1. Sampson Monk Network Description: It is a statnet dataset Number of Nodes: 18, monks Number of Edges: 88, ranked choice out of -3/3 like/dislike of other specified monks (ergo it is a directed network) It is NOT a bipartite/affiliation network. It’s an edgelist. It is weighted, -3 to 3 is weighted, idk what igraph/statnet are on about Node attributes: a.”cloisterville” = classifer for monks that attended seminary of “Cloisterville” previously b.”group” = Sampson identified 3 groups that formed between the monks c.”na” d.”vertex.names” & “name” = chosen names of monks Edge attributes:
  1. “na”
  2. “nominations” = rankings of the monks’ like/dislike survey question. No isolates
  1. Noteworthy Results:

  1. Top Attributes: With their high total degree number, John Bosco and Gregory show that they are popular/infamous and major players in the Young Turks faction. Most people only have two indegrees. Though, Gregory was ousted eventually and John Bosco wasn’t and this may be because JB is highest in status/closeness. Victor is second and highest among the Loyalists. It is interesting a Young Turk is higher in status than all Loyalists. JB is highest in eigenvector value and Bon power, but it isn’t all positive edges. Surprisingly, an Outcast, Basil is second highest in Bon power. So Basil is connected to a lot of connected people and is in a good position in the network, but this contadicts his outcome as being ousted.

  2. The outcasts have lower indegree mean than other groups overall and compared to their outdegree mean. They are less garnering like/dislike from the others, which makes sense since they are called outcasts. The young turks has a the highest of both indegree and outdegree, which makes me think they are pointing a lot of fingers and having a lot of fingers pointed at them. They are shaking up the culture, so that makes sense.

  3. Network Centralization: OutDeg = 0.069 & InDeg = 0.381. Low out degree may have to do with only allowing the monks to choose 3-4 nominations, making the density artificially low. In degree being so high in comparison is interesting. It might have to do with a few people getting a lot of votes like JB.

vcount(ig.sam)
ecount(ig.sam)
is.bipartite(sam)
is.directed(sam)
is.weighted(ig.sam)
vertex_attr_names(ig.sam)
edge_attr_names(ig.sam)
print(sam)
head(sam %v% "vertex.names")
head(sam %v% "group")
head(sam %e% "nominations")
summary(sam %e% "nominations")

igraph::triad.census(ig.sam)
sna::dyad.census(sam)
sna::triad.census(sam, mode="graph")
gtrans(sam) 
isolates(sam)
transitivity(ig.sam, type="global")
transitivity(ig.sam, type="average")
V(ig.sam)[c(1:18)]
V(ig.sam)$name<-V(ig.sam)$vertex.names
transitivity(ig.sam, type="local", vids=V(ig.sam)[c("Basil","John Bosco", "Gregory")])
distances(ig.sam,"Winfrid","Elias", weights = NA)#weights=NA = any available edge weights are ignored. #Calculate distances between two nodes
all_shortest_paths(ig.sam,"Winfrid","Elias", weights=NA)$res # list shortest paths between 2 nodes
average.path.length(ig.sam,directed=T)#find average shortest path for network
igraph::components(ig.sam)$no #Number of components
igraph::components(ig.sam)$csize #Size of each component

gden(sam, diag=FALSE)
graph.density(ig.sam)
monk.nodes<-monk.nodes %>%
    mutate(totdegree=igraph::degree(ig.sam, loops=FALSE),
           indegree=igraph::degree(ig.sam, mode="in", loops=FALSE),
           outdegree=igraph::degree(ig.sam, mode="out", loops=FALSE))
centr_degree(ig.sam, loops = FALSE, mode="in")$centralization
centr_degree(ig.sam, loops = FALSE, mode="out")$centralization
monk.nodes$group<-V(ig.sam)$group
monk.nodes%>%
  group_by(group)%>%
  select(indegree,outdegree)%>%
   summarise_all(.,funs(mean,n()))
hist(monk.nodes$indegree, main="In-degree Distribution", xlab="Nominations Received")
hist(monk.nodes$outdegree, main="Out-degree Distribution ", xlab="Nominations made")
arrange(monk.nodes, desc(totdegree))%>%slice(1:5) #highest indegree
arrange(monk.nodes, desc(bonpow))%>%slice(1:5) #highest outdegree
arrange(monk.nodes, indegree)%>%slice(1:5) #lowest indeg
summary(monk.nodes)