Browse · MathNet
PrintBaltic Way 2019
Baltic Way 2019 counting and probability
Problem
In a directed graph outgoing degree of each vertex equals (loops and bidirectional edges are allowed and considered as cycles). Prove that the graph contains non-intersecting directed cycles.
Solution
We proceed by induction on , the number of vertices.
For the base case we consider where the only possible digraph is the complete digraph which contains disjoint cycles of length .
We now assume that any -out digraph, with vertices, contains disjoint cycles.
Let be a digraph on vertices failing our assumption.
Then has no bidirectional edges. If is a bidirectional edge then is still -out so it contains a cycle, which paired with the -cycle made by the bidirectional edge gives the desired disjoint cycles.
The main idea allowing us to prove the result is using edge contractions. If there is an edge such that have no common parent, we can modify to a new digraph by removing and and adding a vertex for which its outgoing edges come to all children of and its ingoing edges come from all parents of and . has nodes and is still -out, as had no common parent, so by the inductive assumption contains disjoint cycles. If is not in any of the cycles they were contained in and we are done. The other option is that is contained in one of the cycles, this implies the other cycle is in . We distinguish the cases when the cycle in-edge of comes from 's in-edge or from 's in-edge. Replacing in the first case by and in the second by we get the other cycle in and we are done.
The only remaining option is when every edge of has a «witness» (defined as common parent to both its end-vertices). Let be a vertex having the smallest in-degree (we denote in-degree as ). As there are edges in total we have .
Case 1: . Then the edge starting in has no witnesses.
Case 2: . Then the edge ending in has no witnesses.
Case 3: . Let be two parents of . Then has to be the witness to and to implying is a bidirectional edge which is a contradiction.
Case 4: . Since our graph is exactly -out and the minimal in-degree is , then all the vertices have in-degree . Given a vertex and its parents we show that make a -cycle. Indeed, we notice that each of the witnesses of must be among implying that the induced subgraph is -in. This combined with the fact bidirectional edges do not exist, we conclude that make a -cycle as claimed. If we reverse the edges of we notice that all the above arguments still apply, as is both exactly -out and exactly -in, so children of also form a -cycle. As there are no bidirectional edges, the children and parents of are disjoint and give disjoint -cycles.
For the base case we consider where the only possible digraph is the complete digraph which contains disjoint cycles of length .
We now assume that any -out digraph, with vertices, contains disjoint cycles.
Let be a digraph on vertices failing our assumption.
Then has no bidirectional edges. If is a bidirectional edge then is still -out so it contains a cycle, which paired with the -cycle made by the bidirectional edge gives the desired disjoint cycles.
The main idea allowing us to prove the result is using edge contractions. If there is an edge such that have no common parent, we can modify to a new digraph by removing and and adding a vertex for which its outgoing edges come to all children of and its ingoing edges come from all parents of and . has nodes and is still -out, as had no common parent, so by the inductive assumption contains disjoint cycles. If is not in any of the cycles they were contained in and we are done. The other option is that is contained in one of the cycles, this implies the other cycle is in . We distinguish the cases when the cycle in-edge of comes from 's in-edge or from 's in-edge. Replacing in the first case by and in the second by we get the other cycle in and we are done.
The only remaining option is when every edge of has a «witness» (defined as common parent to both its end-vertices). Let be a vertex having the smallest in-degree (we denote in-degree as ). As there are edges in total we have .
Case 1: . Then the edge starting in has no witnesses.
Case 2: . Then the edge ending in has no witnesses.
Case 3: . Let be two parents of . Then has to be the witness to and to implying is a bidirectional edge which is a contradiction.
Case 4: . Since our graph is exactly -out and the minimal in-degree is , then all the vertices have in-degree . Given a vertex and its parents we show that make a -cycle. Indeed, we notice that each of the witnesses of must be among implying that the induced subgraph is -in. This combined with the fact bidirectional edges do not exist, we conclude that make a -cycle as claimed. If we reverse the edges of we notice that all the above arguments still apply, as is both exactly -out and exactly -in, so children of also form a -cycle. As there are no bidirectional edges, the children and parents of are disjoint and give disjoint -cycles.
Techniques
Graph TheoryInduction / smoothingColoring schemes, extremal arguments