Get the name of the vertex in the igraph::transitivity(graph, type = "local") function in R

[silasalvescosta]

When I use the degree(graph) function from the igraph package in R, the output comes up with the name of the vertex and its respective value, however, when I run the transitivity(graph, type = "local") function from the same package, the output is a list without names, I can't tell which node each value corresponds to.

I suggest that the output of the transitivity(graph, type = "local") function to be accompanied by the name of the vertices.

[szhorvat]

Here's a demo to make it clear what this is about:

> g <- make_graph(~ a-b-c-a-d)
> transitivity(g,'local')
[1] 0.3333333 1.0000000 1.0000000       NaN
> degree(g)
a b c d 
3 2 2 1 
> betweenness(g)
a b c d 
2 0 0 0 
> hub_score(g)
$vector
        a         b         c         d 
1.0000000 0.8546377 0.8546377 0.4608111 

Normally, the result vector should include names.

Is this something to fix for 1.6.0 @krlmlr ? (I.e., is it considered breaking?)

[szhorvat]

In the long run (for 2.0?), I think it would be helpful to split this function into three separate function:

• Global clustering coefficient (currently global, globalundirected, undirected are all aliases for this). This returns a single number.
• Local clustering coefficient (currently local, localundirected are aliases for an unweighted version and barrat, weighted are aliases for a weighted version). These return a number for each vertex.
• Average local clustering coefficient (currently average, localaverage, localaverageundirected are aliases). These return a single number, which is the average of the local clustering coefficient for all vertices, but computed more efficiently than averaging the result of the previous category of functions.

These three are semantically different and return a different type of result (number vs vector).

[github-actions]

This old thread has been automatically locked. If you think you have found something related to this, please open a new issue and link to this old issue if necessary.