NeatCombination.java
package net.bmahe.genetics4j.neat.spec.combination;
import org.apache.commons.lang3.Validate;
import org.immutables.value.Value;
import net.bmahe.genetics4j.core.spec.combination.CombinationPolicy;
import net.bmahe.genetics4j.neat.spec.combination.parentcompare.ParentComparisonPolicy;
import net.bmahe.genetics4j.neat.spec.combination.parentcompare.FitnessComparison;
/**
* Configuration policy for NEAT (NeuroEvolution of Augmenting Topologies) genetic crossover operations.
*
* <p>NeatCombination defines how neural network chromosomes should be recombined during genetic crossover,
* including inheritance biases, gene re-enabling policies, and parent comparison strategies. This policy
* controls the fundamental genetic operators that shape network topology evolution in NEAT.
*
* <p>Key crossover parameters:
* <ul>
* <li><strong>Inheritance threshold</strong>: Bias toward fitter parent for gene inheritance</li>
* <li><strong>Gene re-enabling</strong>: Probability of re-enabling disabled genes during crossover</li>
* <li><strong>Parent comparison</strong>: Strategy for determining relative parent fitness</li>
* <li><strong>Genetic alignment</strong>: Innovation-number-based gene matching</li>
* </ul>
*
* <p>NEAT genetic crossover process:
* <ol>
* <li><strong>Gene alignment</strong>: Match genes by innovation number between parents</li>
* <li><strong>Matching genes</strong>: Randomly inherit from either parent (biased by inheritance threshold)</li>
* <li><strong>Disjoint genes</strong>: Inherit from fitter parent based on parent comparison</li>
* <li><strong>Excess genes</strong>: Inherit from fitter parent beyond less fit parent's range</li>
* <li><strong>Gene state</strong>: Apply re-enabling policy to disabled genes</li>
* </ol>
*
* <p>Gene inheritance strategies:
* <ul>
* <li><strong>Matching genes</strong>: Present in both parents with same innovation number</li>
* <li><strong>Disjoint genes</strong>: Present in one parent within other parent's innovation range</li>
* <li><strong>Excess genes</strong>: Present in one parent beyond other parent's highest innovation</li>
* <li><strong>Disabled genes</strong>: May be re-enabled based on re-enabling threshold</li>
* </ul>
*
* <p>Common usage patterns:
* <pre>{@code
* // Default NEAT crossover configuration
* NeatCombination defaultPolicy = NeatCombination.build();
*
* // Custom crossover with fitness bias
* NeatCombination biasedPolicy = NeatCombination.builder()
* .inheritanceThresold(0.7) // 70% bias toward fitter parent
* .reenableGeneInheritanceThresold(0.3) // 30% chance to re-enable genes
* .parentComparisonPolicy(FitnessComparison.build())
* .build();
*
* // Unbiased crossover for diversity
* NeatCombination unbiasedPolicy = NeatCombination.builder()
* .inheritanceThresold(0.5) // No bias toward either parent
* .reenableGeneInheritanceThresold(0.1) // Low re-enabling rate
* .build();
*
* // Use in EA configuration
* var combinationSpec = ChromosomeCombinatorSpec.builder()
* .combinationPolicy(biasedPolicy)
* .build();
* }</pre>
*
* <p>Inheritance threshold effects:
* <ul>
* <li><strong>0.5 (default)</strong>: Unbiased inheritance, equal probability from both parents</li>
* <li><strong>> 0.5</strong>: Bias toward fitter parent, promotes convergence</li>
* <li><strong>< 0.5</strong>: Bias toward less fit parent, increases diversity</li>
* <li><strong>1.0</strong>: Always inherit from fitter parent (if determinable)</li>
* </ul>
*
* <p>Gene re-enabling mechanism:
* <ul>
* <li><strong>Historical information</strong>: Disabled genes preserve connection topology</li>
* <li><strong>Re-activation chance</strong>: Allows previously disabled connections to contribute again</li>
* <li><strong>Topology exploration</strong>: Enables rediscovery of useful connection patterns</li>
* <li><strong>Genetic diversity</strong>: Prevents permanent loss of structural information</li>
* </ul>
*
* <p>Parent comparison integration:
* <ul>
* <li><strong>Fitness comparison</strong>: Standard fitness-based parent ranking</li>
* <li><strong>Custom strategies</strong>: Pluggable comparison policies for different problem domains</li>
* <li><strong>Multi-objective support</strong>: Compatible with complex fitness landscapes</li>
* <li><strong>Equal fitness handling</strong>: Special rules when parents have identical fitness</li>
* </ul>
*
* <p>Performance considerations:
* <ul>
* <li><strong>Innovation sorting</strong>: Leverages pre-sorted connection lists for O(n) crossover</li>
* <li><strong>Memory efficiency</strong>: Minimal allocation during gene inheritance</li>
* <li><strong>Cache-friendly</strong>: Sequential access patterns for better cache performance</li>
* <li><strong>Parallelizable</strong>: Crossover operations can be executed concurrently</li>
* </ul>
*
* @see ParentComparisonPolicy
* @see FitnessComparison
* @see net.bmahe.genetics4j.neat.combination.NeatChromosomeCombinator
* @see CombinationPolicy
*/
@Value.Immutable
public interface NeatCombination extends CombinationPolicy {
public static final double DEFAULT_INHERITANCE_THRESHOLD = 0.5d;
public static final double DEFAULT_REENABLE_GENE_INHERITANCE_THRESHOLD = 0.25d;
/**
* Returns the inheritance threshold for biasing gene selection toward fitter parents.
*
* <p>This threshold controls the probability of inheriting genes from the fitter parent
* during crossover. Higher values bias inheritance toward the better performing parent,
* while lower values provide more equal inheritance or even bias toward the less fit parent.
*
* <p>Inheritance behavior:
* <ul>
* <li><strong>0.5 (default)</strong>: Unbiased inheritance, equal probability from both parents</li>
* <li><strong>> 0.5</strong>: Bias toward fitter parent, promotes convergence to good solutions</li>
* <li><strong>< 0.5</strong>: Bias toward less fit parent, increases population diversity</li>
* <li><strong>1.0</strong>: Always inherit from fitter parent when fitness differs</li>
* <li><strong>0.0</strong>: Always inherit from less fit parent when fitness differs</li>
* </ul>
*
* @return inheritance threshold value between 0.0 and 1.0 (inclusive)
*/
@Value.Default
default public double inheritanceThresold() {
return DEFAULT_INHERITANCE_THRESHOLD;
}
/**
* Returns the threshold for re-enabling disabled genes during crossover.
*
* <p>When a gene (connection) is disabled in one parent but enabled in the other,
* this threshold determines the probability that the gene will be enabled in the
* offspring. This mechanism prevents permanent loss of potentially useful connections
* and allows rediscovery of structural innovations.
*
* <p>Re-enabling behavior:
* <ul>
* <li><strong>0.25 (default)</strong>: 25% chance to re-enable disabled connections</li>
* <li><strong>0.0</strong>: Never re-enable disabled connections</li>
* <li><strong>1.0</strong>: Always re-enable connections that are enabled in either parent</li>
* <li><strong>Higher values</strong>: More aggressive topology exploration</li>
* <li><strong>Lower values</strong>: More conservative structural preservation</li>
* </ul>
*
* @return re-enabling threshold value between 0.0 and 1.0 (inclusive)
*/
@Value.Default
default public double reenableGeneInheritanceThresold() {
return DEFAULT_REENABLE_GENE_INHERITANCE_THRESHOLD;
}
/**
* Returns the policy used to compare parent fitness for inheritance decisions.
*
* <p>The parent comparison policy determines which parent is considered "fitter"
* for the purposes of biased gene inheritance. This affects how disjoint and excess
* genes are inherited and how the inheritance threshold is applied.
*
* <p>Available comparison strategies:
* <ul>
* <li><strong>FitnessComparison (default)</strong>: Compare parents based on their fitness values</li>
* <li><strong>Custom policies</strong>: Pluggable strategies for domain-specific comparisons</li>
* <li><strong>Multi-objective</strong>: Specialized comparisons for multi-objective optimization</li>
* <li><strong>Equal fitness handling</strong>: Specific behavior when parents have identical fitness</li>
* </ul>
*
* @return the parent comparison policy (defaults to fitness-based comparison)
*/
@Value.Default
default public ParentComparisonPolicy parentComparisonPolicy() {
return FitnessComparison.build();
}
@Value.Check
default void check() {
Validate.inclusiveBetween(0, 1, inheritanceThresold());
Validate.inclusiveBetween(0, 1, reenableGeneInheritanceThresold());
}
class Builder extends ImmutableNeatCombination.Builder {
}
static Builder builder() {
return new Builder();
}
/**
* Creates a NEAT combination policy with default settings.
*
* <p>Default configuration:
* <ul>
* <li>Inheritance threshold: 0.5 (unbiased)</li>
* <li>Gene re-enabling threshold: 0.25 (25% chance)</li>
* <li>Parent comparison: Fitness-based comparison</li>
* </ul>
*
* @return a new NEAT combination policy with default settings
*/
static NeatCombination build() {
return builder().build();
}
}