Struct ValidationParameters 

pub struct ValidationParameters<S: Scalar> {
    pub sample_count: usize,
    pub max_nodes: usize,
    pub min_subdivision_size: S,
    pub max_edge_edge_intersection_samples: usize,
    pub max_edge_face_intersection_samples: usize,
    pub face_face_sample_count: usize,
    pub face_face_newton_iterations: usize,
    pub manifold_ray_sample_count: usize,
    pub manifold_seed: u64,
}

Shared tuning knobs for the whole-model validation checks — how many points to sample along a curve, and the budgets passed down to the underlying BFS/DFS geometric searches.

Fields

sample_count: usize

How many points to sample along a curve for sampling-based checks.

max_nodes: usize

Node budget for the BFS containment searches (see contains::curve::curve_could_contain), and for the pairwise curve_curve_intersect/curve_surface_intersect searches in disjointness_check (which now error, rather than silently returning a truncated result, if they exhaust it — see curve_surface_intersect’s own doc comment).

min_subdivision_size: S

Convergence tolerance for those same searches. Empirically, the pairwise curve_curve_intersect/curve_surface_intersect searches need a much looser tolerance than a single-curve curve_could_contain query to reliably converge within max_nodes: 1e-7 can fail to resolve a clean, unambiguous case at all; 1e-5 resolves most cases but not a near-tangential edge/face pair (e.g. two adjacent flat walls of an extruded_cylinder’s polygon approximation meeting at a shallow dihedral angle, or an edge ending exactly at a revolved cap’s own pole) — each halving of the tolerance costs roughly one extra subdivision level, and a near-tangential pair’s hull-overlap pruning barely discriminates at all (see curve_surface_intersect’s own doc comment), so that one extra level can mean an order of magnitude more nodes. 1e-4 is loose enough to resolve every case in this crate’s own basic-shape validations within max_nodes, while still being far tighter than any genuine geometric feature these checks care about.

max_edge_edge_intersection_samples: usize

max_solutions passed to curve_curve_intersect for every edge x edge pair: how many points to return at most. Coincidence is reported directly (Intersections::Coincident), not inferred from reaching this count.

max_edge_face_intersection_samples: usize

Same idea as max_edge_edge_intersection_samples, but for curve_surface_intersect on every edge x face pair.

face_face_sample_count: usize

How many random (point-on-face-a, point-on-face-b) starting pairs face_face_numerical_intersection tries per face pair.

face_face_newton_iterations: usize

How many alternating-projection Newton rounds each of those starting pairs gets to converge in.

manifold_ray_sample_count: usize

How many random points, and how many random ray directions per point, validate_manifold’s ray-direction-consistency check tries per shell.

manifold_seed: u64

Seeds every random draw in validate_manifold’s ray-direction-consistency check.

Trait Implementations

impl<S: Clone + Scalar> Clone for ValidationParameters<S>

fn clone(&self) -> ValidationParameters<S>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<S: Copy + Scalar> Copy for ValidationParameters<S>

impl<S: Debug + Scalar> Debug for ValidationParameters<S>

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

Formats the value using the given formatter.

impl<S: Scalar> Default for ValidationParameters<S>

fn default() -> Self

Returns the “default value” for a type.