how to get a knee point solution from pareto front

Question

I am using a python package DEAP to run NSGA_II algorithm for multi-objective optimization. The out put a a set of pareto optimal solutions in (objective space and parameter space). My question is: How can I code a simple python code to get the knee point solution from pareto front, automatically. The pareto from is mostly convex.

Answer 1

I found a python package implement the Kneedle algorithm.

kn = KneeLocator(
    x,
    y,
    curve='convex',
    direction='decreasing',
    interp_method='interp1d',
)

print(kn.knee)