> ## Documentation Index > Fetch the complete documentation index at: https://proto.evodesign.org/docs/llms.txt > Use this file to discover all available pages before exploring further. # Genetic Algorithm Optimizer > Maintains a population of discrete sequences, generates offspring by crossover and mutation, scores them with constraints, and keeps the lowest-energy candidates.
Genetic Algorithm Optimizer

This optimizer is open source. Any third-party models, product names, or trademarks referenced are the property of their respective owners, and Proto is not affiliated with them.

Source
evo-design/proto-language/proto\_language/optimizer/genetic\_algorithm\_optimizer.py
View source
Optimizer contributors
Population-based genetic algorithm optimizer. ## API Reference
ConfigGeneticAlgorithmOptimizerConfig Source
Configuration for a general genetic algorithm optimizer. Number of genetic algorithm generations. Number of top-scoring candidates to retain as final results. Overrides program count. Number of candidates maintained in the population. Number of children scored per generation. Defaults to population\_size. Fraction of the best parents copied into the next generation before selecting children. Probability that an offspring recombines two parents instead of copying one parent. Crossover operator used for equal-length parent sequences. Options: `single_point`, `two_point`, `uniform` Parent selection strategy. Options: `tournament`, `rank`, `roulette` Use independent parent draws or one shared tournament that returns the winner and runner-up. Options: `independent`, `shared_tournament` Number of candidates sampled for tournament parent selection. Probability of accepting the current best tournament candidate before considering the next. If true, resample parent pairs until they use different population indices. Single creates one child per parent pair; reciprocal creates two children with swapped parents. Options: `single`, `reciprocal` Elitist keeps the best parents and children; generational keeps elites plus top children. Options: `elitist`, `generational` Select survivors by scalar energy or NSGA-II Pareto rank over scoring constraints. Options: `energy`, `nsga2` Zero-based per-segment sequence positions eligible for crossover. Zero-based per-segment sequence positions ineligible for crossover. If true, two-point crossover may select identical cut points and swap no residues. Keep parent structures on crossed-over children so downstream generators can use them. Random seed for reproducible optimization, generator, and constraint tool streams. Save history and log progress every N steps. Step 0 and final step always saved. Save granular per-proposal results (accept/reject) in history snapshots. Emit per-step debug information about proposals, scores, and acceptance through the logger.
## Usage ```python python icon="python" theme={null} >>> config = GeneticAlgorithmOptimizerConfig(num_generations=5, population_size=16) >>> optimizer_config_key = config.parent_selection >>> optimizer_config_key 'tournament' ``` ## Metadata | Property | Value | | ------------------------ | --------------------------- | | Key | `genetic-algorithm` | | Class | `GeneticAlgorithmOptimizer` | | Targets Single Segment | `False` | | Uses GPU | `False` | | Required Constraint Mode | `discrete` |