



Thomassen [T] proved the conjecture when and showed
. In fact, this case is equivalent to the case
of the Bermond-Thomassen Conjecture.
The existence of the corresponding function for the undirected analogue is easy and has been observed by many authors. Stiebitz [S] even proved the following tight result: if the minimum degree of an undirected graph
is
, where each
is a non-negative integer, then the vertex set of
can be partitioned into
pairwise disjoint sets
, so that for all
, the induced subgraph on
has minimum degree at least
. This is clearly tight, as shown by an appropriate complete graph.
Bibliography
*[A] Noga Alon, Disjoint Directed Cycles, Journal of Combinatorial Theory, Series B, 68 (1996), no. 2, 167-178.
[S] M. Stiebitz, Decomposing graphs under degree constraints, J. Graph Theory 23 (1996), 31-324.
[T] C. Thomassen, Disjoint cycles in digraphs, Combinatorica 3 (1983), 393 - 396.
* indicates original appearance(s) of problem.