Type Alias TypedRankRevealingQrResult 

pub type TypedRankRevealingQrResult<T> = RankRevealingQrResult<TypedTensor<T>, TypedTensor<i64>>;

Fixed typed output for rank-revealing QR.

Examples

use tenferro_linalg::{RankRevealingQrResult, TypedRankRevealingQrResult};
use tenferro_tensor::TypedTensor;

let result: TypedRankRevealingQrResult<f64> = RankRevealingQrResult {
    q: TypedTensor::from_vec_col_major(vec![1, 1], vec![1.0])?,
    r: TypedTensor::from_vec_col_major(vec![1, 1], vec![2.0])?,
    column_permutation: TypedTensor::from_vec_col_major(vec![1], vec![0_i64])?,
    rank: TypedTensor::from_vec_col_major(vec![], vec![1_i64])?,
};
assert_eq!(result.rank.as_slice()?, &[1]);

Aliased Type

pub struct TypedRankRevealingQrResult<T> {
    pub q: TypedTensor<T>,
    pub r: TypedTensor<T>,
    pub column_permutation: TypedTensor<i64>,
    pub rank: TypedTensor<i64>,
}

Fields

q: TypedTensor<T>

Thin orthonormal factor with shape [m, min(m, n), batch...].

r: TypedTensor<T>

Upper-trapezoidal factor with shape [min(m, n), n, batch...].

column_permutation: TypedTensor<i64>

Zero-based original column at each factor column, shaped [n, batch...].

rank: TypedTensor<i64>

Leading-prefix numerical rank, shaped [batch...].