Introduction

The ggnet2 function is a visualization function to plot network objects as ggplot2 objects. It accepts any object that can be coerced to the network class, including adjacency or incidence matrices, edge lists, or one-mode igraph network objects.

Rationale

R already provides many ways to plot static and dynamic networks, many of which are detailed in a beautiful tutorial by Katherine Ognyanova.

Furthermore, R can

  • control external network visualization libraries, using tools such as RNeo4j;
  • export network objects to external graph formats, using tools such as ndtv, networkD3 or rgexf; and
  • plot geographic networks, using spatial functions or the dedicated spnet package.

All of these tools, however, require to use a new graph syntax, either within or outside of R, in order to create new network objects with the appropriate properties for plotting.

Instead, for the many users who are familiar with the ggplot2 package, it might be interesting to use a syntax that comes close to its “grammar of graphics” to process and plot network data, in the same format as was used for network analysis.

This idea motivated the very first version of ggnet, by Moritz Marbach, and is also motivating the development of geom_net, a geom object for network data structured as data frames, by Sam Tyner and Heike Hofmann.

ggnet2 is an improved version of ggnet. Both functions are available from the GGally package or as standalone functions. ggnet2 brings several improvements that convey additional control over all plotting parameters.

Installation

ggnet2 is available through the GGally package:

install.packages("GGally")
library(GGally)

Or it can also be installed from its standalone package:

devtools::install_github("briatte/ggnet")
library(ggnet)

Dependencies

The package dependencies of ggnet2 are, on the one hand, the network and sna packages for network manipulation, and the ggplot2 package for plot construction.

library(network)
library(sna)
library(ggplot2)

The ggplot2 package will also load the scales package, which is used internally by ggnet2.

Additionally, ggnet2 suggests the following packages:

  • If the RColorBrewer package is installed, ggnet2 will be able to use ColorBrewer palettes to color network nodes.
  • If the intergraph package is installed, ggnet2 will be able to process one-mode networks objects created with the igraph package.

All packages cited above can be installed from CRAN through install.packages.

Notation

In this vignette, “nodes” designate the vertices of a network, and “edges” designate its ties. Readers who are not familiar with network terminology might want to consult a handbook such as Networks. An Introduction, by Mark Newman.

Contents

Most of this vignette is organized around two simple network examples:

  • node-level plotting parameters are demonstrated on a one-mode network, and
  • edge-level plotting parameters are demonstrated on a two-mode (bipartite) network.

The vignette also contains a section that illustrates some additional capabilities of ggnet2, and another section showing two additional examples of real-world networks plotted with ggnet2. It closes on known limitations of ggnet2.

Example (1): Random graph

Let’s start with an undirected Bernoulli random graph, with 10 nodes named “a, b, …, i, j”, and a rather high likelihood of an edge to exist between them:

# random graph
net = rgraph(10, mode = "graph", tprob = 0.5)
net = network(net, directed = FALSE)

# vertex names
network.vertex.names(net) = letters[1:10]

This graph can be visualized with ggnet2 without any further work:

ggnet2(net)

The net argument is the only compulsory argument of ggnet2. It can be a network object or any object that can be coerced to that class through its edgeset.constructors functions, such as adjacency matrixes, incidence matrixes and edge lists.

If the intergraph package is installed, net can also be an igraph one-mode network object, which is the only type of network that the package can convert from the igraph to the network class.

Node color and size

The most basic properties that one might want to change at that stage are the size and color of the nodes, or the size and color of the edges. Let’s modify each of these properties:

ggnet2(net, node.size = 6, node.color = "black", edge.size = 1, edge.color = "grey")

The vertex-related arguments of ggnet2 start with node, and its edge-related arguments start with edge. The node.color and node.size arguments can be abbreviated:

ggnet2(net, size = 6, color = "black", edge.size = 1, edge.color = "grey")

It also possible to pass a vector of node colors directly to ggnet2, as long as it has the same number of elements as the network has nodes:

ggnet2(net, size = 6, color = rep(c("tomato", "steelblue"), 5))