Re: [igraph] Help for transitivity vs clustering

[Gábor Csárdi]

Hi Charlotte,

there is indeed no consensus about the term AFAIK. The name 'transitivity' is older and comes from social networks. The name 'clustering coefficient' is newer, and came around when 'transitivity' was reinvented about 10-15 years ago.

There are basically two versions, one is a node-level index, the other is a network-level index, they are summarized well on the wikipedia page, I think:

http://en.wikipedia.org/wiki/Clustering_coefficient

igraph can calculate both of them.

As for you application, look at how the measures are defined, and then think about what you actually want to measure in your network, to see which one is appropriate, assuming they are appropriate at all. 

Best,

Gabor

On Fri, Mar 15, 2013 at 6:31 AM, Charlotte Bernard <address@hidden> wrote:

Dear igraph users,

I'm a new user of graph therory. I don't really understand the difference between clustering coefficient and transitivity. I have read a lot things about that and I see no consensus about it. Some authors don't make a difference between the two metrics, others do! In the study of brain network, Rubinov and Sporns (2010) differenciate the two metrics but I don't see a paper which uses transitivity to describe the networks.
In my study, I observed a loss of transvitivity in my patients' group whereas the clustering is not different.
Can somebody help me?
Thanks a lot,
All the best
C.

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

--
Gabor Csardi <address@hidden>     MTA KFKI RMKI