cuqi.problem.BayesianProblem.UQ#

BayesianProblem.UQ(Ns=1000, Nb=None, percent=95, exact=None, experimental=False)#

Run an Uncertainty Quantification (UQ) analysis on the Bayesian problem and provide a summary of the results.

Parameters:

  • Ns (int, Optional) – Number of samples to draw. Defaults to 1000.

  • Nb (int, Optional) – Number of burn-in samples. If not provided, 20% of the samples will be used for burn-in.

  • exact (ndarray or dict[str, ndarray], Optional) – Exact solution to the problem. If provided the summary will include a comparison to the exact solution. If a dict is provided, the keys should be the names of the variables and the values should be the exact solution for each variable.

  • percent (float, Optional) – The credible interval to plot. Defaults to 95%.

  • experimental (bool, Optional) – If set to True, the sampler selection will use the samplers from the cuqi.experimental.mcmc module.

Returns:

samples – Samples from the posterior. The samples can be used to compute further statistics and plots.

Return type:

cuqi.samples.Samples