Struct CartesianGeneticProgramming 

pub struct CartesianGeneticProgramming<B: Backend> { /* private fields */ }

Classical Cartesian GP with (1 + λ) ES.

Example

use burn::backend::Flex;
use rlevo_evolution::algorithms::gp_cgp::{CartesianGeneticProgramming, CgpConfig};

let strategy = CartesianGeneticProgramming::<Flex>::new();
let params = CgpConfig::default_for(1);
assert!(params.genome_len() > 0);
let _ = strategy;

Implementations

impl<B: Backend> CartesianGeneticProgramming<B>

pub fn new() -> Self

Builds a new (stateless) strategy object.

Trait Implementations

impl<B: Clone + Backend> Clone for CartesianGeneticProgramming<B>

fn clone(&self) -> CartesianGeneticProgramming<B>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl<B: Copy + Backend> Copy for CartesianGeneticProgramming<B>

impl<B: Debug + Backend> Debug for CartesianGeneticProgramming<B>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<B: Default + Backend> Default for CartesianGeneticProgramming<B>

fn default() -> CartesianGeneticProgramming<B>

Returns the “default value” for a type.

impl<B: Backend> Strategy<B> for CartesianGeneticProgramming<B>

where B::Device: Clone,

fn init( &self, params: &CgpConfig, rng: &mut dyn Rng, device: &<B as BackendTypes>::Device, ) -> CgpState<B>

Samples the initial parent genome by drawing random node functions and feed-forward input connections via rng, then uploads the genotype as a (1, genome_len) integer tensor.

fn ask( &self, params: &CgpConfig, state: &CgpState<B>, rng: &mut dyn Rng, device: &<B as BackendTypes>::Device, ) -> (Tensor<B, 2, Int>, CgpState<B>)

Returns the offspring population for the current generation.

On the first call (parent fitness not yet set), returns the single parent genome unchanged for initial fitness evaluation. On subsequent calls, produces params.lambda children by cloning the parent and applying per-gene point mutation, with mutation draws taken from a deterministic seed_stream (host-RNG convention).

fn tell( &self, _params: &CgpConfig, offspring: Tensor<B, 2, Int>, fitness: Tensor<B, 1>, state: CgpState<B>, _rng: &mut dyn Rng, ) -> (CgpState<B>, StrategyMetrics)

Applies (1+λ) selection and returns the updated state.

The canonical CGP tie-breaking rule is used: an offspring replaces the parent when its fitness is less than or equal to the parent’s, allowing neutral mutations to accumulate and maintain genetic diversity in the inactive (non-coding) portion of the genome.

The first tell after init bootstraps the parent fitness from the initial single-genome evaluation rather than running selection.

fn best(&self, state: &CgpState<B>) -> Option<(Tensor<B, 2, Int>, f32)>

Returns the best-so-far genome and its fitness, or None before the first tell call.

type Params = CgpConfig

Static parameters for a run.

type State = CgpState<B>

Generation-to-generation state.

type Genome = Tensor<B, 2, Int>

Genome container produced by ask.