tfep.nn.flows.centroid.CenteredCentroidFlow

class tfep.nn.flows.centroid.CenteredCentroidFlow(flow: Module, space_dimension: int, subset_point_indices: Sequence[int] | None = None, weights: Sequence[float] | None = None, fixed_point_idx: int = 0, origin: Sequence[float] | None = None, translate_back: bool = True, return_partial: bool = False)

Bases: PartialFlow

A transformation constraining the (weighted) centroid of the DOFs.

This flow performs the following operations: - It computes the (optionally weighted) centroid of the input features (or

a subset of them).

• Translates the system to move the centroid to the origin or a user-defined point.

• Removes N degrees of freedom (DOFs) from the input features (where N is the dimensionality of a point in space) and runs the remaining features through the wrapped flow.

• Re-adds the N constrained DOFs so that the flow preserves the centroid.

• Optionally, translates the system to restore the original coordinates of the centroid.

The flow assumes that the input features have shape (batch_size, n_dofs), and each batch sample represent a sequence of points in an N-dimensional space. The points must must be listed so that input[b][N*i:N*i+N] are the coordinates of the i-th point for batch sample b.

Parameters:

• flow (torch.nn.Module) – The wrapped flow.

• space_dimension (int) – The dimensionality of a single point in space (e.g., 3 if the input features represent points in a 3D space).

• subset_point_indices (array-like of int, optional) – A vector of shape (n_points,). If passed, the centroid is computed over a subset of n_points points. If not passed, the centroid is computed using all the input features.

  Note that the indices must refer to the points, not the feature indices. For example, in a 3D space, subset_point_indices = [1, 3] will compute the centroid using the feature indices [3, 4, 5, 9, 10, 11].

• weights (array-like of floats, optional) – A vector of shape (n_points,), where n_points is the number of points used to compute the centroid (which depends on subset_point_indices). weights[i] is the (unnormalized) weight used for the i-th point. If not passed, the center of geometry is computed. This can be used for example to center the system on the center of mass.

• fixed_point_idx (int, optional) – The index of the point that is fixed and does not go through the flow. Note that this refers to the index of the point, not the input feature. Also, if subset_point_indices is passed, this index is relative to the subset of atoms used to compute the centroid.

  For example, if fixed_point_idx is 1 and subset_point_indices is None, the fixed DOFs will be [3, 4, 5] in a 3D space. However, if subset_point_indices = [1, 2] then the fixed DOFs will be [6, 7, 8].

• origin (array-like of floats, optional) – A vector of shape (space_dimension,). If the centroid must be moved to a point different than zero.

• translate_back (bool, optional) – If False, the output configuration has the centroid in the origin. Otherwise, it the centroid is restored to the original position.

• return_partial (bool, optional) – If True, only the propagated indices are returned.

Variables:

• flow (torch.nn.Module) – The wrapped flow.

• origin (Tensor of floats) – The position of the origin where the centroid is translated before going through the flow.

• translate_back (bool) – Whether the centroid is restored to its original position in the output configuration or if it is left at the origin.

• return_partial (bool) – If True, only the propagated indices are returned.

__init__(flow: Module, space_dimension: int, subset_point_indices: Sequence[int] | None = None, weights: Sequence[float] | None = None, fixed_point_idx: int = 0, origin: Sequence[float] | None = None, translate_back: bool = True, return_partial: bool = False)

Initialize internal Module state, shared by both nn.Module and ScriptModule.

Methods

__init__(flow, space_dimension[, ...])	Initialize internal Module state, shared by both nn.Module and ScriptModule.
add_module(name, module)	Add a child module to the current module.
apply(fn)	Apply fn recursively to every submodule (as returned by .children()) as well as self.
bfloat16()	Casts all floating point parameters and buffers to bfloat16 datatype.
buffers([recurse])	Return an iterator over module buffers.
children()	Return an iterator over immediate children modules.
compile(*args, **kwargs)	Compile this Module's forward using torch.compile().
cpu()	Move all model parameters and buffers to the CPU.
cuda([device])	Move all model parameters and buffers to the GPU.
double()	Casts all floating point parameters and buffers to double datatype.
eval()	Set the module in evaluation mode.
extra_repr()	Set the extra representation of the module.
float()	Casts all floating point parameters and buffers to float datatype.
forward(x)	Transform the input configuration.
get_buffer(target)	Return the buffer given by target if it exists, otherwise throw an error.
get_extra_state()	Return any extra state to include in the module's state_dict.
get_parameter(target)	Return the parameter given by target if it exists, otherwise throw an error.
get_submodule(target)	Return the submodule given by target if it exists, otherwise throw an error.
half()	Casts all floating point parameters and buffers to half datatype.
inverse(y)	Invert the forward transformation.
ipu([device])	Move all model parameters and buffers to the IPU.
load_state_dict(state_dict[, strict, assign])	Copy parameters and buffers from state_dict into this module and its descendants.
modules()	Return an iterator over all modules in the network.
mtia([device])	Move all model parameters and buffers to the MTIA.
n_parameters()	int: The total number of parameters that can be optimized.
named_buffers([prefix, recurse, ...])	Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
named_children()	Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.
named_modules([memo, prefix, remove_duplicate])	Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.
named_parameters([prefix, recurse, ...])	Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
parameters([recurse])	Return an iterator over module parameters.
register_backward_hook(hook)	Register a backward hook on the module.
register_buffer(name, tensor[, persistent])	Add a buffer to the module.
register_forward_hook(hook, *[, prepend, ...])	Register a forward hook on the module.
register_forward_pre_hook(hook, *[, ...])	Register a forward pre-hook on the module.
register_full_backward_hook(hook[, prepend])	Register a backward hook on the module.
register_full_backward_pre_hook(hook[, prepend])	Register a backward pre-hook on the module.
register_load_state_dict_post_hook(hook)	Register a post-hook to be run after module's load_state_dict() is called.
register_load_state_dict_pre_hook(hook)	Register a pre-hook to be run before module's load_state_dict() is called.
register_module(name, module)	Alias for add_module().
register_parameter(name, param)	Add a parameter to the module.
register_state_dict_post_hook(hook)	Register a post-hook for the state_dict() method.
register_state_dict_pre_hook(hook)	Register a pre-hook for the state_dict() method.
requires_grad_([requires_grad])	Change if autograd should record operations on parameters in this module.
set_extra_state(state)	Set extra state contained in the loaded state_dict.
set_submodule(target, module)	Set the submodule given by target if it exists, otherwise throw an error.
share_memory()	See torch.Tensor.share_memory_().
state_dict(*args[, destination, prefix, ...])	Return a dictionary containing references to the whole state of the module.
to(*args, **kwargs)	Move and/or cast the parameters and buffers.
to_empty(*, device[, recurse])	Move the parameters and buffers to the specified device without copying storage.
train([mode])	Set the module in training mode.
type(dst_type)	Casts all parameters and buffers to dst_type.
xpu([device])	Move all model parameters and buffers to the XPU.
zero_grad([set_to_none])	Reset gradients of all model parameters.

Attributes

T_destination
call_super_init
dump_patches
space_dimension	The dimensionality of a single point in space.
flow	Wrapped flow.
return_partial	If True, only the propagated indices are returned.
training

flow

Wrapped flow.

forward(x: Tensor) → Tuple[Tensor]

Transform the input configuration.

inverse(y: Tensor) → Tuple[Tensor]

Invert the forward transformation.

This works only if the forward transformation was performed with translate_back set to True.

n_parameters()

int: The total number of parameters that can be optimized.

return_partial

If True, only the propagated indices are returned.

property space_dimension

The dimensionality of a single point in space.

For example, 3 if the input features represent points in a 3D space.

Type:

int