MerLean-example

29 Def 19: Edge Boundary and Vertex Incidence for Cell Complexes

Definition 330 Edge Boundary of a Vertex Subset
#

Let \(X\) be a cell complex and let \(S \subseteq X_0\) be a subset of vertices (0-cells). The edge boundary of \(S\) is the set

\[ \delta S \; =\; \bigl\{ \, e \in X_1 \; \big|\; (\exists \, u \in \partial _1 e,\; u \in S) \; \land \; (\exists \, w \in \partial _1 e,\; w \notin S) \, \bigr\} . \]

That is, an edge \(e \in X_1\) belongs to \(\delta S\) if and only if its boundary \(\partial _1 e\) intersects both \(S\) and its complement \(S^c\).

Theorem 331 Membership in the Edge Boundary

Let \(X\) be a cell complex, \(S \subseteq X_0\), and \(e \in X_1\). Then

\[ e \in \delta S \; \Longleftrightarrow \; \bigl(\exists \, u \in \partial _1 e,\; u \in S\bigr) \; \land \; \bigl(\exists \, w \in \partial _1 e,\; w \notin S\bigr). \]
Proof

This follows immediately by unfolding the definition of \(\delta S\) and applying simp.

Lemma 332 Edge Boundary Nonempty

Let \(X\) be a cell complex and \(S \subseteq X_0\). If there exists an edge \(e \in X_1\) such that \(\partial _1 e\) contains a vertex in \(S\) and a vertex not in \(S\), then \(\delta S\) is nonempty.

Proof

Assume there exists \(e \in X_1\) with \((\exists \, u \in \partial _1 e,\; u \in S)\) and \((\exists \, w \in \partial _1 e,\; w \notin S)\). From the hypothesis \(h\), we obtain such an edge \(e\) together with the witness \(h_e\) of both conditions. By the membership characterization \(\texttt{mem\_ edgeBoundaryCell}\), we have \(e \in \delta S\), so \(\delta S\) is nonempty.

Definition 333 Vertices Incident to a Set of Edges
#

Let \(X\) be a cell complex and let \(E \subseteq X_1\) be a set of edges (1-cells). The set of vertices incident to \(E\) is

\[ \Gamma (E) \; =\; \bigl\{ \, v \in X_0 \; \big|\; \exists \, e \in E,\; v \in \partial _1 e \, \bigr\} . \]
Theorem 334 Membership in Incident Vertices

Let \(X\) be a cell complex, \(E \subseteq X_1\), and \(v \in X_0\). Then

\[ v \in \Gamma (E) \; \Longleftrightarrow \; \exists \, e \in E,\; v \in \partial _1 e. \]
Proof

This follows immediately by unfolding the definition of \(\Gamma (E)\) and applying simp.

Lemma 335 Incident Vertices Nonempty

Let \(X\) be a cell complex and \(E \subseteq X_1\). If there exists an edge \(e \in E\) with a vertex \(v \in X_0\) such that \(v \in \partial _1 e\), then \(\Gamma (E)\) is nonempty.

Proof

From the hypothesis \(h\), we obtain \(e \in E\) and \(v \in X_0\) with \(v \in \partial _1 e\). By the membership characterization \(\texttt{mem\_ incidentVerticesCell}\), we conclude \(v \in \Gamma (E)\), so \(\Gamma (E)\) is nonempty.

Definition 336 Edges Incident to a Vertex (Star)
#

Let \(X\) be a cell complex and \(v \in X_0\) a vertex. The set of edges incident to \(v\) (the star of \(v\)) is

\[ \delta v \; =\; \bigl\{ \, e \in X_1 \; \big|\; v \in \partial _1 e \, \bigr\} . \]
Theorem 337 Membership in Incident Edges

Let \(X\) be a cell complex, \(v \in X_0\), and \(e \in X_1\). Then

\[ e \in \delta v \; \Longleftrightarrow \; v \in \partial _1 e. \]
Proof

This follows immediately by unfolding the definition of \(\delta v\) and applying simp.

Lemma 338 Incident Edges Nonempty

Let \(X\) be a cell complex and \(v \in X_0\). If there exists \(e \in X_1\) such that \(v \in \partial _1 e\), then \(\delta v\) is nonempty.

Proof

From the hypothesis \(h\), we obtain \(e \in X_1\) with \(v \in \partial _1 e\). By the membership characterization \(\texttt{mem\_ incidentEdges}\), we conclude \(e \in \delta v\), so \(\delta v\) is nonempty.

Definition 339 Edge Boundary at a Vertex

Let \(X\) be a cell complex, \(S \subseteq X_0\), and \(v \in X_0\). The edge boundary of \(S\) at vertex \(v\) is the intersection

\[ (\delta S)_v \; =\; \delta S \cap \delta v \; =\; \bigl\{ \, e \in X_1 \; \big|\; e \in \delta S \; \land \; v \in \partial _1 e \, \bigr\} . \]
Theorem 340 Membership in the Edge Boundary at a Vertex

Let \(X\) be a cell complex, \(S \subseteq X_0\), \(v \in X_0\), and \(e \in X_1\). Then

\[ e \in (\delta S)_v \; \Longleftrightarrow \; e \in \delta S \; \land \; v \in \partial _1 e. \]
Proof

This follows immediately by unfolding the definition of \((\delta S)_v\) and applying simp.

Theorem 341 Edge Boundary at Vertex is Subset of Edge Boundary

Let \(X\) be a cell complex, \(S \subseteq X_0\), and \(v \in X_0\). Then

\[ (\delta S)_v \; \subseteq \; \delta S. \]
Proof

Since \((\delta S)_v\) is defined as the filter of \(\delta S\) by the predicate \(v \in \partial _1 e\), the inclusion \((\delta S)_v \subseteq \delta S\) holds by the general fact that a filtered finset is a subset of the original finset (Finset.filter_subset).

Theorem 342 Edge Boundary at Vertex is Subset of Incident Edges

Let \(X\) be a cell complex, \(S \subseteq X_0\), and \(v \in X_0\). Then

\[ (\delta S)_v \; \subseteq \; \delta v. \]
Proof

Let \(e\) be arbitrary and assume \(e \in (\delta S)_v\). Rewriting using the membership characterization \(\texttt{mem\_ edgeBoundaryVertex}\), we obtain \(e \in \delta S\) and \(v \in \partial _1 e\). Rewriting the goal using \(\texttt{mem\_ incidentEdges}\), it suffices to show \(v \in \partial _1 e\), which is precisely the second component of the hypothesis.

Lemma 343 Edge Boundary at Vertex Nonempty

Let \(X\) be a cell complex, \(S \subseteq X_0\), and \(v \in X_0\). If there exists \(e \in X_1\) such that \(e \in \delta S\) and \(v \in \partial _1 e\), then \((\delta S)_v\) is nonempty.

Proof

From the hypothesis \(h\), we obtain \(e \in X_1\) with the witness \(h_e\) satisfying \(e \in \delta S\) and \(v \in \partial _1 e\). By the membership characterization \(\texttt{mem\_ edgeBoundaryVertex}\), we conclude \(e \in (\delta S)_v\), so \((\delta S)_v\) is nonempty.

Theorem 344 Alternate Characterization of Edge Boundary at Vertex

Let \(X\) be a cell complex, \(S \subseteq X_0\), \(v \in X_0\), and \(e \in X_1\). Then

\[ e \in (\delta S)_v \; \Longleftrightarrow \; v \in \partial _1 e \; \land \; \bigl(\exists \, u \in \partial _1 e,\; u \in S\bigr) \; \land \; \bigl(\exists \, w \in \partial _1 e,\; w \notin S\bigr). \]
Proof

Rewriting with \(\texttt{mem\_ edgeBoundaryVertex}\), the left-hand side becomes \((e \in \delta S) \land (v \in \partial _1 e)\). Rewriting with \(\texttt{mem\_ edgeBoundaryCell}\), the condition \(e \in \delta S\) expands to \((\exists \, u \in \partial _1 e,\; u \in S) \land (\exists \, w \in \partial _1 e,\; w \notin S)\). The equivalence with the right-hand side then follows by propositional logic (tauto).