Struct AdaptiveFeasibility 

pub struct AdaptiveFeasibility { /* private fields */ }

Adaptive feasibility pressure that adjusts constraint-handling behavior based on the ratio of feasible to infeasible solutions in the population.

When the feasibility ratio is low, the comparison rule leans toward preferring solutions with smaller violations (exploration mode). As feasibility improves, it transitions toward objective-based comparison (exploitation mode).

This prevents premature convergence into infeasible regions while still allowing useful gradient information from near-feasible solutions.

Implementations

impl AdaptiveFeasibility

pub fn new( target_ratio: f64, alpha: f64, feasibility_tol: f64, ) -> OptimizeResult<Self>

Create a new adaptive feasibility handler.

Arguments

• target_ratio — Desired fraction of feasible solutions (0 < t ≤ 1).
• alpha — Update learning rate for the ratio estimate.
• feasibility_tol — Tolerance below which violations are treated as zero.

pub fn update(&mut self, violations: &[ViolationSummary])

Update the feasibility ratio estimate from the current population.

Arguments

• violations — Violation summaries for all population members.

pub fn feasibility_ratio(&self) -> f64

Current estimated feasibility ratio.

pub fn effective_penalty_weight(&self, base_penalty: f64) -> f64

Compute the effective penalty weight for constraint violations.

When feasibility_ratio < target_ratio, penalty is amplified to drive solutions toward feasibility. When ratio >= target_ratio, penalty is reduced to emphasize objective improvement.

The weight is: base_penalty * (target / current)^2 if below target, or base_penalty * (current / target)^(-1) if above.

pub fn compare( &self, f_a: f64, viol_a: &ViolationSummary, f_b: f64, viol_b: &ViolationSummary, ) -> Ordering

Compare two solutions using the adaptive feasibility-weighted rule.

The effective comparison weight between objective and violation is controlled by the current feasibility pressure.

Trait Implementations

impl Clone for AdaptiveFeasibility

fn clone(&self) -> AdaptiveFeasibility

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AdaptiveFeasibility

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

Formats the value using the given formatter.