ml4chem.backends package

Submodules

ml4chem.backends.available module

ml4chem.backends.available.available_backends()[source]: A function that reports available backends

ml4chem.backends.available.cuda_is_available(backend)[source]

Check if cuda is avaible for the current backend

Parameters:: backend (obj) – A backend object
Returns:: cuda_available – Whether or not cuda is there.
Return type:: bool

ml4chem.backends.operations module

class ml4chem.backends.operations.BackendOperations(backend)[source]

Bases: object

A class for centralizing backend operations

This class will be growing systematically. This is probably not the best solution but can be worked out later.

Parameters:: backend (object) – A backend object: numpy, tensorflow, or pytorch.

divide(a, b)[source]: Divide two vectors/tensors

dot(a, b)[source]: Dot product

exp(a)[source]: Exponential of a number

from_numpy(a)[source]: Convert from numpy to right data type

log10(a)[source]: Log base 10

logspace(a, b, num)[source]: Logspace

norm(a)[source]: Norm between two vectors

sum(a)[source]: Sum a list of values

to_numpy(a)[source]: Convert from numpy to right data type

Module contents