Attend the Women in ML Symposium on December 7 Register now

Module: tfp.experimental.sts_gibbs.sample_parameters

Functions for sampling parameters useful in Gibbs Sampling.


inverse_gamma module: The InverseGamma distribution class.

invert module: Invert bijector.

prefer_static module: Operations that use static values when possible.

square module: Square bijector.

transformed_distribution module: A Transformed Distribution class.


normal_scale_posterior_inverse_gamma_conjugate(...): Returns the conditional posterior of Normal scale given observations.

sample_with_optional_upper_bound(...): Samples from the given distribution with an optional upper bound.