I am trying to recreate figure 2 from Hric, Peixoto, and Fortunato (2016):

https://arxiv.org/abs/1604.00255 and would appreciate some assistance. The

data come from the college football network and can be pulled directly from

an online repository using graph-tool:

g = gt.collection.data["football"]

I can recreate figure 2a simply by estimating a nested blockmodel and

graphing it as follow:

state = minimize_nested_blockmodel_dl(g, deg_corr=True)

draw_hierarchy(state, output="./results/test.png")

However, I am stuck on figure 2b, which graphs the bipartite network of

conferences and teams. Does this figure plot the output of estimating the

blockmodel on layered data? If so, how should I extract the metadata, which

are stored as the vertex property "label"? Do I need to create a second edge

list linking the original vertices to the metadata? And if so, is there a

way of doing this without a long loop? Eventually I want to run this code on

a much larger network so I would like to write code that is as efficient as

possible.

Thanks for your help,

Jamie