Computes the shape of seasonal latents.
tfp.experimental.sts_gibbs.get_seasonal_latents_shape(
timeseries, model, num_chains=()
)
Args |
timeseries
|
Timeseries that is being modeled. Used to extract the timeseries
length and batch shape.
|
model
|
The sts.Sum model that the seasonal components will be found in.
Must be a Gibbs-samplable model built with
build_model_for_gibbs_fitting .
|
num_chains
|
Optional int to indicate the number of parallel MCMC chains.
Default to an empty tuple to sample a single chain.
|