x3cflux.run_non_uniform_sampling

class x3cflux.run_non_uniform_sampling(simulator, num_samples: int, starting_point: ~numpy.ndarray = None, bounds: ~typing.Dict[str, ~typing.Tuple[float, float]] = None, num_chains: int = 1, proposal: ~hopsy.core.PyProposal = <class 'hopsy.core.GaussianCoordinateHitAndRunProposal'>, random_seed: int = 42, **kwargs)

Run non-uniform sampling to estimate the posterior distribution induced by the given labeling data. Markov Chain Monte Carlo (MCMC) as implemented in the Polytope sampling toolbox hopsy is used and can be configured by passing appropriate kwargs. See Paul, R. et al. (2024), https://doi.org/10.1093/bioinformatics/btae430.

Parameters:

• simulator – Labeling simulator

• num_samples – Number of samples to generate

• starting_point – Point from where to start sampling

• bounds – Parameter boundary constraints

• num_chains – Number of Markov chains to run

• proposal – Proposal to use for Metropolis-Hastings

• random_seed – Random seed to initialize the Markov chain

• kwargs – Pass hopsy options as kwargs. See https://modsim.github.io/hopsy/generated/hopsy.sample.html.

Returns:

Generated samples as tensor (C, N, M), where C, N, M are the numbers of chains, parameters and samples