Two arguments influence how points are arranged in circleplot. First is the ‘cluster’ command, which interfaces with hclust() to determine the order in which points are arranged around the circumference of the plot. This is useful as it means that points are positioned closer together if they have greater connectedness. Second, the ‘reduce’ command simply deletes all points that have no connections to any other point. Both are set to TRUE by default.
A third argument affect whether points are drawn around the circumference of the plot, and if so, how they appear. This argument (called ‘style’) is set to ‘classic’ by default, but two other options (‘pie’ and ‘clock’) are also available.
Beyond these general methods, you can customise the plot by providing a list of plot attributes to the ‘plot.control’ argument. In particular, users can control how points are drawn on the vertex of the circleplot by passing a
plot.control$points. One column in this
data.frame should be named ‘label’, and should correspond to the labels of the input distance matrix [i.e.
attr(input, "Labels")], but besides that it can contain any information you like. Column names besides ‘labels’ should correspond to commands in
pch, cex, bg etc.), with each row giving the value of that command for a single point. The default is for each point to be represented by a filled grey circle, though if you are in a hurry and don’t mind categorical colour choices, a quick way to specify this is using the
Next: Setting line attributes