Trait SvdFloat 

pub trait SvdFloat:
    FloatOpsTS
    + ScalarOperand
    + MulAcc
    + DivAssign
    + 'static {
    // Required methods
    fn eps() -> Self;
    fn eps34() -> Self;
    fn to_f64(self) -> f64;
    fn from_f64_val(v: f64) -> Self;

    // Provided method
    fn close(a: Self, b: Self) -> bool { ... }
}

Scalar types this crate can decompose.

Implemented for f32 and f64. The supertraits are what the sparse kernels and the ndarray glue need; nothing here leaks a linear-algebra backend, because the small dense factorizations are always performed in f64 and cast back.

Required Methods

fn eps() -> Self

Machine epsilon.

fn eps34() -> Self

eps^(3/4), the accuracy floor LAS2 clamps kappa to.

fn to_f64(self) -> f64

Widen to f64 for the small dense factorizations.

fn from_f64_val(v: f64) -> Self

Narrow back from f64.

Provided Methods

fn close(a: Self, b: Self) -> bool

Equality within one ulp-ish epsilon.

Dyn Compatibility

This trait is not dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementations on Foreign Types

impl SvdFloat for f32

fn eps() -> Self

fn eps34() -> Self

fn to_f64(self) -> f64

fn from_f64_val(v: f64) -> Self

impl SvdFloat for f64

fn eps() -> Self

fn eps34() -> Self

fn to_f64(self) -> f64

fn from_f64_val(v: f64) -> Self

Implementors