Struct NeuralNetwork 

pub struct NeuralNetwork {
    pub genome: EneCode,
    pub graph: DiGraph<Nn, f32>,
    pub node_identity_map: HashMap<String, NodeIndex>,
    /* private fields */
}

NeuralNetwork is a struct that represents a directed graph based feed-forward neural network, initialized from an EneCode genome.

The struct encapsulates the genome (genetic blueprint), graph-based network, node-identity mapping, and network output.

Fields

• genome - The genetic blueprint (EneCode) of the neural network.
• graph - The directed graph (DiGraph) from petgraph that represents the network.
• node_identity_map - A HashMap mapping each neuron ID (String) to its index (NodeIndex) in the graph.
• network_output - A vector holding the output values of the output neurons.

Example Usage

use evo_rl::graph::NeuralNetwork;
use evo_rl::enecode::EneCode;

// Assume genome is a properly initialized EneCode
let mut network = NeuralNetwork::new(genome);

// Assume input is a properly initialized Vec<f32>
network.fwd(input);

let output = network.fetch_network_output();

Fields

genome: EneCode

graph: DiGraph<Nn, f32>

node_identity_map: HashMap<String, NodeIndex>

Implementations

impl NeuralNetwork

pub fn new(genome: EneCode) -> Self

Create a new NeuralNetwork from an EneCode genome.

pub fn serialize_genome(&self) -> String

Returns serialized representation of genome

pub fn recombine_enecode<R: Rng>( &self, rng: &mut R, partner: &NeuralNetwork, ) -> Result<EneCode, GraphConstructionError>

Cross over recombination of genetic code

pub fn mutate( &mut self, mutation_rate: f32, mutation_sd: f32, topology_mutation_rate: f32, )

Run mutation for this network

pub fn transfer(self) -> Self

pub fn fwd(&mut self, input: Vec<f32>)

Forward propagate through the neural network. This function takes a vector of input values and populates the network output.

pub fn fetch_network_output(&self) -> Vec<f32>

Fetch the output of the network as a vector of floating-point numbers.

pub fn write_dot(&self, file_path: &Path)

Write the graph as a .dot file for visualization/inspection

Trait Implementations

impl Clone for NeuralNetwork

fn clone(&self) -> NeuralNetwork

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for NeuralNetwork

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

Formats the value using the given formatter.

impl From<&NeuralNetwork> for EneCode

Creates genome from neural network after recombination and mutation

fn from(network: &NeuralNetwork) -> Self

Converts to this type from the input type.

impl ToPyObject for NeuralNetwork

fn to_object(&self, py: Python<'_>) -> PyObject

Converts self into a Python object.