Metrics
Bench-MR includes various metrics on which the planning algorithms can be evaluated.
The following statistics are gathered for each benchmark in Bench-MR:
| Metric Name | Metric Title |
|---|---|
max_curvature | Maximum Curvature |
normalized_curvature | Normalized Curvature |
aol | Angle-over-length (AOL), an alternative measure of smoothness |
max_clearing_distance | Maximum Clearing |
mean_clearing_distance | Mean Clearing |
median_clearing_distance | Median Clearing |
min_clearing_distance | Minimum Clearing |
path_length | Path Length |
smoothness | Smoothness |
planning_time | Computation Time |
cusps | Cusps |
aggregate | Aggregate |
This mapping is defined in the stat_names dictionary in python/definitions.py.
Visualize Statistics
Use the MPB.plot_planner_stats function with the optional metrics keyword to provide a string of comma-separated metric names to plot:
Example
mpb = MPB()
mpb.set_corridor_grid_env(radius = 3)
mpb.set_planners(['rrt', 'rrt_star', 'informed_rrt_star'])
mpb.set_steer_functions(['reeds_shepp'])
mpb.run(id='test_run', runs=3);
mpb.plot_planner_stats(metrics=", ".join(stat_names.keys()))
Computation Time Drill-Down
Examine the computation times within the different phases of planning:
mpb.plot_planner_timings()
 |  |  |