Experimental CP2K-native GPW real-space-grid path for SKALA TorchScript models. More...

Functions/Subroutines
logical function, public	xc_section_uses_native_skala_grid (xc_section)
	Return true if the GAUXC subsection requests the CP2K-native GPW grid path.
integer function, public	skala_gapw_representation (xc_section)
	Return the pseudopotential GAPW representation selected for an active model.
logical function, public	xc_section_uses_native_skala_evaluator (xc_section)
	Return true when SKALA must be evaluated by the CP2K-native grid machinery.
logical function, public	native_skala_gapw_composite_reference (xc_section)
	Return true if native SKALA should use the full GAPW ORB density on one common grid.
logical function, public	native_skala_gapw_atom_composite_requested (xc_section)
	Return true when the explicit atom-centered composite reference is requested.
logical function, public	native_skala_uses_atom_composite_grid (xc_section)
	Return true when native Skala uses atom-centered grids.
logical function, public	native_skala_gapw_composite_direct_ao (xc_section)
	Return true if the GAPW composite reference uses direct full-ORB collocation.
logical function, public	xc_section_uses_gauxc_model (xc_section)
	Return true if the GAUXC subsection requests a model evaluation.
integer function, public	native_skala_gapw_density_partition (xc_section)
	Return the hard/soft GAPW one-center density partition for native SKALA.
subroutine, public	ensure_native_skala_grid_scope (xc_section)
	Enforce the currently implemented native SKALA GPW input scope.
subroutine, public	skala_gpw_eval (vxc_rho, vxc_tau, exc, rho_r, rho_g, tau, xc_section, weights, pw_pool, particle_set, cell, compute_virial, virial_xc, just_energy, atom_force)
	Evaluate SKALA energy and first derivatives on a CP2K GPW grid.
subroutine, public	skala_gpw_weight_derivative (weight_deriv_r, rho_r, rho_g, tau, xc_section, weights, pw_pool, particle_set, cell)
	Evaluate the derivative of native SKALA XC energy with respect to CP2K's external real-space integration-weight multiplier.
subroutine, public	skala_gapw_atom_vxc_of_r (xc_section, grid_atom, group, atom_coord, rho, drho, tau, weights, lsd, nspins, na, nr, exc, vxc, vxg, vtau, energy_only, atom_force, atom_virial)
	Evaluate SKALA on a GAPW one-center atomic grid.
subroutine, public	skala_gapw_atom_composite_energy (xc_section, group, density, grad, kin, grid_coords, grid_weights, atomic_grid_weights, atomic_grid_sizes, atomic_coords, exc, density_grad_out, grad_grad_out, kin_grad_out, grid_coord_grad_out, grid_weight_grad_out, atomic_grid_weight_grad_out, atom_coord_grad_out)
	Evaluate a rank-local set of complete atom blocks and sum their SKALA energies.
subroutine, public	build_vxc_from_feature_grads (vxc_rho, vxc_tau, rho_r, pw_pool, density_grad, grad_grad, kin_grad, xc_deriv_method_id, global_grid_layout)
	Fill CP2K VXC real-space arrays from Torch feature gradients.
type(section_vals_type) function, pointer, public	get_gauxc_section (xc_section)
	Return the first GAUXC functional subsection, if present.

Variables
integer, parameter, public	skala_gapw_density_partition_hard_minus_soft = 1
integer, parameter, public	skala_gapw_density_partition_hard_only = 2
integer, parameter, public	skala_gapw_density_partition_soft_only = 3
integer, parameter, public	skala_gapw_density_partition_none = 4
type(skala_torch_model_type), save	cached_model

Detailed Description

Experimental CP2K-native GPW real-space-grid path for SKALA TorchScript models.

Function/Subroutine Documentation

◆ xc_section_uses_native_skala_grid()

logical function, public skala_gpw_functional::xc_section_uses_native_skala_grid ( type(section_vals_type), intent(in), pointer xc_section )

Return true if the GAUXC subsection requests the CP2K-native GPW grid path.

Parameters

xc_section ...

Returns: ...

Definition at line 115 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ skala_gapw_representation()

integer function, public skala_gpw_functional::skala_gapw_representation ( type(section_vals_type), intent(in), pointer xc_section )

Return the pseudopotential GAPW representation selected for an active model.

Parameters

xc_section ...

Returns: ...

Definition at line 134 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xc_section_uses_native_skala_evaluator()

logical function, public skala_gpw_functional::xc_section_uses_native_skala_evaluator ( type(section_vals_type), intent(in), pointer xc_section )

Return true when SKALA must be evaluated by the CP2K-native grid machinery.

Parameters

xc_section ...

Returns: ...

Definition at line 163 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ native_skala_gapw_composite_reference()

logical function, public skala_gpw_functional::native_skala_gapw_composite_reference ( type(section_vals_type), intent(in), pointer xc_section )

Return true if native SKALA should use the full GAPW ORB density on one common grid.

Parameters

xc_section ...

Returns: ...

Definition at line 178 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ native_skala_gapw_atom_composite_requested()

logical function, public skala_gpw_functional::native_skala_gapw_atom_composite_requested ( type(section_vals_type), intent(in), pointer xc_section )

Return true when the explicit atom-centered composite reference is requested.

Parameters

xc_section ...

Returns: ...

Definition at line 202 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ native_skala_uses_atom_composite_grid()

logical function, public skala_gpw_functional::native_skala_uses_atom_composite_grid ( type(section_vals_type), intent(in), pointer xc_section )

Return true when native Skala uses atom-centered grids.

Parameters

xc_section ...

Returns: ...

Definition at line 227 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ native_skala_gapw_composite_direct_ao()

logical function, public skala_gpw_functional::native_skala_gapw_composite_direct_ao ( type(section_vals_type), intent(in), pointer xc_section )

Return true if the GAPW composite reference uses direct full-ORB collocation.

Parameters

xc_section ...

Returns: ...

Definition at line 257 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xc_section_uses_gauxc_model()

logical function, public skala_gpw_functional::xc_section_uses_gauxc_model ( type(section_vals_type), intent(in), pointer xc_section )

Return true if the GAUXC subsection requests a model evaluation.

Parameters

xc_section ...

Returns: ...

Definition at line 278 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ native_skala_gapw_density_partition()

integer function, public skala_gpw_functional::native_skala_gapw_density_partition ( type(section_vals_type), intent(in), pointer xc_section )

Return the hard/soft GAPW one-center density partition for native SKALA.

Parameters

xc_section ...

Returns: ...

Definition at line 305 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ ensure_native_skala_grid_scope()

subroutine, public skala_gpw_functional::ensure_native_skala_grid_scope ( type(section_vals_type), intent(in), pointer xc_section )

Enforce the currently implemented native SKALA GPW input scope.

Parameters

xc_section ...

Definition at line 335 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ skala_gpw_eval()

subroutine, public skala_gpw_functional::skala_gpw_eval	(	type(pw_r3d_rs_type), dimension(:), pointer	vxc_rho,
		type(pw_r3d_rs_type), dimension(:), pointer	vxc_tau,
		real(kind=dp), intent(out)	exc,
		type(pw_r3d_rs_type), dimension(:), pointer	rho_r,
		type(pw_c1d_gs_type), dimension(:), pointer	rho_g,
		type(pw_r3d_rs_type), dimension(:), pointer	tau,
		type(section_vals_type), pointer	xc_section,
		type(pw_r3d_rs_type), pointer	weights,
		type(pw_pool_type), pointer	pw_pool,
		type(particle_type), dimension(:), pointer	particle_set,
		type(cell_type), pointer	cell,
		logical, intent(in)	compute_virial,
		real(kind=dp), dimension(3, 3), intent(out)	virial_xc,
		logical, intent(in), optional	just_energy,
		real(kind=dp), dimension(:, :), intent(out), optional	atom_force )

Evaluate SKALA energy and first derivatives on a CP2K GPW grid.

Parameters

vxc_rho	...
vxc_tau	...
exc	...
rho_r	...
rho_g	...
tau	...
xc_section	...
weights	...
pw_pool	...
particle_set	...
cell	...
compute_virial	...
virial_xc	...
just_energy	...
atom_force	...

Definition at line 401 of file skala_gpw_functional.F.

Here is the call graph for this function:

◆ skala_gpw_weight_derivative()

subroutine, public skala_gpw_functional::skala_gpw_weight_derivative	(	type(pw_r3d_rs_type), intent(inout)	weight_deriv_r,
		type(pw_r3d_rs_type), dimension(:), pointer	rho_r,
		type(pw_c1d_gs_type), dimension(:), pointer	rho_g,
		type(pw_r3d_rs_type), dimension(:), pointer	tau,
		type(section_vals_type), pointer	xc_section,
		type(pw_r3d_rs_type), pointer	weights,
		type(pw_pool_type), pointer	pw_pool,
		type(particle_type), dimension(:), pointer	particle_set,
		type(cell_type), pointer	cell )

Evaluate the derivative of native SKALA XC energy with respect to CP2K's external real-space integration-weight multiplier.

Parameters

weight_deriv_r	...
rho_r	...
rho_g	...
tau	...
xc_section	...
weights	...
pw_pool	...
particle_set	...
cell	...

Definition at line 714 of file skala_gpw_functional.F.

Here is the call graph for this function:

◆ skala_gapw_atom_vxc_of_r()

subroutine, public skala_gpw_functional::skala_gapw_atom_vxc_of_r	(	type(section_vals_type), pointer	xc_section,
		type(grid_atom_type), pointer	grid_atom,
		class(mp_comm_type), intent(in)	group,
		real(kind=dp), dimension(3), intent(in)	atom_coord,
		real(kind=dp), dimension(:, :, :), pointer	rho,
		real(kind=dp), dimension(:, :, :, :), pointer	drho,
		real(kind=dp), dimension(:, :, :), pointer	tau,
		real(kind=dp), dimension(:, :), intent(in)	weights,
		logical, intent(in)	lsd,
		integer, intent(in)	nspins,
		integer, intent(in)	na,
		integer, intent(in)	nr,
		real(kind=dp), intent(out)	exc,
		real(kind=dp), dimension(:, :, :), pointer	vxc,
		real(kind=dp), dimension(:, :, :, :), pointer	vxg,
		real(kind=dp), dimension(:, :, :), pointer	vtau,
		logical, intent(in), optional	energy_only,
		real(kind=dp), dimension(3), intent(out), optional	atom_force,
		real(kind=dp), dimension(3, 3), intent(out), optional	atom_virial )

Evaluate SKALA on a GAPW one-center atomic grid.

Parameters

xc_section	...
grid_atom	...
group	...
atom_coord	...
rho	...
drho	...
tau	...
weights	...
lsd	...
nspins	...
na	...
nr	...
exc	...
vxc	...
vxg	...
vtau	...
energy_only	...
atom_force	...
atom_virial	...

Definition at line 891 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ skala_gapw_atom_composite_energy()

subroutine, public skala_gpw_functional::skala_gapw_atom_composite_energy	(	type(section_vals_type), pointer	xc_section,
		class(mp_comm_type), intent(in)	group,
		real(kind=dp), dimension(:, :), intent(in), allocatable	density,
		real(kind=dp), dimension(:, :, :), intent(in), allocatable	grad,
		real(kind=dp), dimension(:, :), intent(in), allocatable	kin,
		real(kind=dp), dimension(:, :), intent(in), allocatable	grid_coords,
		real(kind=dp), dimension(:), intent(in), allocatable	grid_weights,
		real(kind=dp), dimension(:), intent(in), allocatable	atomic_grid_weights,
		integer(kind=int_8), dimension(:), intent(in), allocatable	atomic_grid_sizes,
		real(kind=dp), dimension(:, :), intent(in), allocatable	atomic_coords,
		real(kind=dp), intent(out)	exc,
		real(kind=dp), dimension(:, :), intent(out), optional, allocatable	density_grad_out,
		real(kind=dp), dimension(:, :, :), intent(out), optional, allocatable	grad_grad_out,
		real(kind=dp), dimension(:, :), intent(out), optional, allocatable	kin_grad_out,
		real(kind=dp), dimension(:, :), intent(out), optional, allocatable	grid_coord_grad_out,
		real(kind=dp), dimension(:), intent(out), optional, allocatable	grid_weight_grad_out,
		real(kind=dp), dimension(:), intent(out), optional, allocatable	atomic_grid_weight_grad_out,
		real(kind=dp), dimension(:, :), intent(out), optional, allocatable	atom_coord_grad_out )

Evaluate a rank-local set of complete atom blocks and sum their SKALA energies.

Parameters

xc_section	...
group	...
density	...
grad	...
kin	...
grid_coords	...
grid_weights	...
atomic_grid_weights	...
atomic_grid_sizes	...
atomic_coords	...
exc	...
density_grad_out	...
grad_grad_out	...
kin_grad_out	...
grid_coord_grad_out	...
grid_weight_grad_out	...
atomic_grid_weight_grad_out	...
atom_coord_grad_out	...

Definition at line 1150 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ build_vxc_from_feature_grads()

subroutine, public skala_gpw_functional::build_vxc_from_feature_grads	(	type(pw_r3d_rs_type), dimension(:), pointer	vxc_rho,
		type(pw_r3d_rs_type), dimension(:), pointer	vxc_tau,
		type(pw_r3d_rs_type), dimension(:), pointer	rho_r,
		type(pw_pool_type), pointer	pw_pool,
		real(kind=dp), dimension(:, :), intent(in)	density_grad,
		real(kind=dp), dimension(:, :, :), intent(in)	grad_grad,
		real(kind=dp), dimension(:, :), intent(in)	kin_grad,
		integer, intent(in)	xc_deriv_method_id,
		logical, intent(in), optional	global_grid_layout )

Fill CP2K VXC real-space arrays from Torch feature gradients.

Parameters

vxc_rho	...
vxc_tau	...
rho_r	...
pw_pool	...
density_grad	...
grad_grad	...
kin_grad	...
xc_deriv_method_id	...
global_grid_layout	...

Definition at line 2713 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ get_gauxc_section()

type(section_vals_type) function, pointer, public skala_gpw_functional::get_gauxc_section ( type(section_vals_type), intent(in), pointer xc_section )

Return the first GAUXC functional subsection, if present.

Parameters

xc_section ...

Returns: ...

Definition at line 3025 of file skala_gpw_functional.F.

Here is the call graph for this function:

Here is the caller graph for this function:

Variable Documentation

◆ skala_gapw_density_partition_hard_minus_soft

integer, parameter, public skala_gpw_functional::skala_gapw_density_partition_hard_minus_soft = 1

Definition at line 78 of file skala_gpw_functional.F.

◆ skala_gapw_density_partition_hard_only

integer, parameter, public skala_gpw_functional::skala_gapw_density_partition_hard_only = 2

Definition at line 78 of file skala_gpw_functional.F.

◆ skala_gapw_density_partition_soft_only

integer, parameter, public skala_gpw_functional::skala_gapw_density_partition_soft_only = 3

Definition at line 78 of file skala_gpw_functional.F.

◆ skala_gapw_density_partition_none

integer, parameter, public skala_gpw_functional::skala_gapw_density_partition_none = 4

Definition at line 78 of file skala_gpw_functional.F.

◆ cached_model

type(skala_torch_model_type), save skala_gpw_functional::cached_model

Definition at line 98 of file skala_gpw_functional.F.

Functions/Subroutines

Variables

Detailed Description

Function/Subroutine Documentation

◆ xc_section_uses_native_skala_grid()

◆ skala_gapw_representation()

◆ xc_section_uses_native_skala_evaluator()

◆ native_skala_gapw_composite_reference()

◆ native_skala_gapw_atom_composite_requested()

◆ native_skala_uses_atom_composite_grid()

◆ native_skala_gapw_composite_direct_ao()

◆ xc_section_uses_gauxc_model()

◆ native_skala_gapw_density_partition()

◆ ensure_native_skala_grid_scope()

◆ skala_gpw_eval()

◆ skala_gpw_weight_derivative()

◆ skala_gapw_atom_vxc_of_r()

◆ skala_gapw_atom_composite_energy()

◆ build_vxc_from_feature_grads()

◆ get_gauxc_section()

Variable Documentation

◆ skala_gapw_density_partition_hard_minus_soft

◆ skala_gapw_density_partition_hard_only

◆ skala_gapw_density_partition_soft_only

◆ skala_gapw_density_partition_none

◆ cached_model