On Mon, 6 Jun 2011, David Doria wrote:
1) When is it appropriate to use adjacency_list directly vs using directed_graph and undirected_graph?
I generally just use adjacency_list, since the parameters are easy to look up.
2) I am trying to iterate over the edges of a directed_graph: http://codepad.org/jO3rJLtF
In the first method I tried:
typedef boost::graph_traits < Graph >::out_edge_iterator out_edge_iterator; std::pair<out_edge_iterator, out_edge_iterator> outEdges = boost::out_edges(v1, g);
, memory addresses seem to be output instead of vertex ids.
Those are the vertex descriptors; they are void*'s for adjacency_list graphs with listS vertex containers (of which directed_graph is one).
In the second method:
typedef boost::property_map<Graph, boost::vertex_index_t>::type IndexMap; IndexMap index = get(boost::vertex_index, g);
, vertex ids are output, but I didn't specify if I wanted in edges or out edges?
The loop starting on line 50 of your code iterates through all of the edges of the graph; the concepts of in edges and out edges relate to a single vertex.
When using adjacency_list directly, I used to have to specify boost::bidirectionalS:
typedef boost::adjacency_list<boost::vecS, boost::vecS, boost::bidirectionalS> Graph;
Is something like that required when using directed_graph?
The directed_graph class appears to use a bidirectionalS adjacency_list automatically. -- Jeremiah Willcock