The computational complexity of cordial and equitable labelling (Q1567263)

From MaRDI portal
scientific article
Language Label Description Also known as
English
The computational complexity of cordial and equitable labelling
scientific article

    Statements

    The computational complexity of cordial and equitable labelling (English)
    0 references
    0 references
    0 references
    11 December 2000
    0 references
    For a graph \(G=(V,E)\), a labelling \(f: V \rightarrow \{0,1,\ldots, k-1\}\) is \(k\)-equitable, if for each \(0 \leq i < j \leq k-1\), we have \(|V_i|- |V_j|\in \{-1,0,1\}\) and \(|E_i|- |E_j|\in \{-1,0,1\}\), where \(V_i\) is the set of vertices with label \(i\), and \(E_i\) is the set of edges \(\{v,w\}\) with \(|f(v) - f(w)|= i\). A 2-equitable labelling is called cordial, i.e., the vertices are partitioned into two sets of equal or almost equal size, such that (about) half of the edges are between vertices in different classes. It is shown that it is NP-complete to decide if a given graph has a cordial labelling; and hence it is also NP-complete to decide if a graph has a \(k\)-equitable labelling. As an intermediate result, the authors show that the problem, given a graph with an even number of edges, to decide if the vertices can be partitioned into two classes with exactly half of the edges from one class to the other is NP-complete.
    0 references
    0 references
    0 references
    0 references
    0 references
    graph labelling
    0 references
    cordial graphs
    0 references
    equitable labelling
    0 references
    NP-completeness
    0 references
    0 references