Skip to content

Mean and count priors parameterizations for Beta and Dirichlet #1563

@bob-carpenter

Description

@bob-carpenter
  • add following parameterization of Beta
Beta_mean(theta | mu, phi) = Beta(theta | mu * phi, (1 - mu) * phi)

We could add it without any customized gradients first.

  • ditto for Dirichlet
Dirichlet_mean(theta | mu, phi) = Dirichlet(theta | mu * phi)
  • add a simple symmetric Dirichlet
Dirichlet_sym(theta | alpha) = Dirichlet(theta | rep(alpha, size(theta))
  • and a similar symmetric Beta
Beta_sym(theta | alpha) = Beta(theta | alpha, alpha)

The latter two will save memory by allowing simpler vari implementations.

  • go off and do this in Stan math with tests
  • expose functions in language

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions