pymoo/mutation-prob-per-individual
| Rule | mutation-prob-per-individual |
|---|---|
| Pack | pymoo |
| Severity | warning |
| Fix | unsafe, with xpd check --fix --unsafe-fixes |
| Library versions | pymoo 0.6 and later |
Since pymoo 0.6.0, prob is the share of individuals that mutate, and prob_var is the chance for each variable, so an old prob=1/n mutates far less. Write prob_var= when you mean the chance per variable.
mutation = PM(prob=1 / problem.n_var, eta=20)mutation = PM(prob_var=1 / problem.n_var, eta=20)What the rule reports
Section titled “What the rule reports”The rule reports prob= on PM, PolynomialMutation, BitflipMutation, BFM, GaussianMutation and GM when the value is 1 divided by something, or a number of at most 0.2, and the call has no prob_var. Up to pymoo 0.5.0 prob was the chance that one variable changes. Since 0.6.0 it is the share of individuals that mutate, and prob_var is the chance per variable.
| 20,000 individuals of 30 variables | pymoo 0.5.0 | pymoo 0.6.2 |
|---|---|---|
individuals changed by prob=1/30 |
63 percent | 2 percent |
individuals changed by prob_var=1/30 |
not run | 57 percent |
How to fix it
Section titled “How to fix it”xpd check --fix --unsafe-fixes applies the fixes of this rule. Check the result, because an unsafe fix can change what the code does.
The unsafe fix writes prob_var=.
Message
Section titled “Message”xpd prints this message with the issue. A name that starts with $ stands for code from your file.
prob is the share of individuals that mutate since pymoo 0.6.0, and prob_var is the chance per variable. With prob=1/30 on 30 variables 2 percent of the individuals changed, where prob_var=1/30 changed 57 percent, as prob=1/30 did on pymoo 0.5.0. Write prob_var=$V if $V is the chance per variable.Browse the other rules of the pymoo pack.