summary_space_comparison

bayesflow.diagnostics.summary_space_comparison(observed_data: ~collections.abc.Mapping[str, ~numpy.ndarray], reference_data: ~collections.abc.Mapping[str, ~numpy.ndarray], approximator: ~bayesflow.approximators.continuous_approximator.ContinuousApproximator, num_null_samples: int = 100, comparison_fn: ~collections.abc.Callable = <function maximum_mean_discrepancy>, **kwargs) → tuple[float, ndarray]

Computes the distance between observed and reference data in the summary space and generates a distribution of distance values under the null hypothesis to assess model misspecification.

By default, the Maximum Mean Discrepancy (MMD) is used as a distance function.

[1] M. Schmitt, P.-C. Bürkner, U. Köthe, and S. T. Radev, “Detecting model misspecification in amortized Bayesian inference with neural networks,” arXiv e-prints, Dec. 2021, Art. no. arXiv:2112.08866. URL: https://arxiv.org/abs/2112.08866

Parameters:

observed_datadict[str, np.ndarray]

Dictionary of observed data as NumPy arrays, which will be preprocessed by the approximators adapter and passed through its summary network.

reference_datadict[str, np.ndarray]

Dictionary of reference data as NumPy arrays, which will be preprocessed by the approximators adapter and passed through its summary network.

approximatorContinuousApproximator

An instance of ContinuousApproximator used to compute summary statistics from the data.

num_null_samplesint, optional

Number of null samples to generate for hypothesis testing. Default is 100.

comparison_fnCallable, optional

Distance function to compare the data in the summary space.

**kwargsdict

Additional keyword arguments for the adapter and sampling process.

Returns:

distance_observedfloat

The MMD value between observed and reference summaries.

distance_nullnp.ndarray

A distribution of MMD values under the null hypothesis.

Raises:

ValueError

If approximator is not an instance of ContinuousApproximator or does not have a summary network.