fourier_space_oct_m Module Reference

Data Types
type	fourier_space_op_t
	Definition of a Fourier Space Coulomb Kernel. More...

Functions/Subroutines
subroutine	fourier_space_op_init (this, space, qvector, cam, singularity)
	Initialise a Fourier space Op instance. More...
subroutine	fourier_space_op_end (this)
	Finalize an instance of fourier_space_op_t. More...
subroutine	dfourier_space_op_set_op (this, cube, op, in_device, op_move)
	Set the op buffer, which defines the kernel on the cube, in Fourier space. More...
subroutine, public	dfourier_space_op_apply (this, cube, cf)
	Applies a multiplication factor to the Fourier space grid. This is a local function. More...
subroutine	zfourier_space_op_set_op (this, cube, op, in_device, op_move)
	Set the op buffer, which defines the kernel on the cube, in Fourier space. More...
subroutine, public	zfourier_space_op_apply (this, cube, cf)
	Applies a multiplication factor to the Fourier space grid. This is a local function. More...
subroutine, public	sfourier_space_op_apply (this, cube, cf)
	Applies a multiplication factor to the Fourier space grid. This is a local function. More...
subroutine, public	cfourier_space_op_apply (this, cube, cf)
	Applies a multiplication factor to the Fourier space grid. This is a local function. More...

Variables
real(real64), parameter, public	vanishing_q = 1.0e-5_real64
	Vanishingly small q-vector. More...
real(real64), parameter	tol_vanishing_q = 1e-6_real64

Function/Subroutine Documentation

fourier_space_op_init()

subroutine fourier_space_oct_m::fourier_space_op_init ( class(fourier_space_op_t), intent(inout)  this, class(space_t), intent(in)  space, real(real64), dimension(:), intent(in)  qvector, type(xc_cam_t), intent(in)  cam, real(real64), intent(in), optional  singularity  )

private

Initialise a Fourier space Op instance.

Only the parameters defining the kernel are set here. The op buffer itself is installed by a kernel builder through dset_op/zset_op.

Note

This is not a Fortran finaliser-backed constructor: this must be freshly declared, or have been released with end(), before calling init.

Parameters

[in]	qvector	q-vector
[in]	cam	CAM parameters
[in]	singularity	Must be defined for all but the short-range Coulomb potential

Definition at line 189 of file fourier_space.F90.

fourier_space_op_end()

subroutine fourier_space_oct_m::fourier_space_op_end ( class(fourier_space_op_t), intent(inout)  this )

private

Finalize an instance of fourier_space_op_t.

Definition at line 213 of file fourier_space.F90.

dfourier_space_op_set_op()

subroutine fourier_space_oct_m::dfourier_space_op_set_op ( class(fourier_space_op_t), intent(inout)  this, type(cube_t), intent(in)  cube, real(real64), dimension(:, :, :), intent(in), optional, target  op, logical, intent(in), optional  in_device, real(real64), dimension(:, :, :), intent(inout), optional, allocatable, target  op_move  )

private

Set the op buffer, which defines the kernel on the cube, in Fourier space.

The operator can be provided either through op (which is copied) or through op_move (an allocatable that is taken over via move_alloc, avoiding a full-grid copy). Exactly one of the two must be present.

Definition at line 313 of file fourier_space.F90.

dfourier_space_op_apply()

subroutine, public fourier_space_oct_m::dfourier_space_op_apply	(	type(fourier_space_op_t), intent(in)	this,
		type(cube_t), intent(in)	cube,
		type(cube_function_t), intent(inout)	cf
	)

Applies a multiplication factor to the Fourier space grid. This is a local function.

One handle per multiplication kernel. The kernel is picked by the operator, not by the instantiation, so a single shared handle would keep running the first one built.

Definition at line 443 of file fourier_space.F90.

zfourier_space_op_set_op()

subroutine fourier_space_oct_m::zfourier_space_op_set_op ( class(fourier_space_op_t), intent(inout)  this, type(cube_t), intent(in)  cube, complex(real64), dimension(:, :, :), intent(in), optional, target  op, logical, intent(in), optional  in_device, complex(real64), dimension(:, :, :), intent(inout), optional, allocatable, target  op_move  )

private

Set the op buffer, which defines the kernel on the cube, in Fourier space.

The operator can be provided either through op (which is copied) or through op_move (an allocatable that is taken over via move_alloc, avoiding a full-grid copy). Exactly one of the two must be present.

Definition at line 673 of file fourier_space.F90.

zfourier_space_op_apply()

subroutine, public fourier_space_oct_m::zfourier_space_op_apply	(	type(fourier_space_op_t), intent(in)	this,
		type(cube_t), intent(in)	cube,
		type(cube_function_t), intent(inout)	cf
	)

Applies a multiplication factor to the Fourier space grid. This is a local function.

One handle per multiplication kernel. The kernel is picked by the operator, not by the instantiation, so a single shared handle would keep running the first one built.

Definition at line 803 of file fourier_space.F90.

sfourier_space_op_apply()

subroutine, public fourier_space_oct_m::sfourier_space_op_apply	(	type(fourier_space_op_t), intent(in)	this,
		type(cube_t), intent(in)	cube,
		type(cube_function_t), intent(inout)	cf
	)

Applies a multiplication factor to the Fourier space grid. This is a local function.

One handle per multiplication kernel. The kernel is picked by the operator, not by the instantiation, so a single shared handle would keep running the first one built.

Definition at line 1034 of file fourier_space.F90.

cfourier_space_op_apply()

subroutine, public fourier_space_oct_m::cfourier_space_op_apply	(	type(fourier_space_op_t), intent(in)	this,
		type(cube_t), intent(in)	cube,
		type(cube_function_t), intent(inout)	cf
	)

Applies a multiplication factor to the Fourier space grid. This is a local function.

One handle per multiplication kernel. The kernel is picked by the operator, not by the instantiation, so a single shared handle would keep running the first one built.

Definition at line 1238 of file fourier_space.F90.

Variable Documentation

vanishing_q

real(real64), parameter, public fourier_space_oct_m::vanishing_q = 1.0e-5_real64

Vanishingly small q-vector.

Definition at line 177 of file fourier_space.F90.

tol_vanishing_q

real(real64), parameter fourier_space_oct_m::tol_vanishing_q = 1e-6_real64

private

Definition at line 178 of file fourier_space.F90.