Struct Evaluation 

#[non_exhaustive]
pub struct Evaluation {
    pub value: f64,
    pub violation: f64,
}

What a model gives for one candidate under constraints: its value, and by how much it breaks the constraints, zero if it keeps them all.

Candidates are ranked by K. Deb’s feasibility rules (“An efficient constraint handling method for genetic algorithms”, Computer Methods in Applied Mechanics and Engineering 186(2–4), 311–338 (2000), https://doi.org/10.1016/S0045-7825(99)00389-8, §3): a candidate that keeps every constraint beats one that doesn’t; of two that keep them, the smaller value wins; of two that don’t, the smaller violation wins (here ties in violation go to the smaller value). No penalty weight is needed, as values and violations are never compared with each other. The rules rank, and CMA-ES uses only ranks.

A candidate the model can’t evaluate (a flight that fails) is value and violation both +∞: it ranks behind every other. Both serialize +∞ as none, a JSON null.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

value: f64

The model’s value.

violation: f64

The total violation, Σ max(0, gⱼ) over constraints written gⱼ ≤ 0: zero if the candidate keeps them all.

Implementations

impl Evaluation

pub const fn feasible(value: f64) -> Self

A value with no constraints to break.

pub const fn failed() -> Self

A candidate the model can’t evaluate: value and violation both +∞, behind every other.

pub fn constrained(value: f64, constraints: &[f64]) -> Self

A value under constraints gⱼ(x) ≤ 0, given as the numbers gⱼ: the violation is Σ max(0, gⱼ), Deb’s (2000) overall violation. Deb divides each constraint by a constant so that they count alike (a margin in calibers and a speed in m/s, say); do the same before passing them in. A NaN gⱼ gives a NaN violation, which cmaes::Run::tell_constrained refuses.

pub fn is_feasible(&self) -> bool

Whether the candidate keeps every constraint.

pub fn rank(&self, other: &Self) -> Ordering

Deb’s rules as an ordering: Less if self ranks ahead of other. Violation first, with −0 equal to 0, then value by f64::total_cmp (CMA-ES’s own ranking of plain values).

Trait Implementations

impl Clone for Evaluation

fn clone(&self) -> Evaluation

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for Evaluation

impl Debug for Evaluation

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for Evaluation

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for Evaluation

fn eq(&self, other: &Evaluation) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for Evaluation

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for Evaluation