madness::FunctionCommonData< T, NDIM > Class Template Reference

FunctionCommonData holds all Function data common for given k. More...

#include <function_common_data.h>

Collaboration diagram for madness::FunctionCommonData< T, NDIM >:

Public Types
typedef Tensor< T >	tensorT
	Type of tensor used to hold coeff.

Static Public Member Functions
static void	_init_quadrature (int k, int npt, Tensor< double > &quad_x, Tensor< double > &quad_w, Tensor< double > &quad_phi, Tensor< double > &quad_phiw, Tensor< double > &quad_phit)
	Initialize the quadrature information.
static const FunctionCommonData< T, NDIM > &	get (int k)

Public Attributes
Tensor< double >	g0
Tensor< double >	g0T
Tensor< double >	g1
	The separate blocks of twoscale coefficients.
Tensor< double >	g1T
	The separate blocks of twoscale coefficients.
Tensor< double >	h0
Tensor< double >	h0T
Tensor< double >	h1
Tensor< double >	h1T
Tensor< double >	hg
Tensor< double >	hgsonly
	hg[0:k,:]
Tensor< double >	hgT
	The full twoscale coeff (2k,2k) and transpose.
int	k
	order of the wavelet
Key< NDIM >	key0
	Key for root node.
int	npt
	no. of quadrature points
Tensor< double >	quad_phi
	quad_phi(i,j) = at x[i] value of phi[j]
Tensor< double >	quad_phit
	transpose of quad_phi
Tensor< double >	quad_phiw
	quad_phiw(i,j) = at x[i] value of w[i]*phi[j]
Tensor< double >	quad_w
	quadrature weights
Tensor< double >	quad_x
	quadrature points
Slice	s [4]
	s[0]=Slice(0,k-1), s[1]=Slice(k,2*k-1), etc.
std::vector< Slice >	s0
	s[0] in each dimension to get scaling coeff
std::vector< Slice >	sh
	Slice(0,(k-1)/2) in each dimension for autorefine test.
std::vector< long >	v2k
	(2k,...) used to initialize Tensors
std::vector< long >	vk
	(k,...) used to initialize Tensors
std::vector< long >	vq
	(npt,...) used to initialize Tensors

Private Member Functions
	FunctionCommonData (int k)
	Private. Do first use initialization via get.
void	_init_twoscale ()
	Private. Initialize the twoscale coefficients.

Static Private Attributes
static const FunctionCommonData< T, NDIM > *	data [MAXK]

Detailed Description

template<typename T, std::size_t NDIM>
class madness::FunctionCommonData< T, NDIM >

FunctionCommonData holds all Function data common for given k.

Since Function assignment and copy constructors are shallow it greatly simplifies maintaining consistent state to have all (permanent) state encapsulated in a single class. The state is shared between instances using a shared_ptr. Also, separating shared from instance specific state accelerates the constructor, which is important for massive parallelism, and permitting inexpensive use of temporaries. The default copy constructor and assignment operator are used but are probably never invoked.

Member Typedef Documentation

tensorT

template<typename T, std::size_t NDIM>

typedef Tensor<T> madness::FunctionCommonData< T, NDIM >::tensorT

Type of tensor used to hold coeff.

Constructor & Destructor Documentation

FunctionCommonData()

template<typename T, std::size_t NDIM>

madness::FunctionCommonData< T, NDIM >::FunctionCommonData ( int k )

inlineprivate

Private. Do first use initialization via get.

References _init_quadrature(), _init_twoscale(), k, key0, NDIM, npt, quad_phi, quad_phit, quad_phiw, quad_w, quad_x, s, s0, sh, v2k, vk, and vq.

Referenced by get().

Member Function Documentation

_init_quadrature()

template<typename T, std::size_t NDIM>

void madness::FunctionCommonData< T, NDIM >::_init_quadrature ( int k, int npt, Tensor< double > & quad_x, Tensor< double > & quad_w, Tensor< double > & quad_phi, Tensor< double > & quad_phiw, Tensor< double > & quad_phit )

static

Initialize the quadrature information.

Made public with all arguments thru interface for reuse in FunctionImpl::err_box

References _init_quadrature(), madness::gauss_legendre(), k, madness::legendre_scaling_functions(), mu, npt, quad_phi, quad_phit, quad_phiw, quad_w, quad_x, and madness::transpose().

Referenced by FunctionCommonData(), _init_quadrature(), madness::FunctionImpl< T, LDIM >::errsq_local(), and madness::FunctionImpl< T, NDIM >::print_plane_local().

_init_twoscale()

template<typename T, std::size_t NDIM>

void madness::FunctionCommonData< T, NDIM >::_init_twoscale ( )

private

Private. Initialize the twoscale coefficients.

References madness::_(), _init_twoscale(), copy(), g0, g0T, g1, g1T, h0, h0T, h1, h1T, hg, hgsonly, hgT, k, madness::transpose(), and madness::two_scale_hg().

Referenced by FunctionCommonData(), and _init_twoscale().

get()

template<typename T, std::size_t NDIM>

const FunctionCommonData< T, NDIM > & madness::FunctionCommonData< T, NDIM >::get ( int k )

inlinestatic

References FunctionCommonData(), data, k, MADNESS_ASSERT, MADNESS_PRAGMA_CLANG, madness::MAXK, and madness::pop().

Referenced by madness::FunctionImpl< T, LDIM >::mul(), madness::FunctionImpl< T, NDIM >::pointwise_multiplier< LDIM >::operator()(), madness::FunctionImpl< T, NDIM >::pointwise_multiplier< LDIM >::operator()(), madness::FunctionImpl< T, NDIM >::tnorm(), madness::FunctionImpl< T, NDIM >::tnorm(), and madness::FunctionImpl< T, NDIM >::tnorm().

Member Data Documentation

data

template<typename T, std::size_t NDIM>

const FunctionCommonData<T, NDIM>* madness::FunctionCommonData< T, NDIM >::data[MAXK]

staticprivate

Referenced by get().

g0

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::g0

Referenced by _init_twoscale().

g0T

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::g0T

Referenced by _init_twoscale().

g1

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::g1

The separate blocks of twoscale coefficients.

Referenced by _init_twoscale().

g1T

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::g1T

The separate blocks of twoscale coefficients.

Referenced by _init_twoscale().

h0

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::h0

Referenced by _init_twoscale().

h0T

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::h0T

Referenced by _init_twoscale().

h1

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::h1

Referenced by _init_twoscale().

h1T

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::h1T

Referenced by _init_twoscale().

hg

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::hg

Referenced by _init_twoscale().

hgsonly

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::hgsonly

hg[0:k,:]

Referenced by _init_twoscale().

hgT

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::hgT

The full twoscale coeff (2k,2k) and transpose.

Referenced by _init_twoscale().

k

template<typename T, std::size_t NDIM>

int madness::FunctionCommonData< T, NDIM >::k

order of the wavelet

Referenced by FunctionCommonData(), _init_quadrature(), _init_twoscale(), and get().

key0

template<typename T, std::size_t NDIM>

Key<NDIM> madness::FunctionCommonData< T, NDIM >::key0

Key for root node.

Referenced by FunctionCommonData().

npt

template<typename T, std::size_t NDIM>

int madness::FunctionCommonData< T, NDIM >::npt

no. of quadrature points

Referenced by FunctionCommonData(), and _init_quadrature().

quad_phi

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::quad_phi

quad_phi(i,j) = at x[i] value of phi[j]

Referenced by FunctionCommonData(), and _init_quadrature().

quad_phit

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::quad_phit

transpose of quad_phi

Referenced by FunctionCommonData(), _init_quadrature(), and madness::FunctionImpl< T, LDIM >::mul().

quad_phiw

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::quad_phiw

quad_phiw(i,j) = at x[i] value of w[i]*phi[j]

Referenced by FunctionCommonData(), _init_quadrature(), and madness::FunctionImpl< T, LDIM >::mul().

quad_w

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::quad_w

quadrature weights

Referenced by FunctionCommonData(), and _init_quadrature().

quad_x

template<typename T, std::size_t NDIM>

Tensor<double> madness::FunctionCommonData< T, NDIM >::quad_x

quadrature points

Referenced by FunctionCommonData(), _init_quadrature(), and madness::apply_kernel_helper::slater_apply().

s

template<typename T, std::size_t NDIM>

Slice madness::FunctionCommonData< T, NDIM >::s[4]

s[0]=Slice(0,k-1), s[1]=Slice(k,2*k-1), etc.

Referenced by FunctionCommonData().

s0

template<typename T, std::size_t NDIM>

std::vector<Slice> madness::FunctionCommonData< T, NDIM >::s0

s[0] in each dimension to get scaling coeff

Referenced by FunctionCommonData(), madness::FunctionImpl< T, LDIM >::mul(), and madness::FunctionNode< T, NDIM >::recompute_snorm_and_dnorm().

sh

template<typename T, std::size_t NDIM>

std::vector<Slice> madness::FunctionCommonData< T, NDIM >::sh

Slice(0,(k-1)/2) in each dimension for autorefine test.

Referenced by FunctionCommonData().

v2k

template<typename T, std::size_t NDIM>

std::vector<long> madness::FunctionCommonData< T, NDIM >::v2k

(2k,...) used to initialize Tensors

Referenced by FunctionCommonData().

vk

template<typename T, std::size_t NDIM>

std::vector<long> madness::FunctionCommonData< T, NDIM >::vk

(k,...) used to initialize Tensors

Referenced by FunctionCommonData(), madness::FunctionImpl< T, LDIM >::inner_ext_node(), madness::FunctionImpl< T, LDIM >::mul(), and madness::FunctionNode< T, NDIM >::recompute_snorm_and_dnorm().

vq

template<typename T, std::size_t NDIM>

std::vector<long> madness::FunctionCommonData< T, NDIM >::vq

(npt,...) used to initialize Tensors

Referenced by FunctionCommonData().

---

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

• function_common_data.h
• mraimpl.h