This is just a thread for adding in available tuning parameters, if you add something into the rig that has a tuning parameter, just add it to the list so everyone knows what it does:
epsilon: In model.py on line 122 is the epsilon value for cdom. This is the point at which we decide whether or not the difference is enough to consider a particle as either dominated or non-dominated. If our difference between loss values is less than epsilon we decide that the particle is non-dominated. 0.06 seems like a reasonable value for this.
phi_1 and phi_2: phi_1 is the amount you listen to the global and phi_2 is the amount that you listen to your personal best. These two values must equal 4 when added together. Could start passing these into test.py to test out a number of different potential phi values rather than simply resetting the defaults on line 36 of adaptiveGlobalPSO as I have been so far.
era: This controls how often we output sb and k. Could start to set this higher to cleanup output from the logger.
This is just a thread for adding in available tuning parameters, if you add something into the rig that has a tuning parameter, just add it to the list so everyone knows what it does: