Stellar equivalence of simplicial complexes #
Starring a face at a fresh vertex (PreAbstractSimplicialComplex.stellarSubdivision) is one
stellar move. Two complexes are stellar equivalent when a finite sequence of stellar moves
and inverse stellar moves carries one to the other. This is the combinatorial relation that
layer 11 of the geometric-topology roadmap (TauCetiRoadmap/GeometricTopology/README.md) needs
in order to say that a complex is a combinatorial ball or sphere, namely that it becomes the
standard simplex, respectively the boundary of the standard simplex, after subdivision.
The relation is built as Relation.EqvGen of the one-move relation, so a proof of
StellarEquivalent K L is generated by stellar moves under reflexivity, symmetry, and
transitivity, and induction over that closure is how every invariance result below is proved.
What this relation is, and what it is not #
Newman's and Alexander's theorem is that stellar equivalence coincides with the existence of a common subdivision, hence — once realizations are available — with PL homeomorphism. That theorem is not proved here, and nothing below asserts it: the definition is taken as the combinatorial one, following Lickorish, Simplicial moves on complexes and manifolds, Geom. Topol. Monogr. 2 (1999), 299-320, and Rourke--Sanderson, Introduction to Piecewise-Linear Topology, Chapter 2. Naming the relation after the moves that generate it, rather than after that theorem, keeps the two apart.
A stellar move demands a fresh vertex, one spanning no face of the complex yet. That hypothesis
is what makes the move honest. The same-type relation records literal move sequences, while
StellarEquivalentUpToRelabeling compares complexes intrinsically: it first embeds both into a
common enlarged vertex type and allows arbitrary injective relabelings.
Main definitions #
PreAbstractSimplicialComplex.IsStellarMove: one starring of a face at a fresh vertex.PreAbstractSimplicialComplex.StellarEquivalent: a finite chain of such moves.PreAbstractSimplicialComplex.StellarEquivalentUpToRelabeling: intrinsic stellar equivalence in a common enlarged vertex type.
Main results #
PreAbstractSimplicialComplex.equivalence_stellarEquivalent: stellar equivalence is an equivalence relation.PreAbstractSimplicialComplex.StellarEquivalent.dimension_eq: stellar equivalent complexes have the same dimension.PreAbstractSimplicialComplex.StellarEquivalent.finite_faces_iff: stellar equivalence preserves and reflects finiteness of the face collection.PreAbstractSimplicialComplex.StellarEquivalentUpToRelabeling.induction_on: invariants of intrinsic stellar equivalence follow from the common-relabeling generators.PreAbstractSimplicialComplex.StellarEquivalentUpToRelabeling.dimension_eq,finite_faces_iff, andne_bot_iff: intrinsic stellar equivalence preserves dimension, face finiteness, and nonvoidness.
L is obtained from K by one stellar move: L is the starring of K at a face σ
using a vertex v that K does not already use.
Equations
- K.IsStellarMove L = ∃ (σ : Finset ι) (v : ι), σ ∈ K ∧ {v} ∉ K ∧ L = K.stellarSubdivision σ v
Instances For
The witness description of a stellar move.
Starring a face at a fresh vertex is a stellar move.
A stellar move preserves the dimension.
A stellar move preserves and reflects finiteness of the face collection.
Stellar equivalence: K and L are joined by a finite sequence of stellar moves and
inverse stellar moves.
Equations
Instances For
Every simplicial complex is stellar equivalent to itself.
Stellar equivalences compose.
Stellar equivalence is symmetric.
A single stellar move is a stellar equivalence.
An injective relabeling carries a stellar move to the corresponding stellar move between the image complexes.
A complex is stellar equivalent to any of its starrings at a fresh vertex.
Stellar equivalence is an equivalence relation.
To prove a property of stellar equivalent complexes, it suffices to prove it for a stellar move and show that it is reflexive, symmetric, and transitive.
An injective relabeling carries a stellar equivalence to one between the image complexes.
Stellar equivalent complexes have the same dimension.
Stellar equivalence preserves and reflects finiteness of the face collection.
A complex stellar equivalent to a complex with a face has a face itself.
Intrinsic stellar equivalence #
Stellar equivalence up to relabeling. At each generating step, the two complexes are
injectively relabeled inside the common enlarged vertex type ι ⊕ ℕ. The generator existentially
quantifies these relabelings, and the equivalence closure lets such comparisons compose.
Equations
- K.StellarEquivalentUpToRelabeling L = Relation.EqvGen (fun (A B : PreAbstractSimplicialComplex ι) => ∃ (f : ι ↪ ι ⊕ ℕ) (g : ι ↪ ι ⊕ ℕ), (A.map ⇑f).StellarEquivalent (B.map ⇑g)) K L
Instances For
A same-type stellar equivalence gives an intrinsic stellar equivalence.
Every complex is intrinsically stellar equivalent to itself.
Intrinsic stellar equivalence is symmetric.
Intrinsic stellar equivalences compose.
A stellar equivalence after injectively relabeling both complexes into the common enlarged vertex type gives an intrinsic stellar equivalence.
To prove a property of intrinsically stellar equivalent complexes, it suffices to prove it for a common-relabeling generator and show that it is reflexive, symmetric, and transitive.
An injective relabeling transports intrinsic stellar equivalence.
Intrinsically stellar equivalent complexes have the same dimension.
Intrinsic stellar equivalence preserves and reflects finiteness of the face collection.
Intrinsic stellar equivalence preserves and reflects nonvoidness.
A complex intrinsically stellar equivalent to a nonvoid complex is nonvoid.