Very cool, Renzo! I'll look at it more closely.
On Feb 12, 2015, at 10:02 PM, Renzo Massobrio <[log in to unmask]> wrote:
> I am a newcomer at ECJ. I am currently using it for some multiobjective optimization, and for that purpose I wrote a simple Python code to calculate some common metrics used in MO problems with regards to the pareto front i.e. generational distance, spacing, spread, relative hypervolume.
> The code loads the fronts given by ECJ after multiple runs and compute the global pareto front for all the executions, plots it, and calculate the metrics.
> The source is available at: https://github.com/renzomassobrio/pf_metrics
> Please feel free to use it and change it to fit your needs. You are welcome to report bugs as well.
> Hope is useful for someone :)
> Best regards,