Skip to content

ToP configuration

Configure top through vamos.algorithms.ToPConfig. This page is the parameter reference, not a tutorial about the algorithm; see ToP for the algorithm, its two phases and its budget rule.

New in VAMOS 1.2.0. The 1.0 compatibility baseline does not cover ToP and ToPConfig yet.

from vamos.algorithms import ToPConfig

Configuration overview · Algorithm contracts · Optimization

ToPConfig dataclass

Bases: _SerializableConfig

Configuration of ToP (Liu and Wang, 2019), with NSGA-II in its second phase.

pop_size is the size of the population, at least 1. The variation is ToP's differential evolution for real variables: the crossover must be ("de", {"cr": ..., "f": ...}), whose cr and f are the crossover rate and the scale factor of both phases, and the mutation is applied to every trial vector. Phase 1 lasts while (delta >= delta_threshold or Pf <= feasible_ratio_threshold) and the evaluations done are not more than phase_one_evaluation_fraction of the budget.

default(pop_size=None, n_var=None) classmethod

Create the default ToP configuration: PlatEMO's, which jMetal's port also uses.

pop_size=None means 100. The differential evolution has CR = 1.0 and F = 0.5, and the polynomial mutation probability 1/n and distribution index 20. The switch thresholds are 0.2 for delta, 1/3 for Pf and 90% of the budget.

Builder methods

Use ToPConfig.builder() to obtain the fluent builder; do not import the private builder class directly. The methods below are generated from that returned object, so their signatures are not maintained as a second manual parameter table.

Bases: _ConfigBuilderState, _PopSizeBuilder, _CrossoverBuilder, _MutationBuilder, _RepairBuilder, _InitializerBuilder, _ConstraintModeBuilder, _ResultArchiveBuilder

Fluent builder for ToP configs.

Examples: cfg = ToPConfig.default(n_var=10) cfg = ToPConfig.builder().pop_size(100).crossover("de", cr=1.0, f=0.5).mutation("pm", prob="1/n", eta=20.0).build()

delta_threshold(value)

Phase 1 goes on while delta is at least this value (or Pf is not above its threshold); 0.2 by default.

feasible_ratio_threshold(value)

Phase 1 goes on while Pf is not above this value (or delta is at least its threshold); 1/3 by default.

phase_one_evaluation_fraction(value)

Phase 1 ends when the evaluations done exceed this fraction of the budget; 0.9 by default.