How to conduct network analysis with statistics about attributes

Question

I know this is not a pure-coding question, but I wanted to try anyway, since it could lead to a coding answer!

Supposing I have a dataframe describing frequency of communication between nodes like the following that I turned into a directed network:

sender	receiver	frequency
a	b	5
b	c	7
c	a	4

Now you can notice that the three nodes are connected and, if I make the network with graph_from_data_frame
then I would have that they are ALL connected and the only thing I can do to stress the fact that they are connected with different "weight" is by putting the frequency as E(g)$width to show it in the plot.

What I want to know is:
Is there - and if yes, how - a way to perform descriptive statistics on this type of network (like centrality, betweennes, ecc..) ? Igraph counts only 1 edge for each node and the stats are kinda obvious.

Answer 1

library(igraph)
library(dplyr)

df1 <- data.frame(
  stringsAsFactors = FALSE,
  sender = c("a", "b", "c"),
  receiver = c("b", "c", "a"),
  frequency = c(5L, 7L, 4L)
)

# Rename frequency to weight

df2 <- 
  df1 |> 
  rename(weight = frequency) 

# Create igraph object

g1 <- 
  graph_from_data_frame(df2) 

# Calculate degree centrality

strength(g1)
#>  a  b  c 
#>  9 12 11

# Or you can specify frequency using the weight argument

df1 |> 
  graph_from_data_frame() |> 
  strength(weights = df1$frequency)
#>  a  b  c 
#>  9 12 11