plotHierarchicalEdgeBundles: Hierarchical Edge Bundling

In AEBilgrau/Bmisc: Biostatistical miscellaneous functions

Description

Visualization of networks using hierarchical edge bundles referenced below. Plots a graph using a hierachical tree as guides for the edges.

Usage

plotHierarchicalEdgeBundles(phylo, graph, beta = 0.8,
  include.mrca = FALSE, simplify = FALSE, ..., args.lines = list(),
  args.points = list(pch = 16, cex = 0.1), debug = FALSE, e.cols,
  v.use.only, e.use.only)

Arguments

phylo	A phylo object defining the hierarchy.
graph	A igraph object to be drawn.
beta	A number between 0 and 1 controlling the bundling strength.
include.mrca	Should the only the most recent common ancestor be used in the shortest path used for the splines?
simplify	Simplify the paths by taking the convex hull of the control points. Can sometimes yield better results.
...	Arguments passed to plot.phylo.
args.lines	A list of arguments passed to lines.
args.points	A list of arguments passed to points.
debug	Plot some extra debug info.
e.cols	A character vector giving the colors of the edges. Overrides any use of args.lines$col.
v.use.only	An integer vector giving the nodes from which edges are to be drawn. E.g. use.only = 1 will only plot the edges from vertex 1.
e.use.only	An integer vector giving the edges to be drawn. E.g. use.only = 1 will only draw the first edge.

Value

Plots to a new device.

Author(s)

Anders Ellern Bilgrau <anders.ellern.bilgrau (at) gmail.com>

References

Holten, Danny. "Hierarchical edge bundles: Visualization of adjacency relations in hierarchical data." Visualization and Computer Graphics, IEEE Transactions on 12.5 (2006): 741-748.

See Also

See plot.phylo.

Examples

library("igraph")
library("ape")
n <- 10
  
# Create a graph
corr <- cor(matrix(rnorm(n^2), n, n))
rownames(corr) <- colnames(corr) <- 
  apply(combn(LETTERS, 2), 2, paste0, collapse = "")[1:n]
adj <- abs(corr)
graph <- graph.adjacency(adj, mode = "un", weighted = TRUE, diag = FALSE)
E(graph)$color <- ifelse(get.lower.tri(corr) < 0, "blue", "red")

phylo <- as.phylo(hclust(as.dist(1 - adj), method = "ward.D"))
plotHierarchicalEdgeBundles(phylo, graph, type = "fan",
                            e.cols = E(graph)$color)

par(mfrow = c(1, 3), mar = c(0, 0, 0, 0))
plot(phylo, type = "fan")
plotHierarchicalEdgeBundles(phylo, graph, type = "fan", beta = 0.95,
                            args.lines = list(col = alp("steelblue", 90)))
plotHierarchicalEdgeBundles(phylo, graph, type = "fan", beta = 0.85,
                            args.lines = list(col = alp("steelblue", 90)))
             
# Extra control of plotting and debugging               
par(mfrow = c(1,2))
plot(phylo, type = "unrooted")        
plotHierarchicalEdgeBundles(phylo, graph, type = "unrooted", beta = 0.8,
                            v.use.only = 1, debug = FALSE,
                            args.lines = list(col = alp("red", 1), lwd = 2))

AEBilgrau/Bmisc documentation built on May 5, 2019, 11:28 a.m.