Sketchy, Hand-drawn-like Networks in R

schochastics

2 years ago

[This article was first published on schochastics, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)

Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

This post introduces the R package rougnet, which wraps the java script library rough.js to draw sketchy, hand-drawn-like networks. From all my network packages, this is probably the most useless but most fun one I developed.

The package is so far only available on github, but it already has a fitting pkgdown page.

#install.packages("remotes")
remotes::install_github("schochastics/roughnet")

The package only works with {{igraph}} objects.

library(igraph)
library(roughnet)

The only real function of the package is roughnet() which draws an igraph object as a sketchy network. Without specifying any aesthetics, the function uses default shapes and colours.

g <- make_graph("Frucht")
roughnet(g,width = 600,height = 400)

The function recognizes the following vertex attributes to customize the visualization:

x x-coordinate of vertex
y y-coordinate of vertex
shape one of “circle”, “rectangle”, (novelty shapes: “heart”, “air”, “earth”, “fire”, “water”)
fill vertex fill color
color vertex stroke color
stroke stroke size
fillstyle one of “hachure”, “solid”, “zigzag”, “cross-hatch”, “dots”, “dashed”, “zigzag-line”
size vertex size
label vertex label
pos position of vertex label (c)enter, (n)orth, (e)ast, (s)outh, (w)est

V(g)$shape  <- "rectangle"
V(g)$fill   <- "#e94560"
V(g)$color  <- "#0f3460"
V(g)$stroke <- 2
V(g)$fillstyle <- "zigzag"
V(g)$size <- 40

roughnet(g,width = 600,height = 400)

For edges there are only two supported attributes:

color color of edge
width width of edge

E(g)$color <- "#e94560"
E(g)$width <- 2

roughnet(g,width = 600,height = 400)

Below is a more “realistic” example using the infamous karate network.

g <- make_graph("Zachary")
clu <- membership(cluster_louvain(g))
V(g)$shape <- "circle"
V(g)$shape[c(1,34)] <- "rectangle"
V(g)$fill <- c("#E41A1C", "#377EB8", "#4DAF4A", "#984EA3")[clu]
V(g)$fillstyle <- c("hachure", "zigzag", "cross-hatch", "dots")[clu]
V(g)$color <- "black"
V(g)$size <- 30
V(g)$stroke <- 2
E(g)$color <- "#AEAEAE"
roughnet(g,width = 960,height = 600)

Node labels and s

Node labels can also be placed, but unfortunately it is a bit more tedious if you do not simply want to put them in the center (when, say using some novelty node shapes.)

In the example below, we use the Allies/Enemies relations in “Avatar: The Last Airbender”, available in the signnet package to illustrate all features of roughnet, including node labels.

By default, node labels are placed in the center of the node, but can be placed (n)orth, (e)ast, (s)outh, or (w)est of the node.

# install.packages(c("signnet","graphlayouts"))
library(signnet)
library(graphlayouts)
library(dplyr)

data("avatar") # Allies/Enemies relations in Avatar: The Last Airbender

main <- induced_subgraph(avatar,which(V(avatar)$main)) #only use the main characters

#calculate layout
w <- ifelse(E(main)$sign==1,3,1)
xy <- layout_with_stress(main,weights = 1/w)

V(main)$x <- xy[,1]
V(main)$y <- xy[,2]

V(main)$fill <- case_when(V(main)$affiliation=="earth kingdom"~"#8B6914",
                          V(main)$affiliation=="fire nation"~"#CD2626",
                          V(main)$affiliation=="water tribe"~"white",
                          V(main)$affiliation=="air nomad"~"#98F5FF",
                       TRUE~"grey"
)

V(main)$color <- case_when(V(main)$affiliation=="earth kingdom"~"#8B6914",
                          V(main)$affiliation=="fire nation"~"#CD2626",
                          V(main)$affiliation=="water tribe"~"#1874CD",
                          V(main)$affiliation=="air nomad"~"#98F5FF",
                          TRUE~"grey"
)

V(main)$shape <- case_when(V(main)$affiliation=="earth kingdom"~"earth",
                           V(main)$affiliation=="fire nation"~"fire",
                           V(main)$affiliation=="water tribe"~"water",
                           V(main)$affiliation=="air nomad"~"air",
                           TRUE~"circle"
)


E(main)$width <- ifelse(E(main)$sign==1,1.2,0.3)
E(main)$color <- ifelse(E(main)$sign==1,"#228B22","#CD3278")
V(main)$label <- V(main)$name

# position labels in the (c)enter of the vertex 
# or (n)orth, (e)ast, (s)outh, or (w)est of it 
V(main)$pos <- c("n","s","s","s","s","e","n","e","s","n") 

roughnet(main, width = 600, height = 600,  = "30px Herculanum")

Save plots

Plots can be saved with the function save_roughnet() , which needs {{pagedown}} to be installed.

To leave a comment for the author, please follow the link and comment on their blog: schochastics.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.

Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.