dev/dev_edge_betweenness.R

In antongrau/eliter: Contains the Danish Elite Network for 2012 and 2017 with analytical functions.

library(eliter)

# We want to assign edge weights to a graph on the basis of the edge.betweenness.estimate
# The graph has multiple edges or is unweighted. But for our purpose multiple edges the most important.


g <- graph_from_literal( A-B,
                    A-B,
                    A-C,
                    A-C,
                    B-C,
                    B-C,
                    A-X,
                    A-X,
                    A-X,
                    A-X,
                    A-X,
                    X-Y,
                    X-Z,
                    Z-Y,
                    W-Z,
                    Y-W,
                    X-W,
                    
                    simplify = FALSE)

plot(g)

# We want A - B - C and Z - X - Y to have the same weights. X - A has to be larger.
# We might accept that A - B and A - C is higher. They are the bridgers to a redundent set.
# But right now A-B and A-C are higher than Y - X and Z - X. 
# We should be able to ignore that they are connected to 

eb <- edge.betweenness.estimate(g, cutoff = 2, weights = NA)
E(g)$weight <- eb

data.frame(get.edgelist(g), weight = E(g)$weight)

sg <- simplify(g, edge.attr.comb = "sum")
be <- betweenness.estimate(sg, weights = NA, cutoff = 2)
graph.plot(sg, edge.text = E(sg)$weight, text = T, text.size = 5, vertex.fill = "white", vertex.size = be)

E(simplify(g))$weight

# Størrere eksempel ----
g <- graph_from_literal( A-B,
                         A-B,
                         A-C,
                         A-C,
                         B-C,
                         B-C,
                         A-X,
                         A-X,
                         A-X,
                         A-X,
                         A-X,
                         X-Y,
                         X-Z,
                         Z-Y,
                         W-Z,
                         Y-W,
                         X-W,
                         
                         X-J,
                         J-K,
                         M-J,
                         K-L,
                         L-M,
                         M-X,
                         J-L,
                         K-M,
                         X-L,
                         X-K,
                         
                         
                         simplify = FALSE)


plot(g)

# We want A - B - C and Z - X - Y to have the same weights. X - A has to be larger.
# We might accept that A - B and A - C is higher. They are the bridgers to a redundent set.
# But right now A-B and A-C are higher than Y - X and Z - X. 
# We should be able to ignore that they are connected to 

eb <- edge.betweenness.estimate(g, cutoff = 2, weights = NA)
E(g)$weight <- eb

data.frame(get.edgelist(g), weight = E(g)$weight)

sg <- simplify(g, edge.attr.comb = "sum")
be <- betweenness.estimate(sg, weights = NA, cutoff = 2)

graph.plot(sg, edge.text = E(sg)$weight, text = T, text.size = 5, vertex.fill = "white", vertex.size = be)

E(simplify(g))$weight

?adjacent_vertices
incident_edges(sg, )


# Two-mode example ----

e <- c("Anton", "A",
       "Jacob", "A",
       "Christoph", "A",
       "Andreas", "A",
       
       "Anton", "A2",
       "Jacob", "A2",
       "Christoph", "A2",
       "Andreas", "A2",
       
       #"Jacob", "B",
       "Lasse", "B",
       "Stefan", "B",
       "Hubert", "B",
       "Martin", "B",
       
       "Jacob", "C",
       "Lasse", "C",
       "Jacob", "C2",
       "Lasse", "C2"
       
      )

e  <- t(matrix(e, nrow = 2))

g  <- graph_from_incidence_matrix(table(e[,1], e[,2]), directed = FALSE)
eb <- edge.betweenness.estimate(g, cutoff = 5)
graph.plot(g, edge.alpha = eb, vertex.size = 5, text = T, vertex.fill = "white")

g1 <- bipartite_projection(g, multiplicity = T)$proj1
plot(g1, edge.width = E(g1)$weight, edge.color = E(g1)$weight)

g <- graph_from_literal( A-B,
                         A-B,
                         A-C,
                         A-C,
                         B-C,
                         B-C,
                         A-X,
                         A-X,
                         A-X,
                         A-X,
                         A-X,
                         X-Y,
                         X-Z,
                         Z-Y,
                         W-Z,
                         Y-W,
                         X-W,
                         
                         simplify = FALSE)

plot(g)