Components left by deleting a branch vertex of a tree #
Deleting a vertex c from a tree separates it into one component for each neighbour of c.
When every vertex of the complement has degree at most two there, the resulting components are
paths. This is the graph-theoretic extraction step behind the D and E branches of the
finite-type Dynkin-diagram classification.
Main results #
TauCeti.IsTree.neighborSetEquivConnectedComponentCompl: the neighbours of a vertex index the connected components left after deleting that vertex from a tree.TauCeti.IsTree.exists_equiv_pathGraph_components: deleting a vertex of degreenfrom a finite tree leavesncomponents, each a path when the complement has degree at most two.
References #
See J. E. Humphreys, Introduction to Lie Algebras and Representation Theory, Section 11.4, for the corresponding extraction in the classification of Dynkin diagrams.
The components of a tree with one vertex deleted are indexed by its neighbours.
The forward map sends a neighbour of c to the component of the graph induced on {c}ᶜ that
contains it; the inverse sends a component to the unique neighbour of c that it contains.
Equations
- TauCeti.IsTree.neighborSetEquivConnectedComponentCompl hG c = Equiv.ofBijective (fun (x : ↑(G.neighborSet c)) => (SimpleGraph.induce {c}ᶜ G).connectedComponentMk ⟨↑x, ⋯⟩) ⋯
Instances For
The equivalence sends a neighbour to the connected component containing that neighbour.
Deleting a vertex of degree n from a finite tree leaves n components, and they are paths
as soon as the complement has degree at most two.
The equivalence e records which component begins at each neighbour of c. The second conclusion
is deliberately componentwise: each component carries its own natural path length, which is the arm
length used in the subsequent reindexing onto a star.