(git:a660c7f)
Loading...
Searching...
No Matches
skala_gpw_features Module Reference

Build SKALA TorchScript feature dictionaries from CP2K GPW real-space grids. More...

Data Types

type  skala_gpw_feature_type
 

Functions/Subroutines

subroutine, public skala_gpw_feature_build (features, rho_set, rho_r, particle_set, cell, requires_grad, weights, requires_coordinate_grad, requires_stress_grad, use_atom_chunks, route_atom_chunks, atom_partition)
 Build a flat SKALA molecular feature dictionary from a local GPW grid.
 
subroutine, public skala_gpw_feature_release (features)
 Release Torch objects and backing arrays owned by a feature bundle.
 
integer function, public skala_gpw_atom_subchunk_count (max_rows)
 Return how many atom-contiguous subchunks the cached rank chunk needs.
 
subroutine, public skala_gpw_feature_build_atom_subchunk (parent, features, subchunk_index, max_rows, requires_grad)
 Build an atom-contiguous subchunk feature bundle from a rank-local atom chunk.
 
subroutine, public skala_gpw_smooth_partition_derivatives (grid_point, atom_coords, cell, weights, included, dweights_datom, dweights_dstrain)
 Build smooth atom weights and their atom/cell deformation derivatives.
 

Variables

integer, parameter, public skala_gpw_atom_partition_hard = 1
 
integer, parameter, public skala_gpw_atom_partition_smooth = 2
 
type(skala_gpw_layout_cache_type), save cached_layout
 

Detailed Description

Build SKALA TorchScript feature dictionaries from CP2K GPW real-space grids.

Function/Subroutine Documentation

◆ skala_gpw_feature_build()

subroutine, public skala_gpw_features::skala_gpw_feature_build ( type(skala_gpw_feature_type), intent(inout)  features,
type(xc_rho_set_type), intent(in)  rho_set,
type(pw_r3d_rs_type), dimension(:), intent(in)  rho_r,
type(particle_type), dimension(:), pointer  particle_set,
type(cell_type), pointer  cell,
logical, intent(in), optional  requires_grad,
type(pw_r3d_rs_type), optional, pointer  weights,
logical, intent(in), optional  requires_coordinate_grad,
logical, intent(in), optional  requires_stress_grad,
logical, intent(in), optional  use_atom_chunks,
logical, intent(in), optional  route_atom_chunks,
integer, intent(in), optional  atom_partition 
)

Build a flat SKALA molecular feature dictionary from a local GPW grid.

Parameters
features...
rho_set...
rho_r...
particle_set...
cell...
requires_grad...
weights...
requires_coordinate_grad...
requires_stress_grad...
use_atom_chunks...
route_atom_chunks...
atom_partition...

Definition at line 196 of file skala_gpw_features.F.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ skala_gpw_feature_release()

subroutine, public skala_gpw_features::skala_gpw_feature_release ( type(skala_gpw_feature_type), intent(inout)  features)

Release Torch objects and backing arrays owned by a feature bundle.

Parameters
features...

Definition at line 1552 of file skala_gpw_features.F.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ skala_gpw_atom_subchunk_count()

integer function, public skala_gpw_features::skala_gpw_atom_subchunk_count ( integer, intent(in)  max_rows)

Return how many atom-contiguous subchunks the cached rank chunk needs.

Parameters
max_rows...
Returns
...

Definition at line 1656 of file skala_gpw_features.F.

Here is the caller graph for this function:

◆ skala_gpw_feature_build_atom_subchunk()

subroutine, public skala_gpw_features::skala_gpw_feature_build_atom_subchunk ( type(skala_gpw_feature_type), intent(in)  parent,
type(skala_gpw_feature_type), intent(inout)  features,
integer, intent(in)  subchunk_index,
integer, intent(in)  max_rows,
logical, intent(in)  requires_grad 
)

Build an atom-contiguous subchunk feature bundle from a rank-local atom chunk.

Parameters
parent...
features...
subchunk_index...
max_rows...
requires_grad...

Definition at line 1692 of file skala_gpw_features.F.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ skala_gpw_smooth_partition_derivatives()

subroutine, public skala_gpw_features::skala_gpw_smooth_partition_derivatives ( real(kind=dp), dimension(3), intent(in)  grid_point,
real(kind=dp), dimension(:, :), intent(in)  atom_coords,
type(cell_type), pointer  cell,
real(kind=dp), dimension(:), intent(out)  weights,
logical, dimension(:), intent(out)  included,
real(kind=dp), dimension(:, :, :), intent(out)  dweights_datom,
real(kind=dp), dimension(:, :, :), intent(out)  dweights_dstrain 
)

Build smooth atom weights and their atom/cell deformation derivatives.

Parameters
grid_point...
atom_coords...
cell...
weights...
included...
dweights_datom...
dweights_dstrain...

Definition at line 2205 of file skala_gpw_features.F.

Here is the caller graph for this function:

Variable Documentation

◆ skala_gpw_atom_partition_hard

integer, parameter, public skala_gpw_features::skala_gpw_atom_partition_hard = 1

Definition at line 38 of file skala_gpw_features.F.

◆ skala_gpw_atom_partition_smooth

integer, parameter, public skala_gpw_features::skala_gpw_atom_partition_smooth = 2

Definition at line 38 of file skala_gpw_features.F.

◆ cached_layout

type(skala_gpw_layout_cache_type), save skala_gpw_features::cached_layout

Definition at line 177 of file skala_gpw_features.F.