Operator

continuiti.operators.operator

In continuiti, all models for operator learning are based on the Operator base class.

Operator(shapes=None, device=None)

Bases: Module, ABC

Operator base class.

An operator is a neural network model that maps functions by mapping an observation to the evaluations of the mapped function at given coordinates.

PARAMETER	DESCRIPTION
shapes	Operator shapes. TYPE: Optional[OperatorShapes] DEFAULT: None
device	Device. TYPE: Optional[device] DEFAULT: None

ATTRIBUTE	DESCRIPTION
shapes	Operator shapes. TYPE: OperatorShapes

Source code in src/continuiti/operators/operator.py

def __init__(
    self,
    shapes: Optional[OperatorShapes] = None,
    device: Optional[torch.device] = None,
):
    super().__init__()
    self.shapes = shapes
    self.device = device

forward(x, u, y) abstractmethod

Forward pass through the operator.

PARAMETER	DESCRIPTION
x	Sensor positions of shape (batch_size, x_dim, num_sensors...). TYPE: Tensor
u	Input function values of shape (batch_size, u_dim, num_sensors...). TYPE: Tensor
y	Evaluation coordinates of shape (batch_size, y_dim, num_evaluations...). TYPE: Tensor

RETURNS	DESCRIPTION
Tensor	Evaluations of the mapped function with shape (batch_size, v_dim, num_evaluations...).

Source code in src/continuiti/operators/operator.py

@abstractmethod
def forward(
    self, x: torch.Tensor, u: torch.Tensor, y: torch.Tensor
) -> torch.Tensor:
    """Forward pass through the operator.

    Args:
        x: Sensor positions of shape (batch_size, x_dim, num_sensors...).
        u: Input function values of shape (batch_size, u_dim, num_sensors...).
        y: Evaluation coordinates of shape (batch_size, y_dim, num_evaluations...).

    Returns:
        Evaluations of the mapped function with shape (batch_size, v_dim, num_evaluations...).
    """

save(path)

Save the operator to a file.

PARAMETER	DESCRIPTION
path	Path to the file. TYPE: str

Source code in src/continuiti/operators/operator.py

def save(self, path: str):
    """Save the operator to a file.

    Args:
        path: Path to the file.
    """
    torch.save(self.state_dict(), path)

load(path)

Load the operator from a file.

PARAMETER	DESCRIPTION
path	Path to the file. TYPE: str

Source code in src/continuiti/operators/operator.py

def load(self, path: str):
    """Load the operator from a file.

    Args:
        path: Path to the file.
    """
    self.load_state_dict(torch.load(path))

num_params()

Return the number of trainable parameters.

Source code in src/continuiti/operators/operator.py

def num_params(self) -> int:
    """Return the number of trainable parameters."""
    return sum(p.numel() for p in self.parameters())

__str__()

Return string representation of the operator.

Source code in src/continuiti/operators/operator.py

def __str__(self):
    """Return string representation of the operator."""
    return self.__class__.__name__

---

Last update: 2024-08-20
Created: 2024-08-20