Skip to main content

Module prelude

Module prelude 

Source
Expand description

Common imports for ergonomic usage.

This prelude re-exports the primary types and common constants: Matrix, DeterminantWithErrorBound, Vector, Lu, Ldlt, Tolerance, and LaError. Its typed error categories include ArithmeticOperation, FactorizationKind, InvalidToleranceReason, NonFiniteLocation, NonFiniteOrigin, PositiveSemidefiniteViolation, SingularityReason, and UnrepresentableReason. It also re-exports DEFAULT_SINGULAR_TOL, MAX_STACK_MATRIX_DISPATCH_DIM, and try_with_stack_matrix! for runtime-to-const matrix dispatch. Advanced custom-filter code should import ERR_COEFF_2, ERR_COEFF_3, and ERR_COEFF_4 explicitly from the crate root; those raw coefficients intentionally stay out of the prelude.

When the exact feature is enabled, DeterminantSign, ExactF64Conversion, BigInt, and BigRational are also re-exported. ExactF64Conversion converts an already-computed exact determinant or solution under either the strict or explicitly rounded binary64 contract, without repeating exact elimination. The number types let callers construct expected exact values without adding num-bigint / num-rational to their own dependencies. The most commonly needed num-traits items are re-exported alongside them: FromPrimitive for BigRational::from_f64 / from_i64, ToPrimitive for BigRational::to_f64 / to_i64, and Signed for .is_positive() / .is_negative() / .abs().

Re-exports§

pub use crate::ArithmeticOperation;
pub use crate::DEFAULT_SINGULAR_TOL;
pub use crate::DeterminantWithErrorBound;
pub use crate::FactorizationKind;
pub use crate::InvalidToleranceReason;
pub use crate::LaError;
pub use crate::Ldlt;
pub use crate::Lu;
pub use crate::MAX_STACK_MATRIX_DISPATCH_DIM;
pub use crate::Matrix;
pub use crate::NonFiniteLocation;
pub use crate::NonFiniteOrigin;
pub use crate::PositiveSemidefiniteViolation;
pub use crate::SingularityReason;
pub use crate::Tolerance;
pub use crate::UnrepresentableReason;
pub use crate::Vector;
pub use crate::DeterminantSign;
pub use crate::ExactF64Conversion;

Macros§

try_with_stack_matrix
Fallibly dispatch a runtime dimension to a concrete stack-allocated matrix.

Structs§

BigInt
A big signed integer type.

Traits§

FromPrimitive
A generic trait for converting a number to a value.
Signed
Useful functions for signed numbers (i.e. numbers that can be negative).
ToPrimitive
A generic trait for converting a value to a number.

Type Aliases§

BigRational
Alias for arbitrary precision rationals.