core integral for Yukawa and exponential interactions More...

#include <boys.h>

Public Member Functions
	TennoGmEval (unsigned int mmax, Real precision=maxabsprecision)

unsigned int	max_m () const

Real	precision () const

void	eval_yukawa (Real *Gm, Real one_over_rho, Real T, size_t mmax, Real zeta) const

void	eval_slater (Real *Gm, Real one_over_rho, Real T, size_t mmax, Real zeta) const

Static Public Member Functions
static std::shared_ptr< const TennoGmEval >	instance (int m_max, double=0)
	Singleton interface allows to manage the lone instance; adjusts max m values as needed in thread-safe fashion.

Detailed Description

template<typename Real = double>
struct libint2::TennoGmEval< Real >

core integral for Yukawa and exponential interactions

Evaluates core integral for Gaussian integrals over the Yukawa potential $\exp(- \zeta r) / r$ and the exponential interaction $\exp(- \zeta r)$

Template Parameters

Real real type

Warning: only Real = double is supported; guarantees absolute precision of only about 1e-14

Constructor & Destructor Documentation

template<typename Real = double>

libint2::TennoGmEval< Real >::TennoGmEval	(	unsigned int	mmax,
		Real	precision = maxabsprecision )

inline

Parameters

m_max	maximum value of the Gm function index
precision	the desired absolute precision (relative precision for most intervals will be below epsilon, but for large T/U values and high m relative precision is low); since evaluation is done primarily by interpolation, precision is not controlled dynamically, thus this should not be used.

Exceptions

std::invalid_argument	if `m_max` is greater than `cheb_table_mmax` (see tenno_cheb.h)
std::invalid_argument	if `precision` is smaller than `maxabsprecision`

template<typename Real = double>

void libint2::TennoGmEval< Real >::eval_slater	(	Real *	Gm,
		Real	one_over_rho,
		Real	T,
		size_t	mmax,
		Real	zeta ) const

inline

Parameters

[in]	Gm	pointer to array of `mmax+1` `Real` elements, on return this contains the core integral for Slater-type geminal, $\exp(-zeta r_{12})$ , namely $(G_{m-1}(T,U) - G_m(T,U)) \zeta / (2 \rho), m \in [0,m_\mathrm{max}]$ where $G_{m}(T,U) = \int_0^1 t^{2m} \exp(U(1-t^{-2}) - Tt^2) dt$ , where $T = \rho \|\vec{P} - \vec{Q}\|^2$ and $U = \zeta^2 / (4 \rho)$
[in]	one_over_rho	$1/\rho$
[in]	T
[in]	mmax	$m_\mathrm{max}$
[in]	zeta	$\zeta$

Exceptions

std::invalid_argument if $\zeta^2/(4 \rho)$ exceeds Umax

template<typename Real = double>

void libint2::TennoGmEval< Real >::eval_yukawa	(	Real *	Gm,
		Real	one_over_rho,
		Real	T,
		size_t	mmax,
		Real	zeta ) const

inline

Parameters

[in]	Gm	pointer to array of `mmax+1` `Real` elements, on return this contains the core integral for Yukawa interaction, $\exp(-zeta r_{12})/r_{12}$ , namely $G_{m}(T,U) = \int_0^1 t^{2m} \exp(U(1-t^{-2}) - Tt^2) dt, m \in [0,m_\mathrm{max}]$ , where $T = \rho \|\vec{P} - \vec{Q}\|^2$ and $U = \zeta^2 / (4 \rho)$
[in]	one_over_rho	$1/\rho$
[in]	T
[in]	mmax	$m_\mathrm{max}$
[in]	zeta	$\zeta$