Struct DynQr 

pub struct DynQr<T> { /* private fields */ }

QR decomposition of a dynamically-sized matrix (M >= N).

Stores the packed Householder vectors, R, and tau scalars.

Example

use numeris::{DynMatrix, DynVector};

let a = DynMatrix::from_rows(3, 2, &[
    1.0_f64, 0.0,
    1.0, 1.0,
    1.0, 2.0,
]);
let b = DynVector::from_slice(&[1.0, 2.0, 4.0]);
let x = a.qr().unwrap().solve(&b);
assert!((x[0] - 5.0 / 6.0).abs() < 1e-10);
assert!((x[1] - 3.0 / 2.0).abs() < 1e-10);

Implementations

impl<T: LinalgScalar> DynQr<T>

pub fn new(a: &DynMatrix<T>) -> Result<Self, LinalgError>

Decompose a matrix. Returns an error if a column is rank-deficient.

pub fn r(&self) -> DynMatrix<T>

Extract the upper-triangular R factor (N × N).

pub fn q(&self) -> DynMatrix<T>

Compute the thin Q factor (M × N).

pub fn solve(&self, b: &DynVector<T>) -> DynVector<T>

Solve the least-squares problem min ||Ax - b|| for x.

pub fn det(&self) -> T

Determinant of the original matrix (square only).

Trait Implementations

impl<T: Debug> Debug for DynQr<T>

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

Formats the value using the given formatter.