Module: tfp.experimental.sts_gibbs

Gibbs sampling inference for structural time series models.

Classes

class GibbsSamplerState: GibbsSamplerState(observation_noise_scale, level_scale, weights, level, seed, slope_scale, slope, seasonal_drift_scales, seasonal_levels)

Functions

build_model_for_gibbs_fitting(...): Builds a StructuralTimeSeries model instance that supports Gibbs sampling.

fit_with_gibbs_sampling(...): Fits parameters for an STS model using Gibbs sampling.

get_seasonal_latents_shape(...): Computes the shape of seasonal latents.

one_step_predictive(...): Constructs a one-step-ahead predictive distribution at every timestep.