Rigidity of the non-endpoint theta submodularity families #
The theta branch of Theorem 4.13 needs the canonical correction in the genus-two inversion formula to vanish. Here that is checked directly from the coordinate families of Corollary 3.6.
theorem
Bananas.leftEndpoint_add_not_linearEquiv_canonical
(B : Banana 2)
(w : (Utilities.Certificate.SubdivisionGraph.Spec.graph B).V)
(hw : w ≠ rightEndpoint B)
:
Adding a vertex other than the right endpoint to the left endpoint is not canonical on a theta.
theorem
Bananas.rightEndpoint_add_not_linearEquiv_canonical
(B : Banana 2)
(w : (Utilities.Certificate.SubdivisionGraph.Spec.graph B).V)
(hw : w ≠ leftEndpoint B)
:
Symmetric endpoint version of leftEndpoint_add_not_linearEquiv_canonical.
theorem
Bananas.distinctInterior_strand_pair_not_linearEquiv_canonical
(B : Banana 2)
(alpha beta : Fin 3)
(i : Utilities.Certificate.SubdivisionGraph.Spec.PathPosition B alpha)
(j : Utilities.Certificate.SubdivisionGraph.Spec.PathPosition B beta)
(hi : Utilities.Certificate.SubdivisionGraph.Spec.IsInteriorPosition B alpha i)
(hj : Utilities.Certificate.SubdivisionGraph.Spec.IsInteriorPosition B beta j)
(hab : alpha ≠ beta)
:
¬linearEquiv (Utilities.Certificate.SubdivisionGraph.Spec.graph B)
(oneChip (strandVertex B alpha i) + oneChip (strandVertex B beta j))
(canonicalDivisor (Utilities.Certificate.SubdivisionGraph.Spec.graph B))
Two interior chips on distinct theta strands have rank zero, so cannot be canonical.