MaskedPrior.create_likelihood_variable

MaskedPrior.create_likelihood_variable(name, *, mu, observed, xdist=False)

Create an observed variable over the active subset and expand to full dims.

Parameters:

namestr

Base name for the created variables.

muXTensorLike

Mean/location parameter broadcastable to the masked shape.

observedXTensorLike

Observations broadcastable to the masked shape.

Returns:

XTensorVariable

Deterministic variable over the full dims with observed RV on active entries.