Closed rasmusmarak closed 1 year ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
To answer the second point https://github.com/rasmusmarak/67P-Swarm/issues/31 regarding the convergence rate, perhaps the best way forward is to either benchmark a set of smaller runs or perform one longer run (i.e with a larger number of generations/population).
Yes ideally run a longer to explore the optimum you can achieve and then run a few shorter ones to see how close they get. From the plot you posted it clearly hasn't converged, ideally post results in #31
Perfect, then I will run these tests during the upcoming night and post them in the issue.
Added a fitness over generation plot.
To answer the second point #31 regarding the convergence rate, perhaps the best way forward is to either benchmark a set of smaller runs or perform one longer run (i.e with a larger number of generations/population).