Re: [igraph] Node-level triad census with igraph for R

[Gábor Csárdi]

Hi Marco,

On Wed, Jan 15, 2014 at 5:49 AM, Marco Milano <address@hidden> wrote:

Dear igraph list,

I'm currently working on extracting node-level properties from a social network and I'm stuck with the triad census calculation. As there isn't a function to call on the graph to perform such task, this is the most updated solution I found so far: http://stackoverflow.com/questions/12374534/how-to-mine-for-motifs-in-r-with-igraph. 

actually there is triad.census:

http://igraph.sourceforge.net/doc/R/triad.census.html

 

When I try to iterate the triad census on each node in my graph though, I get some strange results. First of all, being an undirected graph I should be getting 4 values but instead I get 16. Moreover, I don't know which resulting census corresponds to which node. The name attribute I call to iterate on the list of nodes seems to contain the NULL value but it correctly contains the vertex names at graph level before calling the graph.neighborhood function.

To be honest, it is hard to follow what you are doing, without seeing your code. If just calling triad.census() is not an option for you, then please include your code. Thanks.

Best,

Gabor

 

My problem is: how can I have each node printed out along with its triad census? 

And, please forgive me for adding a side-question to my main one but seen that we are talking about triads, I was wondering if anyone is aware of a way to calculate the Kalish & Robins Strong-Weak-NoTie relation?

Thanks. I appreciate any help or advice.

Marco 

-- 

Find me on LinkedIn:

linkedin.com/in/marcomilano

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