Teuchos::details::GivensRotator< ScalarType, false > Class Template Reference

#include <Teuchos_BLAS.hpp>

Public Types
typedef ScalarType	c_type

Public Member Functions
void	ROTG (ScalarType *da, ScalarType *db, ScalarType *c, ScalarType *s) const
ScalarType	SIGN (const ScalarType &x, const ScalarType &y) const

Detailed Description

template<typename ScalarType>
class Teuchos::details::GivensRotator< ScalarType, false >

Definition at line 286 of file Teuchos_BLAS.hpp.

Member Typedef Documentation

template<typename ScalarType >

typedef ScalarType Teuchos::details::GivensRotator< ScalarType, false >::c_type

Definition at line 288 of file Teuchos_BLAS.hpp.

Member Function Documentation

template<typename ScalarType >

void Teuchos::details::GivensRotator< ScalarType, false >::ROTG	(	ScalarType *	da,
		ScalarType *	db,
		ScalarType *	c,
		ScalarType *	s
	)		const

Definition at line 398 of file Teuchos_BLAS.hpp.

template<typename ScalarType >

ScalarType Teuchos::details::GivensRotator< ScalarType, false >::SIGN ( const ScalarType &  x, const ScalarType &  y  ) const

inline

Return ABS(x) if y > 0 or y is +0, else -ABS(x) (if y is -0 or < 0).

Note that SIGN respects IEEE 754 floating-point signed zero. This is a hopefully correct implementation of the Fortran type-generic SIGN intrinsic. ROTG for complex arithmetic doesn't require this function. C99 provides a copysign() math library function, but we are not able to rely on the existence of C99 functions here.

We provide this method on purpose only for the real-arithmetic specialization of GivensRotator. Complex numbers don't have a sign; they have an angle.

Definition at line 307 of file Teuchos_BLAS.hpp.

---

The documentation for this class was generated from the following file:

• Teuchos_BLAS.hpp