Bipartite network analysis

# Package igraph
library(igraph)


# Read Data through CSV files
data=read.csv(file.choose(),header = T)


# Create Network: A directed graph
net<-graph.data.frame(data,directed=T)

# Check whether the network is bipartite or not with is.bipartite function in igraph package
is_bipartite(net)
net                   # The DN-- in output means that it is a directed network

# Although networks loaded are bipartite, the is.bipartite function giving FALSE, therefore
g_el <- as_edgelist(net)
V(net)$type <- ifelse(V(net)$name %in% g_el[,1], TRUE, FALSE)
is_bipartite(net)
net                    # The DN-B in output means that it is a directed network and also bipartite,
# Centrality Measures: https://rpubs.com/pjmurphy/317838

types <- V(net)$type                 ## getting each vertex `type` let's us sort easily
deg <- degree(net)
bet <- betweenness(net)
clos <- closeness(net)
eig <- eigen_centrality(net)$vector

cent_df <- data.frame(types, deg, bet, clos, eig)

cent_df[order(cent_df$type, decreasing = TRUE),] ## sort w/ `order` by `type`

# Bi-partite statistics

# Number of top and bottom nodes
top<-length(V(net)[type==FALSE])
bottom<-length(V(net)[type==TRUE])

# Number of edges
m<-ecount(net)

# Mean degree for top and bottom nodes
ktop<-m/top
kbottom<-m/bottom

# Density for bipartite network
bidens<-m/(top*bottom)

# Largest connected component for top and bottom nodes:
gclust<-clusters(net, mode='weak')
lcc<-induced.subgraph(net, V(net)[gclust$membership==1])
lcctop<-length(V(lcc)[type==FALSE])
lccbottom<-length(V(lcc)[type==TRUE])

# Mean distance for top and bottom nodes
distop<-mean(shortest.paths(lcc, v=V(lcc)[type==FALSE], to=V(lcc)[type==FALSE], mode = 'all'))
disbottom<-mean(shortest.paths(lcc, v=V(lcc)[type==TRUE], to=V(lcc)[type==TRUE], mode = 'all'))

' largest Connected component'
c=clusters(net)
lcomponent=max(c$csize)

'Clustering coefficent for each node'   ### As given in the Sciencedirect whitepaper transitivity and clustering coeff are same
clucoeff=transitivity(net,type="local")