digeo.optim.MeshLossFunc#

class digeo.optim.MeshLossFunc[source]#

Bases: ABC

abstractmethod compute(mesh, points)[source]#

Loss function used in the mesh optimisers.

Parameters:

  • mesh (Mesh) – The mesh used

  • points (MeshPointBatch) – The points on the mesh that are being optimized

Returns:

  • loss (float) – The loss of the function

  • gradient (Tensor) – The gradients in 3d space for each point

Return type:

Tuple[float, Tensor]

get_logs()[source]#

Optionnal method to return any additional information that should be logged during optimization, such as the value of certain terms in the loss function, or any other relevant information.

Returns:

logs – A dictionary containing the information to log, where the keys are the names of the metrics and the values are the corresponding values to log.

Return type:

dict[str, Any]