her2: Hermitian rank 2 update

\(A = \alpha xy^H + conj(\alpha) yx^H + A\) More...

Functions
template<typename TA , typename TX , typename TY >
void	blas::her2 (blas::Layout layout, blas::Uplo uplo, int64_t n, blas::scalar_type< TA, TX, TY > alpha, TX const *x, int64_t incx, TY const *y, int64_t incy, TA *A, int64_t lda)
	Hermitian matrix rank-2 update:
void	blas::her2 (blas::Layout layout, blas::Uplo uplo, int64_t n, float alpha, float const *x, int64_t incx, float const *y, int64_t incy, float *A, int64_t lda)
	CPU, float version.
void	blas::her2 (blas::Layout layout, blas::Uplo uplo, int64_t n, double alpha, double const *x, int64_t incx, double const *y, int64_t incy, double *A, int64_t lda)
	CPU, double version.
void	blas::her2 (blas::Layout layout, blas::Uplo uplo, int64_t n, std::complex< float > alpha, std::complex< float > const *x, int64_t incx, std::complex< float > const *y, int64_t incy, std::complex< float > *A, int64_t lda)
	CPU, complex<float> version.
void	blas::her2 (blas::Layout layout, blas::Uplo uplo, int64_t n, std::complex< double > alpha, std::complex< double > const *x, int64_t incx, std::complex< double > const *y, int64_t incy, std::complex< double > *A, int64_t lda)
	CPU, complex<double> version.

Detailed Description

\(A = \alpha xy^H + conj(\alpha) yx^H + A\)

Function Documentation

her2()

template<typename TA , typename TX , typename TY >

void blas::her2	(	blas::Layout	layout,
		blas::Uplo	uplo,
		int64_t	n,
		blas::scalar_type< TA, TX, TY >	alpha,
		TX const *	x,
		int64_t	incx,
		TY const *	y,
		int64_t	incy,
		TA *	A,
		int64_t	lda
	)

Hermitian matrix rank-2 update:

\[ A = \alpha x y^H + \text{conj}(\alpha) y x^H + A, \]

where alpha is a scalar, x and y are vectors, and A is an n-by-n Hermitian matrix.

Generic implementation for arbitrary data types.

Parameters

[in]	layout	Matrix storage, Layout::ColMajor or Layout::RowMajor.
[in]	uplo	What part of the matrix A is referenced, the opposite triangle being assumed from symmetry. Uplo::Lower: only the lower triangular part of A is referenced. Uplo::Upper: only the upper triangular part of A is referenced.
[in]	n	Number of rows and columns of the matrix A. n >= 0.
[in]	alpha	Scalar alpha. If alpha is zero, A is not updated.
[in]	x	The n-element vector x, in an array of length (n-1)*abs(incx) + 1.
[in]	incx	Stride between elements of x. incx must not be zero. If incx < 0, uses elements of x in reverse order: x(n-1), ..., x(0).
[in]	y	The n-element vector y, in an array of length (n-1)*abs(incy) + 1.
[in]	incy	Stride between elements of y. incy must not be zero. If incy < 0, uses elements of y in reverse order: y(n-1), ..., y(0).
[in,out]	A	The n-by-n matrix A, stored in an lda-by-n array [RowMajor: n-by-lda]. Imaginary parts of the diagonal elements need not be set, are assumed to be zero on entry, and are set to zero on exit.
[in]	lda	Leading dimension of A. lda >= max(1, n).