States

A State in BayesBay is a data structure containing all numerical values associated with a certain Markov chain step. Depending on the inference problem at hand, it could contain information about:

• Discretization. For example, the nuclei positions in a Voronoi tessellation.

• Free parameters. For example, the value of the physical property to be inferred, associated with each Voronoi cell.

• Data noise. This includes standard deviation and noise correlation.

• Cached information. This consists of a Python dictionary containing arbitrary objects that the user might want to store and reuse at the next Markov chain step (see bayesbay.State.save_to_cache()).

bayesbay.State	Data structure that stores a Markov chain state, including all the necessary information to perform the forward operation
bayesbay.ParameterSpaceState	Data structure that stores the state of a parameter space.
bayesbay.DataNoiseState	Data structure that stores the state of the data noise parameters during the inference.

All examples in this documentation use bayesbay.State within a forward function. For usage tips, refer to our Quickstart Tutorial. The classes bayesbay.ParameterSpaceState and bayesbay.DataNoiseState are typically used under the hood.