functional.mcmc.MCMCFit

The result of fitting a forecasting model with MCMC.

Usage

Source

functional.mcmc.MCMCFit(
    samples,
    num_chains=1,
)

Attributes

samples: dict[str, Array]

The posterior samples of the latent sites, sample axis leading and stored flattened (group_by_chain=False).

num_chains: int

The number of chains the fit was run with (frozen; default 1), so chain structure survives into ~numpyro_forecast.convert.to_datatree().