igraph plotting: vertices too close in Kamada-Kawai layout

Question

I'm plotting a directed network from an edge list, and have so far created a tree-like plot (see here).

It looks good, however all of the nodes are too close together. I would like to keep the shape of it while spreading out the nodes more. Here's the code that got me the image above:

library(igraph)
ref <- read.csv("my-ref.csv", as.is=T)
el <- graph.data.frame(ref, directed=T)
lay.kk <- layout.kamada.kawai(el, niter=1000, kkconst=50)
plot.igraph(el, lay=lay.kk, vertex.label=NA, vertex.size=2, vertex.color="black")

I've tried messing around with kkconst, but that doesn't seem to change anything. Any tips are greatly appreciated!

It would be helpful if you provided a [reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) with sample input data. — MrFlick, Jun 24 '15 at 19:12

score 3 · Answer 1 · answered Jun 24 '15 at 21:40

3

The Kamada-Kawai layout does not really work well for disconnected graphs because the disconnected components tend to "drift away" from each other. Since igraph scales the entire plot to fit within the canvas, the farther the components are from each other, the closer the nodes will be within the components. Try the Fruchterman-Reingold layout instead.

answered Jun 24 '15 at 21:40

Tamás

47,239
12
105
124

This makes a lot of sense! Thanks for the tip. – Kira Tebbe Jul 01 '15 at 16:34

score 3 · Answer 2 · answered Jun 25 '15 at 09:26

3

As Tamás suggested, you might get better results with layout.fruchterman.reingold(). You can fine tune this function with the following parameters:

require(igraph)
g <- erdoss.renyi.game(n = 100, p.or.m = 0.04)
lo <- layout.fruchterman.reingold(g, repulserad = vcount(g)^2.8, 
    area = vcount(g)^2.3, niter = 1000)
plot(g, layout = lo, vertex.size = 3, vertex.frame.color = NULL, 
    vertex.label.dist = 0.5, vertex.label.cex = 0.7, edge.width = 0.5)

These values resulted a low overlap, clear but compact layout for me. Try to change them a bit, to see their effect on the layout. Those parameters I set for plot() also help to make the visualization more clear.

answered Jun 25 '15 at 09:26

deeenes

4,148
5
43
59

The only problem I have with this layout is that you can't see the different components; it just plots them evenly within the plotting area. I want there to still be space between the disconnected components. Any idea if FR can do this well? – Kira Tebbe Jul 01 '15 at 16:42
maybe it helps, [here](http://stackoverflow.com/questions/16390221/how-to-make-grouped-layout-in-igraph) you can find two different approaches to achieve a layout emphasising groups: one operates with edge weights, other calculates a separate layout for each component, and after scales each sublayout to a square field of the plotting area – deeenes Jul 02 '15 at 20:12

score 0 · Answer 3 · edited May 23 '17 at 12:30

There is a similar question here to which I have posted the following answer

Option 1: make the vertices smaller

node.size= c(10,10,10)
plot(net, vertex.size=node.size*0.25)

Option 2 (in case the distances between the vertices are not important to you):

# Use the tkplot option to edit your graph in GUI
tkplot (net)

tkplot GUI will allow you to interactively change the layout types.

Note: tkplot outputs the graph as eps. If you want to edit it further or export it to pdf I suggest using inkscape (I use it for all my graph editing - just save the graph as pdf in RStudio and edit it in inkscape). For the case of eps if you are on a windows machine you will need to tweak inkscape to open this format. A very short and simple process which is detailed here:

igraph plotting: vertices too close in Kamada-Kawai layout

3 Answers3