hpp_benchmark

A set of benchmark to track the evolution of performances of HPP.

Running the C++ benchmarks

Compile target benchmarks and run

benchmarks --output {source_dir}/results/cpp/{year}/{month}/{day} --label <version>

Option --output <dir> can be omitted if you do not want to save the results to the disk. See benchmarks --help for more usage.

Save the results

The script will generate a bunch of csv files in the output directory (see --output). To save the benchmarks, commit the output directory.

Running the Python benchmarks

 ./generate_benchmark.py  new_directory_name

The format of the argument new_directory_name is: year-month-day This program creates a copy of the directory future into a new directory which contains the benchmarks.

In each sub-directories of this copy of future, it will run one after another the python scripts (script.py). 2 windows will be opened: a window with hppcorbaserver and another window where the script.py is running.

When the execution of a script is done, the python's window is closed and the output will be written in a file named benchmark in the directory corresponding to the robot's name, next to the script.

The hppcorbaserver window still open and the next script is automatically executed when the previous is done. The name of script which is running is written in the terminal.