Final Best Position Logical Error

[r-karunathilake]

PySwarms version:

Name: pyswarms 1.3.0
Python version: 3.11.0
Operating System: Windows 10

Description

After running global best and local best PSO I noticed that my best cost is NOT reproducible with the final position vector returned after the optimizer is done running.

I think there is a major issue with position return in 'optimize' method in global, local, binary and general optimizers.

final_best_cost = self.swarm.best_cost.copy()
final_best_pos = self.swarm.pbest_pos[self.swarm.pbest_cost.argmin()].copy()

From what I understand from reviewing the code, the 'final_best_pos' is only the best position vector of the last iteration of the swarm. This does NOT return the global best position found. I think it should simply be:

final_best_cost = self.swarm.best_cost.copy()
final_best_position = self.swarm.best_pos.copy()
Thoughts ?