deltakit.decode.noise_sources.UniformErasureNoise.importance_sampling_decomposition

UniformErasureNoise.importance_sampling_decomposition(code_data: CodeT, coefficient_limit: float = 1e-20) → list[tuple[MonteCarloNoise, float]]

Expresses the independent error distribution as a statistical mixture of other error distributions. This decomposition can be used for importance sampling.

Parameters:

• code_data (CodeT) – The code data needed to define the component noise sources.

• coefficient_limit (float, optional) – Only noise models with a probability of occurring above this limit will be kept in the decomposition.

Returns:

List of component noise sources, with normalised weightings.

Return type:

List[Tuple[MonteCarloNoise, float]]