I'm trying to use the fglsnet package in R to run a regression in which my dependent variable is a centrality measure, and the independent variable is an attribute of the vertices.
It is not clear to me the type/class of argument needed for "dependence network" M, and if it needs to be included in the data argument.
This is what I've done:
fglsnet(participation_coefficient ~ leader,communication_M_na5,
directed = FALSE,mcorr = TRUE,k = 10,data = data)
data here is a list including a data frame with info on participation coefficient (Guimerà & Amaral, 2005) and leadership, and the matrix "communication_M_na5" in which I have info on how many communications two vertices have had.