BFS on graph vertices and edges as neighbors

Question

I have a graph data structure, that I copied from this article - http://www.dreamincode.net/forums/topic/377473-graph-data-structure-tutorial/

And I'd like to implement the BFS algorithm on it. I'm not entirely sure how - most articles I see / read about the algorithm use simpler data structures. This data structure stores a hash map of vertices with their string representation as keys, and then also stores a hashmap of edges, using integers as keys.

Here's an example of the problem I run into when trying to implement a BFS example I found -

 public void bfs(Vertex rootNode){ Queue q = new LinkedList(); q.add(rootNode); rootNode.visited=true; while(!q.isEmpty()){ Vertex n = (Vertex)q.poll(); System.out.print(n.toString() + " "); for(Vertex adj : n.getNeighbors()){ -- Here's my problem. Get neighbors doesn't return a list of verts, it returns a list of edges. if(!adj.visited){ adj.visited=true; q.add(adj); } } } }

Do I need to call getNeighbors and then iterate over each unique vertex in the neighborhood?

Thank you.

Answer 1

You do need to call getNeighbors and iterate over each edge (hence each vertex).

public void bfs(Vertex rootNode){

        Queue q = new LinkedList();
        q.add(rootNode);
        rootNode.visited=true;
        while(!q.isEmpty()){
            Vertex n = (Vertex)q.poll();
            System.out.print(n.toString() + " ");
            for(Edge edge : n.getNeighbors()){
                Vertex adj = edge.getNeighbor(n);
                if(!adj.visited){
                    adj.visited=true;
                    q.add(adj);
                }
            }
        }

}