Re: [igraph] I am not understanding something.....

[Tamas Nepusz]

> Now, how can i build a cycle like "for every vertex in G"?
Either you can simply loop over all vertex indices in a for loop:

for i in xrange(g.vcount()):
  print g.neighbors()    # or do whatever you want

Or if the core of your loop is simple, you can use a list  
comprehension from Python 2.4 up:

deg_squared = [g.degree(i) ** 2 for i in xrange(g.vcount())]

(this returns the square of the degrees -- okay, pretty dump example,  
I know)

A simple way to build an adjacency list of the graph is:

adj_list = [g.neighbors(i) for i in xrange(g.vcount())]


If you want to do something with the vertex attributes, you can also  
loop over the vertex sequence of the graph this way:

for vertex in g.vs:
  print vertex["name"], g.degree(vertex.index)

g.vs returns a VertexSeq object, which practically behaves like an  
array, so you can iterate over it, you can index it and so on. See  
help(VertexSeq) for more details.

--
Tamas

> Thanks again!
>
> marco
>
>
> On Wed, Jun 18, 2008 at 1:17 PM, Tamas Nepusz <address@hidden>  
> wrote:
> > > I am not understanding how to cycle on every node of a graph, and  
> > > once
> > > "i am looking" at a one node, how to cycle between the neighbors of
> > > that node....
> >
> > Given a Graph object g, g.neighbors(i) gives you the neighbors of  
> > node i.
> > g.successors(i) gives you the successors of that node (following  
> > outgoing
> > edges), while g.predecessors(i) gives you the predecessors  
> > (following the
> > incoming edges). Note that you'll have to supply vertex IDs to
> > g.neighbors().
> >
> > E.g.:
> >
> > > > > g = Graph.Tree(10,3)
> > > > > g.neighbors(0)
> > [1, 2, 3]
> >
> > --
> > T.
> >
> >
> >
> > _______________________________________________
> > igraph-help mailing list
> > address@hidden
> > http://lists.nongnu.org/mailman/listinfo/igraph-help
>
>
>
> --
> Calvin Coolidge  - "I have never been hurt by what I have not said."
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help