RESEARCH / Focused target
OEIS A400429 second semi-meander diagonal
RELEASED FOUNDATIONS / PROPOSED CONNECTION
Research connections
All source anchors (1)
- D5/S3/Combinatorics/SemiMeanderSecondDiagonal.resultAbsent from this graph
Problem
The question
OEIS A400429 revision 16 states verbatim: "Conjecture: D_2(n) =
(n^2 + 2*n + n mod 2 - 20)/2, for n >= 4." Its diagonal is
D_j(n)=T(n,floor(n/2)-j+1), where T(n,k) counts one-loop semi-meanders
with n crossings and winding 2*(k-1)+(n mod 2). The source's equivalent
arch model fixes a lower rainbow pairing of 2*n vertices and counts
noncrossing upper perfect matchings whose union with it is one loop. The
second diagonal has exactly n-4 upper arches crossing the midpoint.
Motivation
Our foothold
The entry still labels this all-order formula a conjecture. Issue #11034 registered the external open-problem settlement before the Lean proof probe; the issue is a registration locator, not mathematical evidence.
Gap
Missing bridges
Di Francesco, Golinelli and Guitter (1996), Appendix D (D.9)-(D.14), give
a resummed small-t prediction whose second-diagonal coefficient is the same
polynomial. Their conclusion calls (D.14) a "purported re-summation". The
inspected argument does not establish all finite-n coefficient validity.
The formula is credited to that prediction; this result is an independent
proof, with no first-proof claim.
Route
Proposed approach
Index vertices from zero. A Lean UpperMatching n is the fixed-point-free
involution on Fin (2*n) for the upper arches. Its noncrossing condition
excludes a<b<M(a)<M(b). The lower rainbow is x.rev=2*n-1-x.
M.winding counts left-half vertices x<n with M(x)>=n; this counts
each midpoint-crossing upper arch once. M.oneLoop says every two vertices
are connected by the reflexive transitive closure of upper and lower pairing
steps. Therefore the theorem's nested subtype counts precisely source
objects with winding n-4 and one loop, for every natural n>=4.
The proof cuts both halves at the midpoint. Four vertices on each side
belong to non-crossing local pairs; the other n-4 upper arches cross the
midpoint. A connected union with the lower rainbow forces the local pairing
types and their compatible cross-half connections. Their count gives
(n^2+2*n+(n mod 2)-20)/2, with natural division by two. At n=4, the
formula is 2, and the source table has T(4,1)=2.
Falsifier
What would falsify this route
A noncrossing upper matching with one-loop rainbow union and n-4
midpoint-crossing arches counted differently from the source's D_2(n),
or any n>=4 whose model count differs from the polynomial, would defeat
the claimed source bridge or formula respectively. Bounded table agreement
alone does not settle the unbounded claim.
Evidence
Evidence to collect
The Lean model is in
D5/S3/Combinatorics/SemiMeanderSecondDiagonal/Model.lean; its five ordered
Stage1.lean through Stage5.lean modules establish the cut classification,
connectivity and count. The short
D5/S3/Combinatorics/SemiMeanderSecondDiagonal.lean module states the exact
all-n result. The mathematical source mapping and theorem statement were
independently approved before this publication-artifact work. Source
attribution, the Scribe interpretation, deposit, independent publication
review and merge are separate from the kernel statement.
Triage
Scope assessment
Tier 1 external named conjecture (OEIS A400429, revision 16), registered in
issue #11034. theorem; resolution proved; admission basis
open-problem-resolution; proof_shape: content. The proof classifies the
local pairings and compatible cross-half connections on the live path.
ASSUMED-UNVERIFIED
Unverified assumptions
The literature check covers OEIS A400429 revision 16 and the cited 1996
arXiv PDF, including Appendix D and its conclusion. It is not an exhaustive
search of later literature or a priority certification. The separate
declared_validated escape audit remains unfinished (issue #11180).