LocScalePTM.init_dist()

LocScalePTM.init_dist()#

LocScalePTM.init_dist(samples, loc=None, scale=None, newdata=None)[source]#

Construct a batched distribution from posterior samples.

Parameters:

  • samples (dict[str, Any]) – Posterior samples dict used to build the distribution.

  • loc (Any | None) – Optional explicit loc/scale arrays; if provided, newdata is ignored. (default: None)

  • scale (Any | None) – Optional explicit loc/scale arrays; if provided, newdata is ignored. (default: None)

  • newdata (dict[str, Any] | None) – Optional newdata for prediction when loc/scale are not provided. (default: None)

Return type:

LocScaleTransformationDist

Returns:

LocScaleTransformationDist – A batched transformation distribution for prediction.

Contents