faer/linalg/svd/

bidiag.rs

use crate::assert;
use crate::internal_prelude::*;
use linalg::householder::*;
use linalg::matmul::{dot, matmul};
/// computes the layout of the workspace required to compute a matrix's
/// bidiagonalization
pub fn bidiag_in_place_scratch<T: ComplexField>(
	nrows: usize,
	ncols: usize,
	par: Par,
	params: Spec<BidiagParams, T>,
) -> StackReq {
	_ = par;
	_ = params;
	StackReq::all_of(&[
		temp_mat_scratch::<T>(nrows, 1),
		temp_mat_scratch::<T>(ncols, 1),
	])
}
/// bidiagonalization tuning parameters.
#[derive(Debug, Copy, Clone)]
pub struct BidiagParams {
	/// threshold at which parallelism should be disabled
	pub par_threshold: usize,
	#[doc(hidden)]
	pub non_exhaustive: NonExhaustive,
}
impl<T: ComplexField> Auto<T> for BidiagParams {
	fn auto() -> Self {
		Self {
			par_threshold: 192 * 256,
			non_exhaustive: NonExhaustive(()),
		}
	}
}
/// computes a matrix $A$'s bidiagonalization such that $A = U B V^H$
///
/// $B$ is a bidiagonal matrix stored in $A$'s diagonal and superdiagonal
///
/// $U$ is a sequence of householder reflections stored in the unit lower
/// triangular half of $A$, with the householder coefficients being stored in
/// `H_left`
///
/// $V$ is a sequence of householder reflections stored in the unit upper
/// triangular half of $A$ (excluding the diagonal), with the householder
/// coefficients being stored in `H_right`
pub fn bidiag_in_place<T: ComplexField>(
	A: MatMut<'_, T>,
	H_left: MatMut<'_, T>,
	H_right: MatMut<'_, T>,
	par: Par,
	stack: &mut MemStack,
	params: Spec<BidiagParams, T>,
) {
	let params = params.config;
	let m = A.nrows();
	let n = A.ncols();
	let size = Ord::min(m, n);
	let bl = H_left.nrows();
	let br = H_right.nrows();
	assert!(H_left.ncols() == size);
	assert!(H_right.ncols() == size.saturating_sub(1));
	alloca!('stack: {
		let mut y = unsafe { row![uninit::<T>, n] };
		let mut z = unsafe { col![uninit::<T>, m] };
	});
	let mut A = A;
	let mut Hl = H_left;
	let mut Hr = H_right;
	let mut par = par;
	{
		let mut Hl = Hl.rb_mut().row_mut(0);
		let mut Hr = Hr.rb_mut().row_mut(0);
		for k in 0..size {
			let mut A = A.rb_mut();
			let (_, A01, A10, A11) = A.rb_mut().split_at_mut(k, k);
			let (_, A02) = A01.split_first_col().unwrap();
			let (A10, A20) = A10.split_first_row_mut().unwrap();
			let (mut A11, A12, A21, mut A22) = A11.split_at_mut(1, 1);
			let mut A12 = A12.row_mut(0);
			let mut A21 = A21.col_mut(0);
			let a11 = &mut A11[(0, 0)];
			let (y1, mut y2) =
				y.rb_mut().split_at_col_mut(k).1.split_at_col_mut(1);
			let (z1, mut z2) =
				z.rb_mut().split_at_row_mut(k).1.split_at_row_mut(1);
			if k > 0 {
				let ref y1 = y1[0].copy();
				let ref z1 = z1[0].copy();
				let k1 = k - 1;
				let ref up0 = A10[k1];
				let up = A20.rb().col(k1);
				let vp = A02.rb().row(k1);
				*a11 -= up0 * y1 + z1;
				z!(A21.rb_mut(), up.rb(), z2.rb()).for_each(
					|uz!(a, u, z): Zip!(&mut T, &T, &T)| *a -= u * y1 + z,
				);
				z!(A12.rb_mut(), y2.rb(), vp.rb()).for_each(
					|uz!(a, y, v): Zip!(&mut T, &T, &T)| {
						*a -= up0 * y + z1 * v;
					},
				);
			}
			let HouseholderInfo { tau: tl, .. } =
				make_householder_in_place(a11, A21.rb_mut());
			let ref tl_inv = tl.recip();
			Hl[k] = tl.to_cplx();
			if (m - k - 1) * (n - k - 1) < params.par_threshold {
				par = Par::Seq;
			}
			if k > 0 {
				let k1 = k - 1;
				let up = A20.rb().col(k1);
				let vp = A02.row(k1);
				match par {
					Par::Seq => bidiag_fused_op(
						A22.rb_mut(),
						A21.rb(),
						up.rb(),
						z2.rb(),
						y2.rb_mut(),
						vp.rb(),
						simd_align(k + 1),
					),
					#[cfg(feature = "rayon")]
					Par::Rayon(nthreads) => {
						use rayon::prelude::*;
						let nthreads = nthreads.get();
						spindle::for_each(
							nthreads,
							A22.rb_mut()
								.par_col_partition_mut(nthreads)
								.zip_eq(y2.rb_mut().par_partition_mut(nthreads))
								.zip_eq(vp.par_partition(nthreads)),
							|((A22, y2), vp)| {
								bidiag_fused_op(
									A22,
									A21.rb(),
									up.rb(),
									z2.rb(),
									y2,
									vp.rb(),
									simd_align(k + 1),
								);
							},
						);
					},
				}
			} else {
				matmul(
					y2.rb_mut(),
					Accum::Replace,
					A21.rb().adjoint(),
					A22.rb(),
					one(),
					par,
				);
			}
			z!(y2.rb_mut(), A12.rb_mut()).for_each(|uz!(y, a)| {
				*y = (&*y + &*a).mul_real(tl_inv);
				*a -= &*y;
			});
			let norm = A12.rb().norm_l2();
			if norm != zero() {
				// a subnormal row norm has an infinite reciprocal, so rescale the row
				// first: reference `dlarfg` does the same when its normaliser underflows
				let ref norm_inv = if norm >= min_positive() {
					(&norm).recip()
				} else {
					let ref shift = from_f64::<T::Real>(4503599627370496.0);
					z!(A12.rb_mut()).for_each(|uz!(a)| *a = a.mul_real(shift));
					(&norm * shift).recip()
				};
				z!(A12.rb_mut()).for_each(|uz!(a)| *a = a.mul_real(norm_inv));
			}
			matmul(
				z2.rb_mut(),
				Accum::Replace,
				A22.rb(),
				A12.rb().adjoint(),
				one(),
				par,
			);
			if k + 1 == size {
				break;
			}
			let (mut A12_a, mut A12_b) = A12.rb_mut().split_at_col_mut(1);
			let A22_a = A22.rb().col(0);
			let (y2_a, y2_b) = y2.rb().split_at_col(1);
			let y2_a = &y2_a[0];
			let a12_a = &mut A12_a[0];
			let HouseholderInfo {
				tau: tr,
				head_with_beta_inv: m,
				..
			} = make_householder_in_place(a12_a, A12_b.rb_mut().transpose_mut());
			let ref tr_inv = tr.recip();
			Hr[k] = tr.to_cplx();
			let beta = a12_a.copy();
			*a12_a = a12_a.mul_real(norm);
			let ref b = y2_a
				+ dot::inner_prod(
					y2_b,
					Conj::No,
					A12_b.rb().transpose(),
					Conj::Yes,
				);
			if m != infinity() {
				z!(z2.rb_mut(), A21.rb(), A22_a.rb()).for_each(
					|uz!(z, u, a): Zip!(&mut T, &T, &T)| {
						let w = &*z - a * beta.conj();
						let w = w * m.conj();
						let w = w - u * b;
						*z = w.mul_real(tr_inv);
					},
				);
			} else {
				z!(z2.rb_mut(), A21.rb(), A22_a.rb()).for_each(
					|uz!(z, u, a): Zip!(&mut T, &T, &T)| {
						let w = a - u * b;
						*z = w.mul_real(tr_inv);
					},
				);
			}
		}
	}
	let mut j = 0;
	while j < size {
		let bl = Ord::min(bl, size - j);
		let mut Hl = Hl.rb_mut().get_mut(..bl, j..j + bl);
		for k in 0..bl {
			Hl[(k, k)] = Hl[(0, k)].copy();
		}
		upgrade_householder_factor(
			Hl.rb_mut(),
			A.rb().get(j.., j..j + bl),
			bl,
			1,
			par,
		);
		j += bl;
	}
	if size > 0 {
		let size = size - 1;
		let A = A.rb().get(..size, 1..);
		let mut Hr = Hr.rb_mut().get_mut(.., ..size);
		let mut j = 0;
		while j < size {
			let br = Ord::min(br, size - j);
			let mut Hr = Hr.rb_mut().get_mut(..br, j..j + br);
			for k in 0..br {
				Hr[(k, k)] = Hr[(0, k)].copy();
			}
			upgrade_householder_factor(
				Hr.rb_mut(),
				A.transpose().get(j.., j..j + br),
				br,
				1,
				par,
			);
			j += br;
		}
	}
}
fn bidiag_fused_op<T: ComplexField>(
	A22: MatMut<'_, T>,
	u: ColRef<'_, T>,
	up: ColRef<'_, T>,
	z: ColRef<'_, T>,
	y: RowMut<'_, T>,
	vp: RowRef<'_, T>,
	align: usize,
) {
	let mut A22 = A22;
	if const { T::SIMD_CAPABILITIES.is_simd() } {
		if let (Some(A22), Some(u), Some(up), Some(z)) = (
			A22.rb_mut().try_as_col_major_mut(),
			u.try_as_col_major(),
			up.try_as_col_major(),
			z.try_as_col_major(),
		) {
			bidiag_fused_op_simd(A22, u, up, z, y, vp, align);
		} else {
			bidiag_fused_op_fallback(A22, u, up, z, y, vp);
		}
	} else {
		bidiag_fused_op_fallback(A22, u, up, z, y, vp);
	}
}
fn bidiag_fused_op_fallback<T: ComplexField>(
	A22: MatMut<'_, T>,
	u: ColRef<'_, T>,
	up: ColRef<'_, T>,
	z: ColRef<'_, T>,
	y: RowMut<'_, T>,
	vp: RowRef<'_, T>,
) {
	let mut A22 = A22;
	let mut y = y;
	matmul(A22.rb_mut(), Accum::Add, up, y.rb(), -one::<T>(), Par::Seq);
	matmul(A22.rb_mut(), Accum::Add, z, vp, -one::<T>(), Par::Seq);
	matmul(
		y.rb_mut(),
		Accum::Replace,
		u.adjoint(),
		A22.rb(),
		one(),
		Par::Seq,
	);
}
fn bidiag_fused_op_simd<'M, 'N, T: ComplexField>(
	A22: MatMut<'_, T, usize, usize, ContiguousFwd>,
	u: ColRef<'_, T, usize, ContiguousFwd>,
	up: ColRef<'_, T, usize, ContiguousFwd>,
	z: ColRef<'_, T, usize, ContiguousFwd>,
	y: RowMut<'_, T, usize>,
	vp: RowRef<'_, T, usize>,
	align: usize,
) {
	struct Impl<'a, 'M, 'N, T: ComplexField> {
		A22: MatMut<'a, T, Dim<'M>, Dim<'N>, ContiguousFwd>,
		u: ColRef<'a, T, Dim<'M>, ContiguousFwd>,
		up: ColRef<'a, T, Dim<'M>, ContiguousFwd>,
		z: ColRef<'a, T, Dim<'M>, ContiguousFwd>,
		y: RowMut<'a, T, Dim<'N>>,
		vp: RowRef<'a, T, Dim<'N>>,
		align: usize,
	}
	impl<'a, 'M, 'N, T: ComplexField> pulp::WithSimd for Impl<'a, 'M, 'N, T> {
		type Output = ();

		#[inline(always)]
		fn with_simd<S: pulp::Simd>(self, simd: S) -> Self::Output {
			let Self {
				mut A22,
				u,
				up,
				z,
				mut y,
				vp,
				align,
			} = self;
			let m = A22.nrows();
			let n = A22.ncols();
			let simd = SimdCtx::<T, S>::new_align(T::simd_ctx(simd), m, align);
			let indices = simd.batch_indices::<4>();
			for j in n.indices() {
				let mut A = A22.rb_mut().col_mut(j);
				let mut acc = [simd.zero(); 4];
				let yj = simd.splat(-&y[j]);
				let vj = simd.splat(-&vp[j]);
				simd_iter!(for (IDX, i) in [indices; 4] {
					let mut a = simd.read(A.rb(), i);
					a = simd.mul_add(simd.read(up, i), yj, a);
					a = simd.mul_add(simd.read(z, i), vj, a);
					simd.write(A.rb_mut(), i, a);
					acc[IDX] = simd.conj_mul_add(simd.read(u, i), a, acc[IDX]);
				});
				let acc0 = simd.add(acc[0], acc[1]);
				let acc2 = simd.add(acc[2], acc[3]);
				let acc0 = simd.add(acc0, acc2);
				y[j] = simd.reduce_sum(acc0);
			}
		}
	}
	with_dim!(M, A22.nrows());
	with_dim!(N, A22.ncols());
	dispatch!(
		Impl {
			A22: A22.as_shape_mut(M, N),
			u: u.as_row_shape(M),
			up: up.as_row_shape(M),
			z: z.as_row_shape(M),
			y: y.as_col_shape_mut(N),
			vp: vp.as_col_shape(N),
			align,
		},
		Impl,
		T
	)
}
#[cfg(test)]
mod tests {
	use super::*;
	use crate::stats::prelude::*;
	use crate::utils::approx::*;
	use crate::{Mat, assert, c64};
	use dyn_stack::MemBuffer;
	use std::mem::MaybeUninit;
	#[test]
	fn test_bidiag_real() {
		let rng = &mut StdRng::seed_from_u64(0);
		for (m, n) in [(8, 4), (8, 8)] {
			let size = Ord::min(m, n);
			let A = CwiseMatDistribution {
				nrows: m,
				ncols: n,
				dist: StandardNormal,
			}
			.rand::<Mat<f64>>(rng);
			let bl = 4;
			let br = 3;
			let mut Hl = Mat::zeros(bl, size);
			let mut Hr = Mat::zeros(br, size - 1);
			let mut UV = A.clone();
			bidiag_in_place(
				UV.rb_mut(),
				Hl.rb_mut(),
				Hr.rb_mut(),
				Par::Seq,
				MemStack::new(&mut [MaybeUninit::uninit(); 1024]),
				default(),
			);
			let mut A = A.clone();
			let mut A = A.as_mut();
			apply_block_householder_sequence_transpose_on_the_left_in_place_with_conj(
				UV.rb().get(.., ..size),
				Hl.rb(),
				Conj::Yes,
				A.rb_mut(),
				Par::Seq,
				MemStack::new(&mut MemBuffer::new(
					apply_block_householder_sequence_transpose_on_the_left_in_place_scratch::<f64>(n - 1, 1, m),
				)),
			);
			let V = UV.rb().get(..size - 1, 1..size);
			let A1 = A.rb_mut().get_mut(.., 1..size);
			let Hr = Hr.as_ref();
			apply_block_householder_sequence_on_the_right_in_place_with_conj(
				V.transpose(),
				Hr.as_ref(),
				Conj::Yes,
				A1,
				Par::Seq,
				MemStack::new(&mut MemBuffer::new(
					apply_block_householder_sequence_on_the_right_in_place_scratch::<f64>(n - 1, 1, m),
				)),
			);
			let approx_eq = CwiseMat(ApproxEq::<f64>::eps());
			for j in 0..n {
				for i in 0..m {
					if i > j || j > i + 1 {
						UV[(i, j)] = 0.0;
					}
				}
			}
			assert!(UV ~ A);
		}
	}
	#[test]
	fn test_bidiag_cplx() {
		let rng = &mut StdRng::seed_from_u64(0);
		for (m, n) in [(8, 4), (8, 8)] {
			let size = Ord::min(m, n);
			let A = CwiseMatDistribution {
				nrows: m,
				ncols: n,
				dist: ComplexDistribution::new(StandardNormal, StandardNormal),
			}
			.rand::<Mat<c64>>(rng);
			let bl = 4;
			let br = 3;
			let mut Hl = Mat::zeros(bl, size);
			let mut Hr = Mat::zeros(br, size - 1);
			let mut UV = A.clone();
			let mut UV = UV.as_mut();
			bidiag_in_place(
				UV.rb_mut(),
				Hl.rb_mut(),
				Hr.rb_mut(),
				Par::Seq,
				MemStack::new(&mut [MaybeUninit::uninit(); 1024]),
				default(),
			);
			let mut A = A.clone();
			let mut A = A.as_mut();
			apply_block_householder_sequence_transpose_on_the_left_in_place_with_conj(
				UV.rb().subcols(0, size),
				Hl.rb(),
				Conj::Yes,
				A.rb_mut(),
				Par::Seq,
				MemStack::new(&mut MemBuffer::new(
					apply_block_householder_sequence_transpose_on_the_left_in_place_scratch::<c64>(n - 1, 1, m),
				)),
			);
			let V = UV.rb().get(..size - 1, 1..size);
			let A1 = A.rb_mut().get_mut(.., 1..size);
			let Hr = Hr.rb();
			apply_block_householder_sequence_on_the_right_in_place_with_conj(
				V.transpose(),
				Hr,
				Conj::Yes,
				A1,
				Par::Seq,
				MemStack::new(&mut MemBuffer::new(
					apply_block_householder_sequence_on_the_right_in_place_scratch::<c64>(n - 1, 1, m),
				)),
			);
			let approx_eq = CwiseMat(ApproxEq::eps());
			for j in 0..n {
				for i in 0..m {
					if i > j || j > i + 1 {
						UV[(i, j)] = c64::ZERO;
					}
				}
			}
			assert!(UV ~ A);
		}
	}
}