Claudio Martella wrote:
I'm writing an application with heavy graph operations. It's written
in python, so i've tried networkx, igraph and lately i've tried to use
boost graph library bindings to python (BGL-PY).
Thanks god i ran into graph-tool which looks very complete and also is
doing a transition into something very handy for me (python accessible
I'm having a look at the API i found here:
and i see i add vertices with g = Graph() g.add_vertices() and
calculate the betweeness centrality with
As i was looking for an implementation of a minimum spanning tree i
realized i couldn't find it in the reference i pointed a couple of
So i had a look at the graph_tool.py that uses the library and
realized that the interface is different there.
Graph = GraphInterface()
Graph.GetBetweennessCentrality() or Graph.GetMinimumSpanningTree()
etc, and i cannot find any add_vertex() in this API.
I'm a bit confused, so i'm trying to understand what i should use to
have full power over the graph-tool API.
The minimum spanning tree code has not yet been ported to the new
scheme, so it does not work at the moment. If you dig in the source code
you will find the old code, but as you noticed it will not make much
sense. Right now, you cannot really use it, sorry. :-(
The interface from graph-tool with C++ (with the BGL) is also something
I plan to document, so that it will be easy for people to write
extensions. There is even a module called run_action which allows inline
embedding of C++ code, which is automatically compiled and run (via
I'm currently finishing the documentation of the features which are
implemented. I will then include more algorithms from BGL (such as
minimum spanning tree, which is a high priority), which are relatively
I cannot promise any time schedule, but hopefully all this will be done
soon, as I'm working at it regularly. I'll post something in this list
when it is done.