kaira.data.datasets.GaussianDataset

Inheritance diagram for GaussianDataset

class kaira.data.datasets.GaussianDataset(length: int, shape: int | Tuple[int, ...] = (128,), mean: float = 0.0, std: float = 1.0, seed: int | None = None)

Bases: Dataset

Dataset for Gaussian distributed tensor data.

Generates tensors with Gaussian distributed random values on-demand. Useful for noise modeling and channel simulation.

Methods

__init__

Initialize the Gaussian dataset.

__init__(length: int, shape: int | Tuple[int, ...] = (128,), mean: float = 0.0, std: float = 1.0, seed: int | None = None)

Initialize the Gaussian dataset.

Parameters:

• length – Number of samples in the dataset

• shape – Shape of each tensor (int for 1D, tuple for multi-dimensional)

• mean – Mean of the Gaussian distribution

• std – Standard deviation of the Gaussian distribution

• seed – Random seed for reproducibility