pymoo/reused-algorithm-without-copy
| Rule | reused-algorithm-without-copy |
|---|---|
| Pack | pymoo |
| Severity | error |
| Fix | none, edit by hand |
| Library versions | pymoo 0.6 and later |
With copy_algorithm=False, a second minimize on the same algorithm does not set it up again and returns the earlier result. Build a new algorithm for each run, or drop copy_algorithm=False.
algorithm = NSGA2(pop_size=100)for problem in problems: algorithm = NSGA2(pop_size=100) results.append(minimize(problem, algorithm, ("n_gen", 200), seed=1, copy_algorithm=False))What the rule reports
Section titled “What the rule reports”The rule reports a second minimize on the same algorithm when both calls pass copy_algorithm=False, in one block or in a loop that never binds the name. With copy_algorithm=False, minimize sets the algorithm up only when it has no problem yet. A second call with another problem, ("n_gen", 30) and another seed returned the F of the first call, still at generation 11 and still on the first problem. A loop over the algorithms, as in for algorithm in [NSGA2(), SMSEMOA()], gives each pass another object and is left alone.
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.
$ALG already ran in an earlier minimize with copy_algorithm=False, so this call does not set it up again and returns the earlier result. A second call with another problem, ("n_gen", 30) and another seed returned the F of the first call, still at generation 11. Build a new algorithm for this run, or drop copy_algorithm=False.Browse the other rules of the pymoo pack.