High convergens using GRankSelector

11 views

Skip to first unread message

Saeed Soltani Moghadam

unread,

Jun 28, 2017, 8:20:33 AM6/28/17

to pyevolve

Hi Guys

As i know about the difference between Rank & RouletteWheel selections methods, when we used the first one we should see some slight convergence and high for the second one. But if you run the 1st example of the package and compare the result using pyevolve_graph.py [consider the files ex1_maxmin_fitness.png] you will see the vice versa! Anybody could explain why?

result using GRankselector:

result using GRouletteWheelselector:

Reply all

Reply to author

Forward

0 new messages