Visualization of static and dynamic networks on R, part 3

In the first part :

• network visualization: why? how?
• visualization parameters
• best practices - aesthetics and performance
• data formats and preparation
• description of the data sets used in the examples
• getting started with igraph

In the second part : colors and fonts in graphs R.

In this part: parameters of graphs, vertices and edges.

Network Plotting

Building with igraph: network graphics have an extensive set of options. In particular, there are vertex settings (start with vertex. ) And edge settings (start with edge. ). Some of the settings are listed below, and more information can be obtained by running the ?igraph.plotting .

Build options in igraph (among others):
')

Vertices

vertex.color	vertex color
vertex.frame.color	vertex outline color
vertex.shape	vertex designation form, one of the values ​​“none”, “circle”, “square”, “csquare”, “rectangle”, “crectangle”, “vrectangle”, “pie”, “raster”, “sphere”
vertex.size	vertex size (default is 15)
vertex.size2	second parameter of vertex size (for example, for a rectangle)
vertex.label	vector of symbols for vertices
vertex.label.family	font family for vertex labels (for example, Times, Helvetica)
vertex.label.font	font: 1 - regular, 2 - bold, 3 - italic, 4 - bold italic, 5 - character
vertex.label.cex	font size (multiplier, device dependent)
vertex.label.dist	the distance between the mark and the top
vertex.label.degree	location of the label in relation to the vertex: 0 - on the right, "pi" - on the left, "pi / 2" - on the bottom, "-pi / 2" - on the top

Ribs

edge.color	edge color
edge.width	edge width default 1
edge.arrow.size	default arrow size 1
edge.arrow.width	arrow width, default 1
edge.lty	line type, either 0 or “blank”, 1 or “solid”, 2 or “dashed” (discontinuous), 3 or “dotted” (dash), 4 or “dotdash” (dot-dash) ), 5 or “longdash” (long dash), 6 or “twodash” (double dash)
edge.label	edge symbol vector
edge.label.family	font family for edge labels (for example, Times, Helvetica)
edge.label.font	font: 1 - regular, 2 - bold, 3 - italic, 4 - bold italic, 5 - character
edge.label.cex	font size of edge marks
edge.curved	rib bend, in the range from 0 to 1 (FALSE sets to 0, TRUE to 0.5)
arrow.mode	a vector that determines whether the edges should have arrows, possible values: 0 - no arrows, 1 - reverse, 2 - straight, 3 - in both directions

Other

margin	empty borders around the graph, vector of length 4
frame	if TRUE, the graph will be framed
main	if set, adds a caption to the graphic
sub	if set, adds a subtitle to the graphic

Settings for vertices and edges can be set in two ways. The first is to set them using the plot() function, as shown below.

 #     (edge.curved=.1)    : plot(net, edge.arrow.size=.4, edge.curved=.1) 

 #        #     ,   "media" plot(net, edge.arrow.size=.2, edge.color="orange", vertex.color="orange", vertex.frame.color="#ffffff", vertex.label=V(net)$media, vertex.label.color="black") 

Another way to set attributes is to add them to the igraph object. Let's say we want the color of the vertices of the network to depend on the media, and the size depends on the degree of centrality (more links - more vertices). We will also change the width of the ribs depending on their weight.

 #        : colrs <- c("gray50", "tomato", "gold") V(net)$color <- colrs[V(net)$media.type] #    ( )       : deg <- degree(net, mode="all") V(net)$size <- deg*3 #      : V(net)$size <- V(net)$audience.size*0.6 #   -  . #     NA,   : V(net)$label <- NA #      : E(net)$width <- E(net)$weight/6 #      : E(net)$arrow.size <- .2 E(net)$edge.color <- "gray80" E(net)$width <- 1+E(net)$weight/12 plot(net) 

You can also override attributes directly in the graphic:

 plot(net, edge.color="orange", vertex.color="gray50") 

It is useful to add a legend explaining the meaning of the colors used:

 plot(net) legend(x=-1.5, y=-1.1, c("Newspaper","Television", "Online News"), pch=21, col="#777777", pt.bg=colrs, pt.cex=2, cex=.8, bty="n", ncol=1) 

Sometimes, especially in semantic networks, you only need to display vertex labels:

 plot(net, vertex.shape="none", vertex.label=V(net)$media, vertex.label.font=2, vertex.label.color="gray40", vertex.label.cex=.7, edge.color="gray85") 

Let's color the edges of the graph in the colors of the vertices from which they emanate. You can find the source vertex of each edge using the get.edges function from igraph.

 edge.start <- get.edges(net, 1:ecount(net))[,1] edge.col <- V(net)$color[edge.start] plot(net, edge.color=edge.col, edge.curved=.1)