0

I have to graphs which I want to unite, that is, create a new graph composed by the union of both graph's edges and nodes (without repetition). Is there an implementation for that avaliable in JUNG or do I have do so on my own?

Paulo
  • 267
  • 4
  • 14

1 Answers1

0

There isn't an implementation for that in JUNG, but it's about six lines of code assuming that the graphs, vertices, and edges are of the same types:

// given Graph g1, g2
Graph g = new [appropriate Graph implementation]
for (V v : Collections.union(g1.getVertices(), g2.getVertices())) {
  g.addVertex(v);
}
for (E e : g1.getEdges()) {
  g.addEdge(e, g1.getEndpoints(e));
}
for (E e : g2.getEdges()) {
  g.addEdge(e, g2.getEndpoints(e));
}

You can skip the vertex adding if there are no isolated vertices (i.e., vertices that have no incident edges); addEdge() will add any incident vertices.

If the graph is directed, you'll want to change the above to

g.addEdge(e, g1.getSource(e), g1.getDest(e));

Duplicates are silently ignored (if you want to know whether an add had an effect, check the return value).

Joshua O'Madadhain
  • 2,704
  • 1
  • 14
  • 18
  • Thanks, but now I have another doubt. Is there a way of getting the giant connected component of a graph or do I have to use the WeakComponentClusterer class and iterate over all the clusters in order to find the giant one? – Paulo Aug 24 '11 at 01:13
  • Your second question is answered in http://stackoverflow.com/questions/7182052/jung-how-to-get-the-giant-connected-component-of-a-graph – Joshua O'Madadhain Aug 30 '11 at 22:58