Edge splitting is the reverse of edge contraction. In a simple graph, it replaces a graph
vertex
by adjacent vertices
and
and distributes the former neighbors
of
between them, allowing a former neighbor
to be adjacent to both. Contracting
, deleting the resulting graph
loop, and identifying parallel edges recovers
the original graph.
Under this inverse-contraction convention, the -dimensional operation of vertex
splitting is a restricted edge split, with exactly
common neighbors of the
two new vertices. The terminology is not uniform.
In rigidity theory, "edge splitting"
also denotes the different operation that deletes an graph
edge
and adds a new graph vertex joined to
,
, and
other vertices (Schulze
and Whiteley 2017, pp. 1597-1599). That operation splits an graph
edge, whereas vertex splitting splits an
existing graph vertex.