Algorithm configuration¶

Choose the configuration class for the algorithm passed to optimize. These classes are exported through vamos.algorithms; their documented fields follow the 1.x stability policy.

Algorithm identifier	Public configuration
`nsgaii`	`NSGAIIConfig`
`nsgaiii`	`NSGAIIIConfig`
`moead`	`MOEADConfig`
`smsemoa`	`SMSEMOAConfig`
`spea2`	`SPEA2Config`
`ibea`	`IBEAConfig`
`smpso`	`SMPSOConfig`
`agemoea`	`AGEMOEAConfig`
`rvea`	`RVEAConfig`

Explicit configuration¶

from vamos import optimize
from vamos.algorithms import NSGAIIConfig
from vamos.problems import ZDT1

problem = ZDT1(n_var=30)
config = NSGAIIConfig.default(pop_size=40, n_var=problem.n_var)
result = optimize(
    problem,
    algorithm="nsgaii",
    algorithm_config=config,
    max_evaluations=400,
    seed=42,
)

Use a configuration that matches the selected algorithm. Read the generated signature on its page for defaults and construction methods; use Config.builder() rather than importing a private builder class.

Algorithms and backends explains result modes, archive behavior, capabilities, and probability shorthand such as "1/n". Optimization documents how the configuration is passed to the run.