Chromatic Polynomial and Edge Coloring
The chromatic polynomial encodes the number of proper colorings as a function of the number of colors. Edge coloring, a parallel theory, assigns colors to edges so that edges sharing a vertex receive different colors.
Chromatic Polynomial
The chromatic polynomial counts the number of proper -colorings of . Key properties: (1) is a polynomial in of degree ; (2) the leading coefficient is 1; (3) the coefficient of is ; (4) . For connected : the coefficients alternate in sign and for all real .
The deletion-contraction recurrence: for any edge . Base cases: (empty graph), . For the cycle : . For a tree : .
Edge Coloring
A proper edge coloring assigns colors to edges so that edges sharing a vertex get different colors. The chromatic index is the minimum number of colors needed. Since each vertex has degree , we need . Vizing's theorem: for simple graphs. Graphs with are Class 1; those with are Class 2.
Classifying graphs as Class 1 or 2 is NP-complete. Known results: bipartite graphs are Class 1 (KΓΆnig's theorem), planar graphs with are Class 1 (Vizing), and the Petersen graph is Class 2 (, ). Multigraphs satisfy (Shannon's theorem) and the stronger bound where is the maximum edge multiplicity.
List Coloring
In list coloring, each vertex has a list of permitted colors, and we seek a proper coloring with . The list chromatic number is the minimum such that is -colorable for every assignment of lists of size . Always , but equality can fail: but . The list coloring conjecture asserts for every graph.
Galvin (1995) proved the list coloring conjecture for bipartite graphs: for bipartite . The proof uses a clever orientation argument and the theory of kernel-perfect graphs.