quemb.shared.typing

Define some types that do not fit into one particular module

In particular it enables barebone typechecking for the shape of numpy arrays

Inspired by https://stackoverflow.com/questions/75495212/type-hinting-numpy-arrays-and-batches

Note that most numpy functions return ndarray[Any, Any] i.e. the type is mostly useful to document intent to the developer.

Module Attributes

T_dtype_co	Type annotation of a generic covariant type.
Vector	Type annotation of a vector.
Matrix	Type annotation of a matrix.
Tensor3D	Type annotation of a tensor.
Tensor4D	Type annotation of a tensor.
Tensor5D	Type annotation of a tensor.
Tensor6D	Type annotation of a tensor.
Tensor7D	Type annotation of a tensor.
Tensor	Type annotation of a tensor.
PathLike	Type annotation for pathlike objects.
KwargDict	Type annotation for dictionaries holding keyword arguments.
T	A generic type variable, without any constraints.
AOIdx	The index of an atomic orbital.
GlobalAOIdx	The global index of an atomic orbital, i.e. not per fragment.
OwnRelAOIdx	The relative AO index.
OtherRelAOIdx	The relative AO index, relative to another fragment.
FragmentIdx	The index of a Fragment.
MotifIdx	The index of a heavy atom, i.e. of a motif.
CenterIdx	In a given fragment, this is the index of a center.
EdgeIdx	An edge is the complement of the set of centers in a fragment.
OriginIdx	In a given BE fragment, this is the origin of the remaining fragment after subsets have been removed.