Enum SolverRecipe 

pub enum SolverRecipe {
    NalgebraTrfLegacy,
    FlatGaussianFirstTie,
    DenseGaussianLastTie,
    ScipyHostLapackReference,
    OwnedDeterministicCholesky,
    OwnedDeterministicTrf,
}

Linear-solve / factorization operation order. Determinism note: the legacy SPP path is nalgebra LU (not bit-portable end-to-end), preserved as a named variant; the owned deterministic kernel (P5) owns the dense subproblem factorization with its own goldens. Its determinism scope is the factorization, not the surrounding nalgebra reductions that build the subproblem – see Self::OwnedDeterministicTrf.

Variants

NalgebraTrfLegacy

nalgebra trust-region least squares, the current SPP solver (spp / crate::astro::math::least_squares). Existing SPP goldens use this; kept unchanged.

FlatGaussianFirstTie

Flat first-tie Gaussian elimination (RTK baseline/filter solve).

DenseGaussianLastTie

Dense last-tie Gaussian elimination (PPP solve, crate::astro::math::linear::solve_linear_last_tie).

ScipyHostLapackReference

scipy host LAPACK reference solve (machine-dependent; only as a fingerprinted CI reference, never canonical).

OwnedDeterministicCholesky

Owned deterministic Cholesky (square-root) linear solve, the canonical RTK (P6 increment 2) and canonical PPP (P6 increment 3) solver: the SPD normal system is factored Λ = L Lᵀ and solved by forward/back substitution through the owned crate::astro::math::linear::solve_flat_normal_square_root_into kernel, with no nalgebra LU and no black-box BLAS. Paired with NormalRecipe::CanonicalSquareRoot. Both the RTK and PPP canonical paths are owned scalar arithmetic and f64 sqrt is IEEE-754 correctly rounded, so unlike Self::OwnedDeterministicTrf (whose surrounding reductions ride nalgebra) its bit guarantee covers the full solve and is portable across platforms.

OwnedDeterministicTrf

Owned deterministic trust-region subproblem solve added in P5: a fixed-reduction-order dense Gaussian elimination (the OwnedGaussianFirstTie kernel) with no nalgebra LU and no black-box BLAS in the factorization, pinned to its OWN frozen-bits goldens. Scope: it owns ONLY the subproblem factorization; the normal-matrix / gradient / norm reductions that build the subproblem still flow through nalgebra’s CPU-dispatched dense algebra, so the cross-platform bit guarantee is scoped to the factorization, not the full solve.

Trait Implementations

impl Clone for SolverRecipe

fn clone(&self) -> SolverRecipe

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for SolverRecipe

impl Debug for SolverRecipe

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

Formats the value using the given formatter.

impl Default for SolverRecipe

fn default() -> SolverRecipe

Returns the “default value” for a type.

impl Eq for SolverRecipe

impl Hash for SolverRecipe

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for SolverRecipe

fn eq(&self, other: &SolverRecipe) -> 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 StructuralPartialEq for SolverRecipe