Apply analysis methods exclusively for Pareto Front

Hey everyone!

I have a fairly simple goal, which involves reproducing Navarro-Mateu et al. (2018) method to select the best solutions from the Pareto Front. To do so, I want to apply the Relative Difference and Average of Fitness Ranks analysis methods only to non-dominated solutions. However, if I understood it correctly, Wallacei X component only enables to carry those analyses on the whole population.

What are your thoughts on that? I figured that maybe it could be resolved by exporting the whole pareto front solutions, and use their branches to filter what fitness values should serve as input for the Parallel coordinate plot analysis component.

Oh, and since we’re talking about selection methods, I couldn’t find any articles that applied the K-means clustering algorithm to select optimal solutions of MOOA. Do anyone have good references on that?

By the way, what an elegant and robust tool. Congrats!

Hi Gustavo,

I figured that maybe it could be resolved by exporting the whole pareto front solutions, and use their branches to filter what fitness values should serve as input for the Parallel coordinate plot analysis component.

You’re exactly right. Here is a quick definition (using the Wallacei X example file) that does just that. FYI, when you open the definition, you’ll have to export solutions from within Wallacei X first.

Wallacei X_Example File_visualisaing PCP for Pareto Fronts.gh (391.5 KB)

1 Like