Re: [igraph] Generation of Random Graphs According to Some Rules

[Bernie Hogan]

Dear Lorenzo,

I recommend sending this request to the socnet mailing list.
http://www.insna.org/pubs/socnet.html

Social network analysts have been doing random graph generation
according to a variety of rules for a long time. You will find some
help in the statnet package for R.

Betweenness, being a global measure, would seem to me to be the least
likely to be available in a random method, and it is not something
I've seen. I'm almost certain I've seen methods that give prescribed
clustering coefficients. But I do not do this sort of thing myself, so
I can't give a great answer.

Take care,
BERNiE

Bernie Hogan
Research Fellow, Oxford Internet Institute
University of Oxford




These graphs are the basis of ERGm modelling, which is a popular but
computationally expensive tool used in the social sciences.

On Sun, Oct 25, 2009 at 4:46 PM, Jonathan Donges
<address@hidden> wrote:
> Dear Lorenzo,
> There is a paper by G. Zamora-Lopez et al., namely "The generation of random
> directed networks with
> prescribed 1-node and 2-node degree correlations", in
> J. Phys. A: Math. Theor. 41 (2008) 224006 (15pp)
> doi:10.1088/1751-8113/41/22/224006
> This could be exactly what you are looking for, at least for the
> degree-degree correlation case.
> Cheers,
> Jonathan
> Am 25.10.2009 um 16:44 schrieb Tamas Nepusz:
>
> Dear Lorenzo,
>
> I have no idea of how to deal with requirements (2) to (5) [it does not
> sound trivial to me] hence any suggestion is really appreciated.
>
> I am not aware of any graph rewiring methods that are able to preserve the
> properties you mentioned, except for the degree distribution. The problem is
> that a single rewiring step (i.e., the addition, removal or rearrangement of
> an edge) has a very unpredictable effect on the properties you mentioned. I
> could be wrong, though...
>
> --
> Tamas
>
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help
>
>