[igraph] multilevel.community

[joe j]

Dear all,

I have a single mode network on which I tried to implement the so
called Louvain algorithm to uncover community structure (I've been
playing with other community algorithms before). For this I've used
the "multilevel.community' in igraph6. I've two questions for
clarification:

1. It seems to me that the the program reports only the final
community structure (?) I am asking this because with
"cohesive.blocks" for example you get the hierarchy level and
communities/blocks within each hierarchy.
2. I also notice that only nodes that fall into a community are
retained (I created the data frame after using "membership()" ; I
guess this is the correct approach). This probably makes sense, but it
might also be good to know if there are smaller "components" for
example, although they may not satisfy the community definition.

Best regards,
Joe.