torch_sim.transforms

Coordinate transformations and periodic boundary conditions.

This module provides functions for handling coordinate transformations and periodic boundary conditions in molecular simulations, including matrix inversions and general PBC wrapping.

Functions

build_linked_cell_neighborhood	Build the neighbor list of a given set of atomic structures using the linked cell algorithm.
build_naive_neighborhood	Build a naive neighborhood list for atoms based on positions
compute_cell_shifts	Compute the cell shifts based on the provided indices and cell matrix.
compute_distances_with_cell_shifts	Compute distances between pairs of positions, optionally
get_cell_shift_idx	Generate the indices for cell shifts based on the number of repeats.
get_fractional_coordinates	Convert Cartesian coordinates to fractional coordinates.
get_fully_connected_mapping	Generate a fully connected mapping of atom indices with optional cell shifts.
get_linear_bin_idx	Calculate the linear bin index for each position within a defined box.
get_number_of_cell_repeats	Determine the number of cell repeats required for a given
get_pair_displacements	Compute displacement vectors and distances between atom pairs.
high_precision_sum	Sums tensor elements over specified dimensions at 64-bit precision.
inverse_box	Compute the inverse of an affine transformation.
linked_cell	Determine the atomic neighborhood of the atoms of a given structure for a particular cutoff using the linked cell algorithm.
minimum_image_displacement	Apply minimum image convention to displacement vectors.
multiplicative_isotropic_cutoff	Creates a smoothly truncated version of an isotropic function.
pbc_wrap_batched	Apply periodic boundary conditions to batched systems.
pbc_wrap_general	Apply periodic boundary conditions using lattice
ravel_3d	Convert 3D indices into linear indices for an array of given shape.
safe_mask	Safely applies a function to masked values in a tensor.
scatter_bin_index	Convert a linear table of bin indices into a structured bin ID table.
strides_of	Calculate the cumulative strides of a flattened tensor.
translate_pretty	ASE pretty translation in pytorch.
unravel_3d	Convert linear indices back into 3D indices for an array of given shape.
wrap_positions	ASE wrap_positions in pytorch.