Re: [igraph] Creation of a random graph with high clustering coefficiant

[Jeff Hemsley]

Hi Dov.
Great suggestion. I tried this and can see how it is useful in many places.

On Tue, Dec 4, 2012 at 10:05 AM, Dov Pechenick <address@hidden> wrote:
> Hi Jeff,
>
> Perhaps there is a smarter way to do this, but I've found that a rewiring
> approach works well (though it can be slow) for a number of different
> network properties.  You first use the appropriate graph constructor to get
> a network with the desired degree distribution.  Then you can use the rewire
> function to swap a single pair of edges in a way that preserves the degree
> sequence.  After the swap, you check to see if average clustering
> coefficient is changed from before.  If it is greater than or equal to the
> previous value, you accept the change, otherwise you reject.
>
> This approach is nice in that it can be generally applied to any network and
> any network property you are trying to maximize/minimize.  However, this
> also makes it naive and general in a way that may not suit you, and someone
> may know of a more specific way to construct a random graph that has a
> certain property of interest.  Either way, I hope this helps.
>
> Dov
>
>
> On Tue, Dec 4, 2012 at 12:40 PM, Jeff Hemsley <address@hidden>
> wrote:
>>
>> looking for a function to create a random graph where nodes have a
>> high clustering coefficient.
>
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/igraph-help
>



-- 
Jeff Hemsley
Doctoral Candidate
The Information School
University of Washington
http://ischool.uw.edu/
address@hidden