feral 0.11.2

Sparse symmetric indefinite direct solver in pure Rust, with certified inertia counts.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

// Deny `.unwrap()` and `.expect()` in production code, but allow them in
// test modules (inside `#[cfg(test)]` blocks) where panics are acceptable.
// This is a structural enforcement of the CLAUDE.md hard rule against
// unwrap in `src/`, replacing the ad-hoc grep check in CI.
#![cfg_attr(not(test), deny(clippy::unwrap_used))]
#![cfg_attr(not(test), deny(clippy::expect_used))]
// Style lints that fire only in test scaffolding — relaxed under cfg(test).
// The lib build keeps default clippy strictness.
#![cfg_attr(test, allow(clippy::needless_range_loop))]

pub mod capi;
pub mod dense;
pub mod error;
pub mod inertia;
pub mod io;
pub mod lu;
pub mod numeric;
pub mod ordering;
pub mod scaling;
pub mod sparse;
pub mod symbolic;

// Flat public API re-exported at crate root:
pub use dense::factor::{
    factor, factor_single_front, BunchKaufmanParams, Factors, ZeroPivotAction,
};
pub use dense::matrix::SymmetricMatrix;
pub use dense::solve::{solve, solve_refined};
pub use error::FeralError;
pub use inertia::Inertia;
pub use io::mtx::{parse_mtx, read_mtx, MtxMatrix};
pub use io::sidecar::{read_sidecar, KktSidecar, SidecarInertia};
// Unsymmetric LU basis engine (issue #81). LU operations are methods on the
// factor structs, so the symmetric crate-root `factor`/`solve` names are
// unaffected.
pub use lu::dense_factor::DenseLu;
pub use lu::dense_matrix::GeneralMatrix;
pub use lu::sparse_factor::SparseLu;
pub use lu::sparse_matrix::SparseColMatrix;
pub use lu::sparse_symbolic::SparseLuSymbolic;
pub use lu::{should_use_dense_lu, LuParams, LuScaling, LuSingularAction};
pub use numeric::condition::{estimate_condition_1norm, estimate_inverse_norm_1, matrix_norm_1};
pub use numeric::factorize::{
    factorize_multifrontal_with_schur, LdltExport, NumericParams, ProfileReport, SchurBlock,
};
pub use numeric::solve::{
    solve_sparse, solve_sparse_refined, solve_sparse_refined_with_diagnostics,
    RefinementDiagnostics, RefinementStep,
};
pub use numeric::solver::{FactorStats, FactorStatus, QualityLevel, Solver};
pub use sparse::csc::{CscMatrix, CscPattern};
pub use symbolic::SymbolicProfileReport;