Re: [igraph] random graph

[Thiago M. Venancio]

Hi Gabor,

I definitely do not expect all the edges to be different. However, what is happening is that the a big fraction of the randomized edges are the same in, for example 900/1000 random graphs generated.

If I compare the random graph with the original, that's surely expected to have an amount of edges retained. However, I think they should not be the same across several randomized graphs.

Thanks.

Thiago

On Thu, Aug 21, 2008 at 10:39 AM, Csardi Gabor <address@hidden> wrote:

Thiago,

why do you expect that ALL edges will be different in the
randomized graph? Some edges will be the same, just by chance, and
especially the big hubs have a good chance for this.

Gabor

On Thu, Aug 21, 2008 at 10:35:38AM -0400, Thiago M. Venancio wrote:
> Hi Guys...
> I have tested the suggestions, but am experiencing a problem here. By simple
> rewiring I can reach up to 85% of rewiring, what seems pretty good. However,
> the 15% remaining edges are retained in a large fraction of the random
> graphs I created. Thus, I think they are not that random...
> I have also tested the method Gabor suggested, but the problem persists.
> Typically the problematic edges are involving one of the large hubs of the
> graph.
> Any clues ?
> Thanks.
> Thiago
>
> On Thu, Aug 14, 2008 at 3:09 AM, Csardi Gabor <address@hidden> wrote:
>
> > Hmmm, degree.sequence.game is another option, especially
> > in the new 0.5.1 R package, with method="vl". It is perhaps
> > faster for larger graphs and it is known to be a uniform
> > sampler.
> >
> > Gabor
> >
> > On Wed, Aug 13, 2008 at 07:55:38PM +0100, Tamas Nepusz wrote:
> > > Hi Thiago,
> > >
> > > >Is it possible to make an undirected random graph from a reference
> > > >one, keeping the same degree for each vertex ?
> > > Yes, it is. In C, see igraph_rewire. In Python, it's Graph.rewire()
> > > and I think it's simply called "rewire" in R.
> > >
> > > --
> > > Tamas
> > >
> > >
> > >
> > > _______________________________________________
> > > igraph-help mailing list
> > > address@hidden
> > > http://lists.nongnu.org/mailman/listinfo/igraph-help
> >
> > --
> > Csardi Gabor <address@hidden>    MTA RMKI, ELTE TTK
> >
> >
> > _______________________________________________
> > igraph-help mailing list
> > address@hidden
> > http://lists.nongnu.org/mailman/listinfo/igraph-help
> >
>
>
>
> --
> ========================
> Thiago Motta Venancio, M.Sc., PhD
> http://tmvenancio.googlepages.com
> ========================

> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help


--
Csardi Gabor <address@hidden>    MTA RMKI, ELTE TTK


_______________________________________________
igraph-help mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/igraph-help

--
========================
Thiago Motta Venancio, M.Sc., PhD
http://tmvenancio.googlepages.com
========================