pymoo/unseeded-run
| Rule | unseeded-run |
|---|---|
| Pack | pymoo |
| Severity | info |
| Fix | none, edit by hand |
| Library versions | every version |
minimize without seed starts from a new random state on every run, so two runs return different fronts. Pass seed.
res = minimize(problem, NSGA2(pop_size=100), ("n_gen", 200))res = minimize(problem, NSGA2(pop_size=100), ("n_gen", 200), seed=1)What the rule reports
Section titled “What the rule reports”The rule reports minimize without seed. Two runs of NSGA2 on ZDT1 then returned different fronts, and two runs with seed=1 returned the same front. The rule leaves alone a call that takes **kwargs, a call whose algorithm was built with seed, and a file that calls np.random.seed, which numpy-seed-ignored reports.
How to fix it
Section titled “How to fix it”xpd has no automatic fix, so make the change by hand.
Message
Section titled “Message”xpd prints this message with the issue. A name that starts with $ stands for code from your file.
minimize without seed starts from a new random state on every run, and two runs returned different fronts. Pass seed, such as seed=1, to get the same result again.Browse the other rules of the pymoo pack.