Converting Igraph to VisNetwork

Question

I have this network graph that I made using the "igraph" library:

library(tidyverse)
library(igraph)


set.seed(123)
n=15
data = data.frame(tibble(d = paste(1:n)))

relations = data.frame(tibble(
  from = sample(data$d),
  to = lead(from, default=from[1]),
))

data$name = c("new york", "chicago", "los angeles", "orlando", "houston", "seattle", "washington", "baltimore", "atlanta", "las vegas", "oakland", "phoenix", "kansas", "miami", "newark" )

graph = graph_from_data_frame(relations, directed=T, vertices = data) 
(edge_fac <- forcats::as_factor(get.edgelist(graph)[,1]))
n2 <- as.integer(factor(data$name,levels = levels(edge_fac)))


V(graph)$color <- ifelse(data$d == relations$from[1], "red", "orange")
V(graph)$label <- paste0(data$name,"\n\n\n",n2)
plot(graph, layout=layout.circle, edge.arrow.size = 0.2, main = "my_graph")

Is it somehow possible to convert the above graph into a "visnetwork" graph, so that it looks like this?

I know there is a function ( visIgraph() ) meant for converting "igraph" graps to "visnetwork" graphs: https://www.rdocumentation.org/packages/visNetwork/versions/2.1.0/topics/visNetwork-igraph

But I am not sure if I can transform the first "igraph" graph (with both "numeric" and "text" labels) into an interactive "visnetwork" graph.

I tried to do this with the following code :

visIgraph(graph)

But this creates an interactive graph without the "number" labels.

Is it possible to do this?

Thank you!

Answer 1

You have to do a bit of manipulation to make this work because this uses base R plotting.

Essentially, these are two different igraph objects lying on top of each other. This is the only way I could think of to have two different 'cex' sizes. It may require a bit of finesse, depending on where you go from here.

library(tidyverse)
library(igraph)
library(gridGraphics)    # <--- I'm new!
library(grid)            # <--- I'm new!

#----------- from question -----------
set.seed(123)
n=15
data = data.frame(tibble(d = paste(1:n)))    

relations = data.frame(tibble(
  from = sample(data$d),
  to = lead(from, default=from[1]),
))

data$name = c("new york", "chicago", "los angeles", "orlando",
              "houston", "seattle", "washington", "baltimore", 
              "atlanta", "las vegas", "oakland", "phoenix", 
              "kansas", "miami", "newark" )


graph = graph_from_data_frame(relations, 
                              directed=T, 
                              vertices = data) 
(edge_fac <- forcats::as_factor(get.edgelist(graph)[,1]))
n2 <- as.integer(factor(data$name,levels = levels(edge_fac)))
V(graph)$color <- ifelse(data$d == relations$from[1], 
                         "red", "orange")

This is where the changes begin.

#---------- prepare the first plot -----------
# make label text larger
V(graph)$label.cex = 1.5
# V(graph)$label <- paste0(data$name,"\n",n2)
V(graph)$label <- paste0(n2) # just the number instead

#---------- prepare to collect grob ----------
# collect base plot grob
grabber <- function(){
  grid.echo()
  grid.grab()
}

# create a copy for the top layer
graph2 <- graph

#-------------- plot and grab ----------------
# without arrow sizes
plot(graph, layout=layout.circle, main = "my_graph")

# grab the grob
g1 = grabber() 

Now for the second graph; the top layer

#----------- create the top layer -------------
# with the copy, make the vertices transparent
V(graph2)$color <- "transparent"

# reset the font size
V(graph2)$label.cex = 1 

# shift the labels below (while keeping the plot design the same)
V(graph2)$label <- paste0("\n\n\n\n", data$name)

# show me
plot(graph2, layout=layout.circle,
     main = "my_graph", 
     edge.color = "transparent") # invisible arrows/ only 1 layer of arrows

# grab the grob
g2 = grabber()

Layer them!

#-------------- redraw the plots -------------
# make the plot background transparent on the top layer
g2[["children"]][["graphics-background"]][["gp"]][["fill"]] <- "transparent"

# draw it!
grid.draw(g1)
grid.draw(g2)

You might find it interesting that the graphs going into the grob look different than what comes out of them...grid essentially adjusts them. I thought that was kind of awesome.