Finite bicolored strand routing with explicit rotation and exit data.
A rotation system is not by itself a planar disk embedding. Same-color edges are allowed. Termination, the bijective boundary map, fixed-point decorations, and any later identification with positroid cells must not be inferred merely from the existence of a finite graph.
Internal vertex colors for the prescribed strand-turning convention.
Instances For
A permutation with a color supplied exactly where a fixed point needs decoration.
- perm : Equiv.Perm (Fin boundaryCount)
Instances For
Identity boundary routing with every fixed point decorated white.
Instances For
Distinct boundary and internal vertex identifiers; each internal identifier has one color.
Instances For
A finite graph with one incident edge per boundary vertex and cyclic internal rotations.
rotation is interpreted as counterclockwise; black uses it and white uses its
inverse. singleOrbit rules out a purported cyclic order consisting of several
cycles. A genus-zero disk embedding and reducedness are not supplied here.
- graph : SimpleGraph (Vertex boundaryCount internalCount)
- color : Fin internalCount → NodeColor
- boundaryNeighbor : Fin boundaryCount → Vertex boundaryCount internalCount
- boundaryAdjacent (index : Fin boundaryCount) : self.graph.Adj (Sum.inl index) (self.boundaryNeighbor index)
- boundaryUnique (index : Fin boundaryCount) (vertex : Vertex boundaryCount internalCount) : self.graph.Adj (Sum.inl index) vertex → vertex = self.boundaryNeighbor index
Instances For
A directed edge carries its adjacency proof throughout routing.
Instances For
Directed edges are equivalent to a vertex paired with one of its outgoing neighbors.
Instances For
Finite vertex and neighbor sets give a finite directed-edge carrier without extra graph data.
Directed edges agree once their source and destination agree; adjacency proofs are irrelevant.
Reverse each directed edge; graph symmetry makes this an involutive permutation.
Instances For
Neighbor turns use the chosen color internally and the unique boundary edge externally.
Instances For
Continue strands through boundaries to obtain a derived permutation on all directed edges.
Instances For
Each continuing turn leaves from the previous directed edge's destination.
Every dart returns in a positive number of turns bounded by the finite dart count.
The unique outgoing edge at a boundary vertex initializes its strand.
Instances For
One routing step: black takes the next neighbor, white the previous, and boundary stops.
Instances For
Routing at an exit boundary leaves the directed edge unchanged.
At an internal destination the continuing dart turn agrees with the original strand step.
A boundary arrival continues through its unique incident edge to the corresponding start.
Evaluate a strand for a bounded number of steps, retaining a valid directed edge.
Instances For
Once a strand reaches a boundary, every further bounded evaluation preserves the exit.
Moving the first step outside a bounded route leaves the same strand endpoint.
Splitting a route into two consecutive bounded evaluations preserves its endpoint.
Search for the first boundary exit, returning its step count or an explicit fuel failure.
Instances For
A successful search reaches the reported boundary within fuel, with no earlier boundary exit.
Any boundary reached by the fuel-limited route is found by the first-exit search.
Fuel exhaustion is equivalent to reaching no boundary by the bounded route endpoint.
Before any boundary arrival, the absorbing route agrees with continuing dart iteration.
An exhausted search budget also excludes all boundary arrivals at earlier steps.
Every boundary start has an exit found using at most the finite dart count as search fuel.
Extract the guaranteed first exit from the computed dart-count search, not a supplied match.
Instances For
The derived boundary data are exactly the successful finite search result.
Derived boundary data retain the dart-count bound, route endpoint and earliest-exit condition.
One turn after the earliest boundary arrival reaches the derived destination's start edge.
No positive iterate reaches any boundary start before the derived first-return time.
The positive boundary first-return time is bounded by the finite dart count.
Invertible dart dynamics and first-return minimality prevent two starts sharing an exit.
Construct the boundary permutation from guaranteed earliest exits, without a supplied permutation.
Instances For
The constructed boundary permutation is exactly the label returned by the first-exit search.
Any two successful fuel budgets return the same earliest step count and boundary label.
A checked finite exit certificate and bijective boundary labeling.
The permutation and finite step counts are supplied and checked against the route evaluator, not derived from unproved termination or planarity premises.
- perm : Equiv.Perm (Fin boundaryCount)
- steps : Fin boundaryCount → ℕ
Instances For
Extract a decorated permutation only from checked boundary routing data.
Instances For
The extracted permutation agrees with each finite routed endpoint.
Bijectivity is available only through the checked routing certificate's permutation.
Populate the compatibility certificate from derived exits; fixed-point decorations remain explicit.
Instances For
Nonnegative edge weights with a checked strict level increase on every nonzero edge.
- level : Fin vertexCount → Fin (height + 1)
- weight : Matrix (Fin vertexCount) (Fin vertexCount) NNReal
Instances For
Matrix-power path weights vanish when their length exceeds the available level difference.
The checked level bound derives nilpotency; no cyclic-series convergence is assumed.
Sum all weighted path lengths allowed by the certified acyclic height, including length zero.
Instances For
Increasing the finite path cutoff adds only zero terms, rather than a convergence assumption.
The finite path sum satisfies the exact identity-plus-one-edge continuation equation.
Boundary measurements select source/sink entries of the derived finite weighted path sum.
Instances For
A single weighted directed edge with source zero and sink one.
Instances For
A subdivided directed edge with an intermediate vertex and two independently supplied weights.
Instances For
A single edge's actual finite path sum recovers its supplied edge weight.
A two-edge path's actual finite sum multiplies its two edge weights.
Subdivision by a nonzero factor preserves the complete one-source/one-sink boundary matrix.
One source feeding two sinks with independent nonnegative edge weights.
Instances For
The fork's two boundary entries are derived from its finite weighted-path sum.
Two sources at level zero feed two sinks at level one with independent edge weights.
Instances For
The complete two-channel boundary matrix is derived from the finite weighted path sum.
Two sources feed two sinks through one shared intermediate vertex.
Instances For
Each measured hub entry is the product of its independently supplied incoming and outgoing edge.
One source at level zero feeds two parallel intermediate vertices at level one, which both feed one sink at level two.
Instances For
The diamond network's source-to-sink transfer sums the two parallel two-edge path products.
Reducing two parallel two-edge paths into a single edge with summed path-product weight preserves the complete one-source/one-sink boundary measurement matrix.
Multi-Layer Lindström–Gessel–Viennot (LGV) Disjoint-Path Composition & Three-Channel ($k = 3$) LGV Positivity #
The determinant of a finite list of composable $k \times k$ layer transfer matrices $L = [M_1, M_2, \dots, M_m]$ equals the product of the individual layer determinants: $\det\bigl(\prod_{r=1}^m M_r\bigr) = \prod_{r=1}^m \det(M_r)$.
Multi-Layer Lindström–Gessel–Viennot (LGV) Total Nonnegativity Preservation: If every layer transfer matrix $M_r \in L$ in an arbitrary-length planar acyclic cascade has nonnegative $k$-channel determinant $\det(M_r) \ge 0$, then the composite multi-step transfer matrix $T = \prod_{r=1}^m M_r$ has $\det(T) \ge 0$.
Multi-Layer Lindström–Gessel–Viennot (LGV) Strict Disjoint-Path Positivity: If every layer transfer matrix $M_r \in L$ in an arbitrary-length planar acyclic cascade admits a strictly positive $k$-tuple of vertex-disjoint paths ($\det(M_r) > 0$), then the composite multi-step transfer matrix $T = \prod_{r=1}^m M_r$ has strictly positive determinant $\det(T) > 0$.
Planar upper-bidiagonal elementary $3 \times 3$ step matrix $E_{\mathrm{upper}}^{(3)}(d_0, d_1, d_2, u_{01}, u_{12})$ with horizontal channel weights $d_0, d_1, d_2$ and non-crossing upward steps $0 \to 1$ ($u_{01}$) and $1 \to 2$ ($u_{12}$).
Instances For
Planar lower-bidiagonal elementary $3 \times 3$ step matrix $E_{\mathrm{lower}}^{(3)}(d_0, d_1, d_2, l_{10}, l_{21})$ with horizontal channel weights $d_0, d_1, d_2$ and non-crossing downward steps $1 \to 0$ ($l_{10}$) and $2 \to 1$ ($l_{21}$).
Instances For
Three-Channel ($k = 3$) Lindström–Gessel–Viennot Disjoint-Path Determinant & Shared-Vertex Obstruction Theorem:
- For any planar upper-bidiagonal or lower-bidiagonal 3-channel step, every intersecting path triple cancels or vanishes by planarity, leaving the unique vertex-disjoint horizontal path-triple product: $$\det E_{\mathrm{upper}}^{(3)}(d_0, d_1, d_2, u_{01}, u_{12}) = d_0 d_1 d_2, \qquad \det E_{\mathrm{lower}}^{(3)}(d_0, d_1, d_2, l_{10}, l_{21}) = d_0 d_1 d_2,$$
- For their two-step planar composite $T^{(3)} = E_{\mathrm{lower}}^{(3)}(1, 1, 1, l_{10}, l_{21}) \, E_{\mathrm{upper}}^{(3)}(d_0, d_1, d_2, u_{01}, u_{12})$, the 3-channel LGV determinant equals $\det T^{(3)} = d_0 d_1 d_2 > 0$ whenever $d_0, d_1, d_2 > 0$,
- Conversely, if three sources route through a two-vertex bottleneck (
A : Matrix (Fin 3) (Fin 2) ℝcomposed withB : Matrix (Fin 2) (Fin 3) ℝ), the pigeonhole principle forces every path triple to share an intermediate vertex and the $3 \times 3$ determinant vanishes identically: $\det(A B) = 0$.
Vector-Relation Configurations (VRC) for Complex Grassmannian $\mathrm{Gr}_{\mathbb{C}}(k, n)$ Plabic Networks #
Standard Postnikov plabic graphs (WeightedAcyclicNetwork above) assign strictly positive real
weights $x_e \in \mathbb{R}_{>0}$ (or nonnegative ℝ≥0 weights) to parameterize positroid cells
of the Totally Nonnegative Grassmannian $\mathrm{Gr}_{\ge 0}(k, n)$, where boundary-measurement
denominators and cluster-mutation denominators $x_X + 1 > 1 > 0$ never vanish.
When moving to the Complex Grassmannian $\mathrm{Gr}_{\mathbb{C}}(k, n)$—as required for off-shell virtual states, complex Kohn-Sham / QED-DFT electronic orbitals, hybrid polariton and plasmon superpositions, and quantum causal channels—rational face weights $x_e \in \mathbb{C}^*$ parameterize only an open algebraic torus chart $(\mathbb{C}^*)^{k(n-k)}$ and suffer from two fundamental obstructions:
- On-Shell Resonance / Destructive-Interference Poles ($x_X = -1$): Under the complex square move (urban renewal) $x_X' = 1 / x_X$, $x_1' = (x_1 x_X + x_2) / (x_X + 1)$, the denominator $x_X + 1$ vanishes at $x_X = -1$ (on-shell polaritonic resonance or destructive loop interference).
- Chart-Boundary Degeneration ($x_X = 0$): Rational entries such as $x_X^{-1}$ blow up on lower-rank positroid / Schubert boundary strata.
Following Lam (2015) and Affolter, Glick, Pylyavskyy, & Ramassamy (2019, arXiv:1908.06959),
a Vector-Relation Configuration (VectorRelationConfiguration) resolves both obstructions
by placing vectors $v_w \in R^k$ on the white vertices $w \in W$ (vectors : Matrix (Fin rankK) (Fin whiteCount) R)
and homogeneous linear relations $\sum_{w} K_{b,w} v_w = 0$ on the black vertices $b \in B$
(relations : Matrix (Fin blackCount) (Fin whiteCount) R), governed by the polynomial matrix
identity $V K^\top = 0$ (vectors * relations.transpose = 0).
A Vector-Relation Configuration (VRC) on a bicolored network over a commutative ring $R$:
vectors : Matrix (Fin rankK) (Fin whiteCount) Rassigns a column vector $v_w \in R^{\mathrm{rankK}}$ to each white vertex $w \in \mathrm{Fin}\,\mathrm{whiteCount}$,relations : Matrix (Fin blackCount) (Fin whiteCount) Rassigns linear relation coefficients $K_{b, w} \in R$ to each black vertex $b \in \mathrm{Fin}\,\mathrm{blackCount}$,boundaryWhite : Fin boundaryCount → Fin whiteCountidentifies the $n$ boundary white vertices,relation_satisfied : vectors * relations.transpose = 0enforces $\sum_w K_{b, w} v_w = 0$ at every black vertex $b$ without any division or non-vanishing denominator assumption.
- vectors : Matrix (Fin rankK) (Fin whiteCount) R
- relations : Matrix (Fin blackCount) (Fin whiteCount) R
- boundaryWhite : Fin boundaryCount → Fin whiteCount
Instances For
Extract the $k \times n$ boundary generator matrix $C_{\mathrm{bdry}} \in R^{k \times n}$
whose $j$-th column is the vector at boundary white vertex boundaryWhite j.
Instances For
At every black vertex $b$ and vector component $i$, the weighted sum of white-vertex vectors vanishes identically: $\sum_w V_{i, w} K_{b, w} = 0$.
Global $\mathrm{GL}_k(R)$ / Unitary Orbital Rotation Gauge Invariance of a VRC: Left-multiplying all white-vertex vectors by $g \in R^{k \times k}$ (such as a unitary orbital transformation $U \in U(N_{\mathrm{occ}})$ in Kohn-Sham DFT) preserves all black-vertex relations $V K^\top = 0$.
Instances For
Global $\mathrm{GL}_k(R)$ action on a VRC acts by left multiplication on the boundary matrix: $(g \cdot \mathrm{vrc}).\mathrm{boundaryMatrix} = g \cdot \mathrm{vrc}.\mathrm{boundaryMatrix}$.
Black-Vertex Row-Operation / Urban-Renewal Elimination Invariance of a VRC: Left-multiplying the black-vertex relation matrix $K$ by any row-mixing matrix $M \in R^{|B'| \times |B|}$ preserves $V (M K)^\top = (V K^\top) M^\top = 0$ and leaves the boundary Grassmannian generator matrix unchanged.
Instances For
Black-vertex relation row operations leave the boundary subspace matrix identically unchanged.
The $2 \times 2$ Plücker minor $p_{j_1, j_2}(C) = C_{0, j_1} C_{1, j_2} - C_{0, j_2} C_{1, j_1}$ of a $2 \times 4$ generator matrix $C \in R^{2 \times 4}$.
Instances For
The quadratic Plücker relation residual $\mathcal{R}_{\mathrm{Pl\ddot{u}cker}}(C) = p_{01} p_{23} - p_{02} p_{13} + p_{03} p_{12}$ (in 1-based indexing: $p_{12} p_{34} - p_{13} p_{24} + p_{14} p_{23}$) for $C \in R^{2 \times 4}$.
Instances For
Universal Quadratic Plücker Identity on $\mathrm{Gr}_R(2, 4)$: Every $2 \times 4$ matrix $C \in R^{2 \times 4}$ over any commutative ring $R$ (including every VRC boundary matrix over $\mathbb{C}$) identically satisfies the quadratic Plücker relation $p_{01} p_{23} - p_{02} p_{13} + p_{03} p_{12} = 0$.
The rational face-weight generator matrix $C_{\mathrm{rat}}(x_1, x_2, x_X)$ on $\mathrm{Gr}_F(2, 4)$
from the QCI-GXC / polariton plabic network (Google-Search_Density-Functional-Theory_20261105-1512.md,
lines 576–580, 735–739, 1118–1122):
$$C_{\mathrm{rat}}(x_1, x_2, x_X) = \begin{pmatrix} 1 & 0 & -x_1 & -x_X \\ 0 & 1 & x_X^{-1} & -x_2 \end{pmatrix}.$$
Instances For
The rational complex cluster mutation (square move) around the central face $x_X$
(Google-Search_Density-Functional-Theory_20261105-1512.md, lines 373–378, 761–763, 1129–1135):
$$(x_1', x_2', x_X') = \left(\frac{x_1 x_X + x_2}{x_X + 1},\; x_2,\; \frac{1}{x_X}\right).$$
Instances For
Pole-Free Homogeneous Vector-Relation Configuration (gr24HomogeneousVRC) for the $\mathrm{Gr}_R(2, 4)$ Plabic Network:
Clears the rational chart singularity $x_X^{-1}$ by scaling white vertex $w_2$ by $x_X$, yielding
white-vertex vectors
$$V_{\mathrm{hom}}(x_1, x_2, x_X) = \begin{pmatrix} 1 & 0 & -x_1 x_X & -x_X \\ 0 & 1 & 1 & -x_2 \end{pmatrix}$$
and black-vertex linear relations
$$K_{\mathrm{hom}}(x_1, x_2, x_X) = \begin{pmatrix} x_1 x_X & -1 & 1 & 0 \\ x_X & x_2 & 0 & 1 \end{pmatrix},$$
which satisfy $V_{\mathrm{hom}} K_{\mathrm{hom}}^\top = 0$ over any commutative ring $R$
(including at $x_X = 0$ and $x_X = -1$)!
Instances For
Pole-Free Mutated Vector-Relation Configuration (gr24MutatedHomogeneousVRC) Across the $x_X = -1$ On-Shell Polaritonic Resonance:
In the rational cluster mutation (gr24ClusterMutation), $x_1' = (x_1 x_X + x_2)/(x_X + 1)$ and
$x_X' = 1 / x_X$ exhibit rational poles at the on-shell polariton resonance $x_X = -1$ ($x_X + 1 = 0$)
and at $x_X = 0$. In the Vector-Relation Configuration framework, scaling white vertex $w_2$ by
$x_X + 1$ and white vertex $w_3$ by $x_X$ produces the polynomial VRC:
$$V_{\mathrm{mut}}^{\mathrm{VRC}}(x_1, x_2, x_X) = \begin{pmatrix} 1 & 0 & -(x_1 x_X + x_2) & -1 \\ 0 & 1 & x_X(x_X + 1) & -x_2 x_X \end{pmatrix},$$
$$K_{\mathrm{mut}}^{\mathrm{VRC}}(x_1, x_2, x_X) = \begin{pmatrix} x_1 x_X + x_2 & -x_X(x_X + 1) & 1 & 0 \\ 1 & x_2 x_X & 0 & 1 \end{pmatrix},$$
which satisfies $V_{\mathrm{mut}}^{\mathrm{VRC}} (K_{\mathrm{mut}}^{\mathrm{VRC}})^\top = 0$ and the
quadratic Plücker relation identically for all $x_1, x_2, x_X \in R$, remaining completely
smooth through the on-shell polariton resonance $x_X = -1$!
Instances For
Master Theorem: Complex Grassmannian Vector-Relation Configurations, Pole-Free Square-Move Resonance at $x_X = -1$, and Exact SymPy Benchmark Check Figures:
- Every $2 \times 4$ rational chart matrix $C_{\mathrm{rat}}(x_1, x_2, x_X)$ ($x_X \neq 0$) has
exact Plücker minors $p_{01} = 1$, $p_{02} = x_X^{-1}$, $p_{03} = -x_2$, $p_{12} = x_1$,
$p_{13} = x_X$, $p_{23} = x_1 x_2 + 1$, and satisfies
pluckerQuadraticRelation2x4 = 0, - Both the pre-mutation
gr24HomogeneousVRCand post-mutationgr24MutatedHomogeneousVRCsatisfy $V K^\top = 0$ andpluckerQuadraticRelation2x4 = 0for all $x_1, x_2, x_X \in \mathbb{C}$, - At the on-shell polariton resonance pole $x_X = -1$ ($x_X + 1 = 0$), where the rational
cluster mutation divides by zero,
gr24MutatedHomogeneousVRC x1 x2 (-1)remains regular with finite Plücker minors $P_{01} = 1$, $P_{02} = 0$, $P_{03} = x_2$, $P_{12} = x_2 - x_1$, $P_{13} = 1$, and $P_{23} = x_2(x_1 - x_2)$, - On the concrete off-shell complex test point from
qci_gxc_sympy.py(lines 1190–1212), $x_1 = \frac{1}{2} + \frac{1}{5}i$, $x_2 = -\frac{3}{10} + \frac{7}{10}i$, $x_X = \frac{6}{5} - \frac{2}{5}i$:- $p_{03}(C_{\mathrm{rat}}(x_1, x_2, x_X)) = -x_2 = \frac{3}{10} - \frac{7}{10}i$, and
- $(\operatorname{gr24ClusterMutation}(x_1, x_2, x_X)).2.2 = x_X^{-1} = \frac{3}{4} + \frac{1}{4}i$.