Skip to contents

A result from paths() records the optimality criterion it was found under, and such a result is drawn as a trajectory tree by plot_path_trajectories(). That is the right picture for an endpoint-local family: the routes need not share prefix-optimal subpaths, so a vertex reached under two different temporal histories appears twice rather than being forced into one predecessor tree.

Older serialised results carry no criterion. Those are drawn by the legacy predecessor-tree renderer, which only defines a tree when the sessions were collapsed; a bounded or separate-session result of that vintage raises dynet_unsupported_plot instead of implying a tree the criterion never promised.

Usage

# S3 method for class 'dynet_paths'
plot(x, palette = "okabe", ...)

Arguments

x

A dynet_paths from paths().

palette

Palette specification, as in plot.dynet(). Passed to plot_path_trajectories(), and for a legacy result colours vertices by how many hops they are from the source.

...

Passed to plot_path_trajectories(), or to cograph::splot() for a legacy result.

Value

A ggplot object for a result carrying criterion metadata, which is every result paths() returns. A legacy result without it is drawn on the current device by cograph::splot() and x is returned invisibly.

A legacy result raises dynet_unsupported_plot when its sessions were not collapsed, dynet_empty_result when the source reaches no other vertex, dynet_bad_palette for an unusable palette, and dynet_needs_cograph when cograph is not installed.

Examples

dn <- dynet(school_contacts)
journeys <- paths(dn, from = "Ana")
plot(journeys)