obsweatherscale.data

Submodules

• obsweatherscale.data.dataset

Classes

GPDataset

Abstract base class for Gaussian Process datasets.

Package Contents

class GPDataset

Bases: torch.utils.data.Dataset, abc.ABC

Abstract base class for Gaussian Process datasets.

This class extends PyTorch’s Dataset class to provide a standardized interface for datasets used in Gaussian Process models. It defines required methods for data access, transformation, and device management.

All subclasses must implement the abstract methods defined here.

See also

torch.utils.data.Dataset

PyTorch’s base dataset class

abstract get_dataset() → tuple[torch.Tensor, Ellipsis]

Get the entire dataset as tensors.

Returns:

A tuple of tensors representing the entire dataset. Typically contains input features and target values.

Return type:

tuple[torch.Tensor, …]

abstract to(device: torch.device) → None

Move the dataset to the specified device.

Parameters:

device (torch.device) – The device to move the dataset to (e.g., CPU or CUDA).

Returns:

The dataset instance moved to the specified device.

Return type:

self