nalgebra/tests/linalg/hessenberg.rs

51 lines
1.6 KiB
Rust
Raw Permalink Normal View History

#![cfg(feature = "arbitrary")]
use na::Matrix2;
#[test]
fn hessenberg_simple() {
2018-10-22 13:00:10 +08:00
let m = Matrix2::new(1.0, 0.0, 1.0, 3.0);
let hess = m.hessenberg();
let (p, h) = hess.unpack();
assert!(relative_eq!(m, p * h * p.transpose(), epsilon = 1.0e-7))
}
macro_rules! gen_tests(
($module: ident, $scalar: ty) => {
mod $module {
use na::{DMatrix, Matrix2, Matrix4};
use std::cmp;
#[allow(unused_imports)]
2019-03-23 21:29:07 +08:00
use crate::core::helper::{RandScalar, RandComplex};
quickcheck! {
fn hessenberg(n: usize) -> bool {
let n = cmp::max(1, cmp::min(n, 50));
let m = DMatrix::<$scalar>::new_random(n, n).map(|e| e.0);
let hess = m.clone().hessenberg();
let (p, h) = hess.unpack();
relative_eq!(m, &p * h * p.adjoint(), epsilon = 1.0e-7)
}
fn hessenberg_static_mat2(m: Matrix2<$scalar>) -> bool {
let m = m.map(|e| e.0);
let hess = m.hessenberg();
let (p, h) = hess.unpack();
relative_eq!(m, p * h * p.adjoint(), epsilon = 1.0e-7)
}
fn hessenberg_static(m: Matrix4<$scalar>) -> bool {
let m = m.map(|e| e.0);
let hess = m.hessenberg();
let (p, h) = hess.unpack();
relative_eq!(m, p * h * p.adjoint(), epsilon = 1.0e-7)
}
}
}
}
);
gen_tests!(complex, RandComplex<f64>);
2020-04-06 00:49:48 +08:00
gen_tests!(f64, RandScalar<f64>);