neml2::linalg Namespace Reference

Functions
SSR4	dsptrf (const Vec &evals, const R2 &evecs, const Vec &f, const Vec &df)
	Derivative of a spectral transformation.
std::pair< Vec, R2 >	eigh (const SR2 &m)
	Calculate the eigenvalues and eigenvectors of a symmetric second order tensor.
SR2	ieigh (const Vec &eigvals, const R2 &eigvecs)
	Reconstruct the symmetric second order tensor from its eigenvalues and eigenvectors.
std::tuple< Tensor, Tensor >	lu_factor (const Tensor &A, bool pivot)
Tensor	lu_solve (const Tensor &LU, const Tensor &pivots, const Tensor &B, bool left, bool adjoint)
Tensor	solve (const Tensor &A, const Tensor &B)
	Solve the linear system A X = B.

Function Documentation

dsptrf()

SSR4 dsptrf	(	const Vec &	evals,
		const R2 &	evecs,
		const Vec &	f,
		const Vec &	df )

Derivative of a spectral transformation.

A spectral transformation \( f \) is defined as

\[ f(\mathbf{M}) = \sum_i f(\lambda_i) \mathbf{v}_i \otimes \mathbf{v}_i \]

where \( \lambda_i \) are the eigenvalues and \( \mathbf{v}_i \) are the eigenvectors of \( \mathbf{M} \).

eigh()

std::pair< Vec, R2 > eigh

(

const SR2 &

m

)

Calculate the eigenvalues and eigenvectors of a symmetric second order tensor.

ieigh()

SR2 ieigh	(	const Vec &	eigvals,
		const R2 &	eigvecs )

Reconstruct the symmetric second order tensor from its eigenvalues and eigenvectors.

lu_factor()

std::tuple< Tensor, Tensor > lu_factor	(	const Tensor &	A,
		bool	pivot )

lu_solve()

Tensor lu_solve	(	const Tensor &	LU,
		const Tensor &	pivots,
		const Tensor &	B,
		bool	left,
		bool	adjoint )

solve()

Tensor solve	(	const Tensor &	A,
		const Tensor &	B )

Solve the linear system A X = B.