util.prefix_condition()

Condition a (t+f)-length distribution on a t-length data prefix.

Usage

Source

util.prefix_condition(
    noise_dist,
    data,
)

For independent-over-time noise (the default) the conditional reduces to the forecast-horizon marginal, i.e. a time slice [t:]. Only independent families are supported today; correlated families (e.g. MultivariateNormal) would need a registered dispatch implementing a genuine Gaussian conditional, which is not yet provided.

Parameters

noise_dist: dist.Distribution

The observation distribution over the full horizon (*batch, t+f, obs).

data: Array

The observed prefix with shape (*batch, t, obs).

Returns

dist.Distribution

The forecast-horizon distribution over (*batch, f, obs).