On 01/26/2014 11:23 PM, Ankit Arora wrote:
I'm working on a project which involves network graphs. Is there a library
that can help me do this:
I want to create multi-layered graphs i.e. graphs which contain a set
number of vertices but multiple 'layers' of edges i.e. same set of vertices
representing two or more properties in the same data structure.
One rather hacky solution can be to form a complete graph in igraph and
deal with the layers as if they were igraph edge attributes, though when
dealing with tens of thousands of vertices on a complete graph it will be
inefficient.
Any clue if something proper exists? If not, any more intelligent solutions
using existing libraries such as igraph/networkx?
Just a personal point of view: I usually end up implementing custom graphs or
trees because it is rather simple [*] and because of the variety of structures
and features. (Maybe that's why there are no general purpose node/tree/graph
libs, only highly specialised one, as for XML parsing.)
If you take this route and need help or advice on given points or features, we
can probably be of some use.
Denis
[*] compared the overall app: if a graph is complex, then the app it is a
component of is even more complex
_______________________________________________
Tutor maillist - Tutor@python.org
To unsubscribe or change subscription options:
https://mail.python.org/mailman/listinfo/tutor
