Jido.Evolve.Config (Jido Evolve v1.0.0)

Canonical configuration structure for evolutionary algorithms.

This module uses Zoi for validation and provides sensible defaults.

Summary

Types

t()

Functions

elite_count(config)

Get the number of elite entities to preserve.

init_random_seed(config)

Initialize random seed if configured.

new(opts \\ [])

Create a new configuration with validation.

new!(opts \\ [])

Create a new configuration, raising on validation errors.

Types

t()

@type t() :: %Jido.Evolve.Config{
  checkpoint_interval: nil | nil | integer(),
  crossover_rate: number(),
  crossover_strategy: atom(),
  elitism_rate: number(),
  evaluation_timeout: integer() | :infinity,
  generations: integer(),
  max_concurrency: integer(),
  metrics_enabled: boolean(),
  mutation_rate: number(),
  mutation_strategy: atom(),
  population_size: integer(),
  random_seed: nil | nil | integer(),
  selection_pressure: number(),
  selection_strategy: atom(),
  termination_criteria: [
    max_generations: integer(),
    target_fitness: number(),
    no_improvement: integer()
  ],
  tournament_size: integer()
}

Functions

elite_count(config)

@spec elite_count(t()) :: non_neg_integer()

Get the number of elite entities to preserve.

init_random_seed(config)

@spec init_random_seed(t()) :: :ok

@spec init_random_seed(t()) :: :ok

Initialize random seed if configured.

Uses explicit :exs1024 algorithm with deterministic seed tuple for reproducibility across OTP versions.

new(opts \\ [])

@spec new(keyword() | map()) :: {:ok, t()} | {:error, term()}

Create a new configuration with validation.

Examples

iex> {:ok, config} = Jido.Evolve.Config.new(population_size: 50)
iex> config.population_size
50

iex> {:error, _error} = Jido.Evolve.Config.new(population_size: -1)

new!(opts \\ [])

@spec new!(keyword() | map()) :: t()

Create a new configuration, raising on validation errors.

Examples

iex> config = Jido.Evolve.Config.new!(population_size: 50)
iex> config.population_size
50