWord C P j)*signedPow (m (affineYWord C P (adjacentLeft j) false))
        (m (affineYWord C P (adjacentLeft j) true)) ell*m (affineContractionWord C P j) =
        aMomentAt C P n (n+2) ell (by omega)*m (affineContractionWord C P j)) ∧
    (∀ n : ℕ,
      let j : Adj (n+2) := ⟨n, by omega⟩
      let c := algebraMap C (AW C P (n+2))
      let e := ps (awMap C P (n+2) (affineContractionWord C P j))
      let h := 1+pairResolvent (1 : AW C P (n+2))
      (aMomentSeries C P n (n+2) (by omega) true-ps (c ((P.z : C)*(↑P.Q⁻¹ : C)))-h)*
        (aMomentSeries C P n (n+2) (by omega) false+ps (c ((↑P.z⁻¹ : C)*(↑P.Q⁻¹ : C)))-h)*e =
        -(ps (c ((↑P.Q⁻¹ : C)^2))*aScalarFraction (c (P.q : C)) (c (↑P.q⁻¹ : C)))*e) ∧
    (∀ n : ℕ,
      let j : Adj (n+3) := ⟨n+1, by omega⟩
      let y : Fin (n+3) := ⟨n, by omega⟩
      let m := awMap C P (n+3)
      let c := algebraMap C (AW C P (n+3))
      let e := ps (m (affineContractionWord C P j))
      let offset := ps (c ((↑P.z⁻¹ : C)*(↑P.Q⁻¹ : C)))-(1+pairResolvent (1 : AW C P (n+3)))
      (aMomentSeries C P (n+1) (n+3) (by omega)+offset)*e =
        (aMomentSeries C P n (n+3) (by omega)+offset)*
        aRatio (c (P.q : C)) (c (↑P.q⁻¹ : C)) (m (affineYWord C P y false)) (m (affineYWord C P y true))*e) ∧
    (∀ (k : ℕ) (hk : 0 < k),
      let j : Adj (k+1) := ⟨k-1, by omega⟩
      let m := awMap C P (k+1)
      let c := algebraMap C (AW C P (k+1))
      let e := ps (m (affineContractionWord C P j))
      let offset := ps (c ((↑P.z⁻¹ : C)*(↑P.Q⁻¹ : C)))-(1+pairResolvent (1 : AW C P (k+1)))
      (aMomentSeries C P (k-1) (k+1) (by omega)+offset)*e =
        (aMomentSeries C P 0 (k+1) (by omega)+offset)*
        (List.ofFn (fun a : Fin (k-1) =>
          let y : Fin (k+1) := ⟨a.val, by have := a.isLt; omega⟩
          aRatio (c (P.q : C)) (c (↑P.q⁻¹ : C)) (m (affineYWord C P y false)) (m (affineYWord C P y true)))).prod*e) := by sorry

end

end QAlgorithms.Papers.Independent.Intakenext100.P34
Source
Affine and degenerate affine BMW algebras: The center, arXiv:1105.4207v2, PDF p. 17, Theorem 3.6
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

Let CCC be a commutative ring, and let PPP consist of units q,z,Q∈C×q,z,Q\in C^\timesq,z,Q∈C× and moments ω:Z→C\omega:\mathbb Z\to Cω:Z→C, satisfying

Q=q−q−1,ω0=1+(z−z−1)Q−1.Q=q-q^{-1},\qquad \omega_0=1+(z-z^{-1})Q^{-1}.Q=q−q−1,ω0​=1+(z−z−1)Q−1.

For every natural number NNN, let WNW_NWN​ be the associative CCC-algebra generated by invertible T0,…,TN−2,XT_0,\ldots,T_{N-2},XT0​,…,TN−2​,X, and define

Er=1−Q−1(Tr−Tr−1),Yj=zTj−1⋯T0XT0⋯Tj−1,Y‾j=z−1Tj−1−1⋯T0−1X−1T0−1⋯Tj−1−1.E_r=1-Q^{-1}(T_r-T_r^{-1}),\quad Y_j=zT_{j-1}\cdots T_0XT_0\cdots T_{j-1},\quad \overline Y_j=z^{-1}T_{j-1}^{-1}\cdots T_0^{-1}X^{-1}T_0^{-1}\cdots T_{j-1}^{-1}.Er​=1−Q−1(Tr​−Tr−1​),Yj​=zTj−1​⋯T0​XT0​⋯Tj−1​,Yj​=z−1Tj−1−1​⋯T0−1​X−1T0−1​⋯Tj−1−1​.

For an integer mmm, write Yj[m]=YjmY_j^{[m]}=Y_j^mYj[m]​=Yjm​ if m≥0m\ge0m≥0, and Yj[m]=Y‾j−mY_j^{[m]}=\overline Y_j^{-m}Yj[m]​=Yj−m​ if m<0m<0m<0. The quotient relations, for all valid indices, are

TrTs=TsTr (∣r−s∣≠1),TrTr+1Tr=Tr+1TrTr+1,XT0XT0=T0XT0X,XTr=TrX (r>0),ErTrδ=TrδEr=z−δEr,ErTsδEr=zδEr (∣r−s∣=1, δ=±1),E0Y0[m]E0=ωmE0 (m∈Z),ErYrYr+1=Er=YrYr+1Er.\begin{gathered} T_rT_s=T_sT_r\ (|r-s|\ne1),\qquad T_rT_{r+1}T_r=T_{r+1}T_rT_{r+1},\\ XT_0XT_0=T_0XT_0X,\qquad XT_r=T_rX\ (r>0),\\ E_rT_r^\delta=T_r^\delta E_r=z^{-\delta}E_r,\qquad E_rT_s^\delta E_r=z^\delta E_r \ (|r-s|=1,\ \delta=\pm1),\\ E_0Y_0^{[m]}E_0=\omega_mE_0\ (m\in\mathbb Z),\qquad E_rY_rY_{r+1}=E_r=Y_rY_{r+1}E_r. \end{gathered}Tr​Ts​=Ts​Tr​ (∣r−s∣=1),Tr​Tr+1​Tr​=Tr+1​Tr​Tr+1​,XT0​XT0​=T0​XT0​X,XTr​=Tr​X (r>0),Er​Trδ​=Trδ​Er​=z−δEr​,Er​Tsδ​Er​=zδEr​ (∣r−s∣=1, δ=±1),E0​Y0[m]​E0​=ωm​E0​ (m∈Z),Er​Yr​Yr+1​=Er​=Yr​Yr+1​Er​.​

The following notation specifies the moment expressions used in the assertions. In any coefficient algebra, regard scalars as their images and algebra elements as constant series, and put

G(u)=∑r≥1u2r,H(u)=1+G(u),F(u)=1+(2−q2−q−2)(G+G2),O(u)=z−1Q−1−H(u).G(u)=\sum_{r\ge1}u^{2r},\qquad H(u)=1+G(u),\qquad F(u)=1+(2-q^2-q^{-2})(G+G^2),\qquad O(u)=z^{-1}Q^{-1}-H(u).G(u)=r≥1∑​u2r,H(u)=1+G(u),F(u)=1+(2−q2−q−2)(G+G2),O(u)=z−1Q−1−H(u).

For elements a,aˉa,\bar aa,aˉ define the ordered formal series

R(a,aˉ)=(1−au)2(1−q−2aˉ u)(1−q2aˉ u)((1−aˉ u)2(1−q2au)(1−q−2au))−1,R(a,\bar a)= (1-au)^2(1-q^{-2}\bar a\,u)(1-q^2\bar a\,u) \left((1-\bar a\,u)^2(1-q^2au)(1-q^{-2}au)\right)^{-1},R(a,aˉ)=(1−au)2(1−q−2aˉu)(1−q2aˉu)((1−aˉu)2(1−q2au)(1−q−2au))−1,

whose denominator has constant coefficient 111. In the Laurent polynomial algebra C[x0±1,…,xn−1±1]C[x_0^{\pm1},\ldots,x_{n-1}^{\pm1}]C[x0±1​,…,xn−1±1​], consisting of finite sums of integer-exponent monomials, define

Sn(u)=(∑r≥0ωrur+O(u))∏j=0n−1R(xj,xj−1),μn,m={ωm,n=0,[um](Sn−z−1Q−1+H),n>0, m≥0,[u−m](zQ−1+H−Q−2FSn−1),n>0, m<0.\begin{gathered} S_n(u)= \left(\sum_{r\ge0}\omega_ru^r+O(u)\right) \prod_{j=0}^{n-1}R(x_j,x_j^{-1}),\\ \mu_{n,m}= \begin{cases} \omega_m,&n=0,\\ [u^m](S_n-z^{-1}Q^{-1}+H),&n>0,\ m\ge0,\\ [u^{-m}](zQ^{-1}+H-Q^{-2}FS_n^{-1}),&n>0,\ m<0. \end{cases} \end{gathered}Sn​(u)=(r≥0∑​ωr​ur+O(u))j=0∏n−1​R(xj​,xj−1​),μn,m​=⎩⎨⎧​ωm​,[um](Sn​−z−1Q−1+H),[u−m](zQ−1+H−Q−2FSn−1​),​n=0,n>0, m≥0,n>0, m<0.​​

The inverse of SnS_nSn​ uses its unit constant coefficient zQ−1zQ^{-1}zQ−1. For n≤Nn\le Nn≤N, if μn,m=∑dcdxd\mu_{n,m}=\sum_dc_dx^dμn,m​=∑d​cd​xd, put

Mn,mN=∑d: cd≠0ιN(cd)Y0[d0]⋯Yn−1[dn−1],ZnN,+(u)=∑ℓ≥0Mn,ℓNuℓ,ZnN,−(u)=∑ℓ≥0Mn,−ℓNuℓ,M^N_{n,m}= \sum_{d:\,c_d\ne0}\iota_N(c_d) Y_0^{[d_0]}\cdots Y_{n-1}^{[d_{n-1}]}, \qquad Z_n^{N,+}(u)=\sum_{\ell\ge0}M^N_{n,\ell}u^\ell,\qquad Z_n^{N,-}(u)=\sum_{\ell\ge0}M^N_{n,-\ell}u^\ell,Mn,mN​=d:cd​=0∑​ιN​(cd​)Y0[d0​]​⋯Yn−1[dn−1​]​,ZnN,+​(u)=ℓ≥0∑​Mn,ℓN​uℓ,ZnN,−​(u)=ℓ≥0∑​Mn,−ℓN​uℓ,

where ιN:C→WN\iota_N:C\to W_NιN​:C→WN​ is the scalar map and products are ordered by increasing index. The theorem asserts the conjunction of the following five universally quantified statements: for every n>0n>0n>0, every m∈Zm\in\mathbb Zm∈Z, and every a∈Wna\in W_na∈Wn​,

Mn,mna=aMn,mn;M^n_{n,m}a=aM^n_{n,m};Mn,mn​a=aMn,mn​;

for every n≥0n\ge0n≥0 and m∈Zm\in\mathbb Zm∈Z, in Wn+2W_{n+2}Wn+2​,

EnYn[m]En=Mn,mn+2En;E_nY_n^{[m]}E_n=M^{n+2}_{n,m}E_n;En​Yn[m]​En​=Mn,mn+2​En​;

for every n≥0n\ge0n≥0, in Wn+2[[u]]W_{n+2}[[u]]Wn+2​[[u]],

(Znn+2,−−zQ−1−H)(Znn+2,++z−1Q−1−H)En=−Q−2FEn;\bigl(Z_n^{n+2,-}-zQ^{-1}-H\bigr) \bigl(Z_n^{n+2,+}+z^{-1}Q^{-1}-H\bigr)E_n =-Q^{-2}FE_n;(Znn+2,−​−zQ−1−H)(Znn+2,+​+z−1Q−1−H)En​=−Q−2FEn​;

for every n≥0n\ge0n≥0, in Wn+3[[u]]W_{n+3}[[u]]Wn+3​[[u]],

(Zn+1n+3,++O)En+1=(Znn+3,++O)R(Yn,Y‾n)En+1;\bigl(Z_{n+1}^{n+3,+}+O\bigr)E_{n+1} = \bigl(Z_n^{n+3,+}+O\bigr) R(Y_n,\overline Y_n)E_{n+1};(Zn+1n+3,+​+O)En+1​=(Znn+3,+​+O)R(Yn​,Yn​)En+1​;

and for every k>0k>0k>0, in Wk+1[[u]]W_{k+1}[[u]]Wk+1​[[u]],

(Zk−1k+1,++O)Ek−1=(Z0k+1,++O)(∏a=0k−2R(Ya,Y‾a))Ek−1.\bigl(Z_{k-1}^{k+1,+}+O\bigr)E_{k-1} = \bigl(Z_0^{k+1,+}+O\bigr) \left(\prod_{a=0}^{k-2}R(Y_a,\overline Y_a)\right)E_{k-1}.(Zk−1k+1,+​+O)Ek−1​=(Z0k+1,+​+O)(a=0∏k−2​R(Ya​,Ya​))Ek−1​.

All products retain the displayed order. The second, third, and fourth assertions include n=0n=0n=0; here M0,mN=ιN(ωm)M^N_{0,m}=\iota_N(\omega_m)M0,mN​=ιN​(ωm​). The final assertion includes k=1k=1k=1, when its product is empty and equals 111.

Human review
  • Endorsed by Shuze Chen · Oct 31, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 31, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me