Part 2a: Social Network Analysis

Network Diffusion

epiworldR also supports diffusion networks to simulate the spread of ideas or rumors throughout a population. The below example demonstrates this capability.

Example

Goal: Create a diffusion network model that simulates a rumor’s spread throughout the UKfaculty network of size n = 81. We start by loading the corresponding R packages and data.

We now load the UKfaculty data and visualize the network. The file part2a.rda has the network, the vertex attributes as a numeric matrix, and the edge list as a numeric matrix.

Code

# Loading the data for the section
load("part2a.rda")

# Visualizing the Network
# library(netplot)
nplot(UKfaculty)

Code

# First few lines of the vertex attributes
head(UKfaculty_vertex)

  Group age
1     3  34
2     1  47
3     3  56
4     3  42
5     2  63
6     2  65

Code

# First few lines of the edge list
head(UKfaculty_edgelist)

     [,1] [,2]
[1,]   57   52
[2,]   76   42
[3,]   12   69
[4,]   43   34
[5,]   28   47
[6,]   58   51

To simulate the rumor, we use the ModelDiffNet function:

Code

# Creating the diffusion model
library(epiworldR)
adopt_rumor <- ModelDiffNet(
  name       = "The dean is leaving!",
  prevalence = 2/vcount(UKfaculty), # Two adopter
  prob_adopt = .1,
  data       = UKfaculty_vertex,
  params     = c(0, 5)
)

# Reading in the network
agents_from_edgelist(
  adopt_rumor,
  size     = vcount(UKfaculty),
  source   = UKfaculty_edgelist[,1] - 1L,
  target   = UKfaculty_edgelist[,2] - 1L,
  directed = TRUE
  )

# Running the model for 50 steps
run(adopt_rumor, 100);adopt_rumor

_________________________________________________________________________
Running the model...
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| done.
 done.

________________________________________________________________________________
Diffusion of Innovations - The dean is leaving!
It features 81 agents, 1 virus(es), and 0 tool(s).
The model has 2 states.
The final distribution is: 0 Non adopters, and 81 Adopters.

Code

# Plotting the model
plot(adopt_rumor)

We can extract the adoption network:

Code

transmissions <- get_transmissions(adopt_rumor)

Warning in get_transmissions.epiworld_diffnet(adopt_rumor): The transmission
network is not necesarily relevant for the diffnet model

Code

# Turning it into a network via igraph
tnet <- graph_from_data_frame(
  transmissions[, c(2, 3)] + 1, # To start from 1 (not 0)
  directed = TRUE,
  vertices = cbind(id = 1:81, UKfaculty_vertex)
  )

# Adding back vertex attributes
nplot(tnet)