CODEWITHSUNDEEP
pythongenetic-algorithmevolutionary-algorithmgenetic-programmingpymoo
iammetehan
iammetehan

Reputation: 11

Couldn't get more than two solutions in pymoo

I have a problem with two objectives. The problem is minimized by NSGA-2 algorithm. I run minimize with different population sizes and different termination criterias, but every time the algorithm just gives only two solutions. I know from logs that there are some other proper solutions but I can't see it in pareto-front. How can I get more solutions? Do you have an idea?

Thank you...

I tried to increase population size, number of generations, termination criteria. I tried to change objective function. I tried to apply different mutation, selection and crossover functions. However, all of these resulted in just two solutions. I want to get more solutions, I know there are more solutions but the algorithm just gives me two.

Upvotes: 0

Views: 416

Answers (1)

Pieter-Jan
Pieter-Jan

Reputation: 540

To have an idea whether your problem has converged or not, you can plot the hypervolume. It shows how much your pareto-front is moving in function of number of evaluations. Documentation is here:

https://pymoo.org/getting_started/part_4.html?highlight=hypervolume

Secondly, are you sure your known solution is not being scrapped by any constraints? Is the range of design variables set so that they can find your optimum?

Upvotes: 0

Related Questions