Learn R Programming

ndtv (version 0.13.4)

timeline: Plot a timeline for the edge and vertex spells of a network

Description

Produces a `phase plot' or timeline showing the durations of the activity spells in a networkDynamic object. Spells are traced out horizontally, with all the activity for each element (vertex or edge) in a single row.

Usage

timeline(x, v = seq_len(network.size(x)), e = seq_along(x$mel), 
        plot.vertex.spells = TRUE, plot.edge.spells = TRUE,
        slice.par = NULL,
        displaylabels = TRUE, e.label=TRUE, e.label.col='purple',
        edge.lwd=1,
        v.label, v.label.col='blue',
        vertex.cex=1, cex, adj=0,
        edge.col = rgb(0.5, 0.2, 0.2, 0.5), 
        vertex.col = rgb(0.2, 0.2, 0.5, 0.5), 
        xlab, ylab, xlim, ylim, ...)

Value

A plot is produced.

Arguments

x

a networkDynamic object that will have its spells plotted.

v

numeric vector of vertex ids to include

e

numeric vector of edge ids to include

plot.vertex.spells

logical, should vertex spells be plotted?

plot.edge.spells

logical, should edge spells be plotted?

slice.par

(optional) `slice.par' list giving network binning parameters. If included, rectangles corresponding to each bin will be plotted over the spells to indicate which spell will land in bins. The bins will be drawn with slightly darker left edge more transparent right edge to evoke the effect of a right-open interval.

displaylabels

logical, should labels be drawn for each spell

e.label

character vector of edge labels or edge attribute name. Default is edge.id

e.label.col

color name or character vector of colors for edge labels (or name of edge attribute to provide them)

v.label

character vector of vertex labels or vertex attribute name. Default is network.vertex.names

v.label.col

color name or character vector of colors for vertex labels (or name of vertex attribute to provide them)

vertex.cex

numeric scaling factor, vector of numeric scaling factors or attribute name. Translated width of line (lwd) corresponding to each vertex.

edge.lwd

numeric scale factor, numeric vector, or character edge attribute name providing a numeric value for the width of the lines corresponding to each edge. Note that this does not behave exactly as edge.lwd in plot.network as it does not perform scaling based on attribute values.

cex

text size scaling for both vertex and edge labels (see plot.default)

adj

text justification parameter (see par) for both vertex and edge labels. Labels are positioned relative to onset of spell.

edge.col

color to be used to draw lines for edge spells, or vector of color names corresponding to edges, or name of edge attribute.

vertex.col

color to be used to draw lines for vertex spells, or vector of color names corresponding to vertices, or name of vertex attribute.

xlab

x-axis label for plot

ylab

y-axis label for plot

xlim

two-element numeric vector giving the x-range (time bounds) of the plot to show. Defaults to (non-infinite) max and min time of network.

ylim

two-element numeric vector giving the y-range (effectively the number of entities) of plot to show. Defaults to an appropriate mapping of the number of entities to the available plot size.

...

additional arguments to be passed to plot subroutines. See plot.default,lines,text.

Author

skyebend@uw.edu

Details

When the v argument is included, edges involving vertices not in v are excluded (but the reverse is not true for the e argument). If xlim range is provided and the spells corresponding to a vertex or an edge lie entirely outside its bounds they will not be shown.

Many of the arguments correspond to arguments in plot.network but are translated to the timeline plot context. For example, vertex.cex actually controls the lwd (line width) of the lines corresponding to vertex spells. The arguments are expanded using plotArgs.network so that they should give the expansion behavior and attribute look up as plot.network

Additional plotting arguments can be passed in to modify drawing. For example, lty for line style. Vertices and edges that are never active are not included on the plot.

See Also

See also plot.network.

Examples

Run this code
  data(stergm.sim.1)
  timeline(stergm.sim.1)
  
  # color vertices by priorates, don't show edges
  timeline(stergm.sim.1,vertex.col='priorates',plot.edge.spells=FALSE)
  
  # show only relationships among a few vertices
  timeline(stergm.sim.1,v=1:8)
  
  # zoom in on a region of time
  timeline(stergm.sim.1,xlim=c(20,40))
  
  # label vertices with numbers
  # and label edges by the tail and head vertices they link
  timeline(stergm.sim.1,xlim=c(0,5),v.label=1:network.size(stergm.sim.1),
      e.label=sapply(stergm.sim.1$mel,function(e){paste(e$inl,e$outl,sep='->')}) )
      
  # show only edge spells, hi-lite edge id 20
  set.edge.attribute(stergm.sim.1,'my_color','gray')
  set.edge.attribute(stergm.sim.1,'my_color','red',e=20)
  timeline(stergm.sim.1,edge.col='my_color',plot.vertex.spells=FALSE)
  
  # show binning over the edges
  timeline(stergm.sim.1,slice.par=list(start=0,
                                       end=100,
                                       interval=10, aggregate.dur=5,
                                       rule='latest'),
                      plot.vertex.spells=FALSE)

Run the code above in your browser using DataLab