Skip to contents

Compile an unnormalised Bayesian posterior into a gmm_target

Source: R/from_posterior.R
gmm_target_from_posterior.Rd

Generic S3 constructor that turns a Bayesian posterior — represented either by a fitted model object (e.g. from brms or Stan) or by a bare callable — into a gmm_target suitable for regime (iii) of fit_proxymix() / fit_kld_em().

Usage

gmm_target_from_posterior(model, ...)

# Default S3 method
gmm_target_from_posterior(model, ...)

# S3 method for class '`function`'
gmm_target_from_posterior(
  model,
  ...,
  parameter_names = NULL,
  log_normalizer = NA_real_,
  name = NULL
)

Arguments

model

One of:

  • a function — a bare callable satisfying the contract above;

  • a fitted model object whose class registers a gmm_target_from_posterior.<class> method in its own package.

...

Forwarded to method-specific implementations.

parameter_names

Character vector of parameter names. Required for the function method (or attached as attr(model, "parameter_names")). The length determines n_dim.

log_normalizer

Numeric scalar log Z of the posterior, if known. NA_real_ (the default) otherwise; downstream diagnostics will label any KLD estimate as shifted.

name

Optional human-readable target name. Defaults to "posterior".

Value

A gmm_target with normalised = FALSE and the user-supplied log_normalizer (or NA_real_).

Details

The contract for the underlying callable is:

  • Vectorised: accepts a numeric matrix with rows indexing independent parameter draws and columns indexing parameters; returns a length-nrow(theta) numeric vector of log p(theta | data) + const.

  • Unnormalised is fine: the marginal likelihood log Z is not required. Where the source package can supply it, pass log_normalizer.

  • Side-effect free: no plotting, no mutable state. Pure function.

  • Domain-safe: returns -Inf outside support rather than raising an error.

The default method errors with a hint pointing the user at either (a) a Bayesian package that registers a method, or (b) the function method below.

Examples

# A trivial unnormalised log-posterior: a 2D banana centred near (1, 0).
log_post <- function(theta) {
  x <- theta[, 1L]
  y <- theta[, 2L]
  -0.5 * (x^2 + (y - 0.1 * x^2 + 1)^2)
}
tgt <- gmm_target_from_posterior(
  log_post,
  parameter_names = c("x", "y")
)
tgt
#> <gmm_target>: "posterior" in p = 2 dimensions
#>   log_density : supplied
#>   samples     : <absent>
#>   normalised  : FALSE