Struct ExponentialPlane 

pub struct ExponentialPlane { /* private fields */ }

Drive a species’ signed distance to a plane, ξ = x·n̂ − offset, to an exponential distribution ∝ e^(−ξ/λ) (ξ ≥ 0) — a layer densest at the plane and decaying along +n̂ with decay length λ (a diffuse layer).

strength (λ_pen) scales the squared-Wasserstein penalty; scale/scale2 are accepted for trait symmetry but unused.

Implementations

impl ExponentialPlane

pub fn new(normal: [F; 3], offset: F, strength: F, lambda: F) -> Self

• normal — plane normal (normalised internally; need not be unit).
• offset — plane offset so ξ = x·n̂ − offset (the wall is at ξ = 0).
• strength — overall penalty multiplier.
• lambda — exponential decay length (λ > 0).

Panics

If the normal is the zero vector or lambda <= 0.

Trait Implementations

impl Clone for ExponentialPlane

fn clone(&self) -> ExponentialPlane

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ExponentialPlane

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

Formats the value using the given formatter.

impl Restraint for ExponentialPlane

fn f(&self, coords: &[[F; 3]], _scale: F, _scale2: F) -> F

Penalty value for the group’s current configuration.

fn fg( &self, coords: &[[F; 3]], _scale: F, _scale2: F, grads: &mut [[F; 3]], ) -> F

Fused value + gradient. Accumulates ∂L/∂coords[i] INTO grads[i] with +=; returns the same value f would. grads.len() == coords.len().

fn name(&self) -> &'static str

Human-readable identifier.

fn is_parallel_safe(&self) -> bool

If false, the scheduler serializes this restraint (Python-backed collective restraints MUST return false).