post_dim: Obtain MCMC dimensions from an mcmc.list

In postpack: Utilities for Processing Posterior Samples Stored in 'mcmc.lists'

Obtain MCMC dimensions from an mcmc.list

Description

Quickly query the number of burn-in samples, post-burnin, thinning, number of chains, etc. from a mcmc.list object.

Usage

post_dim(post, types = NULL)

Arguments

post	A mcmc.list object.
types	The dimension types to return. Must contain some of "burn", "post_burn", "thin", "chains", "nodes". Defaults to NULL, in which case all of these are returned.

Value

A numeric vector with named elements, which may contain:

• burn: The burn-in period + adapting phase (per chain).

• post_burn: The post-burn-in period (per chain).

• thin: The thinning interval post-burn-in.

• chains: The number of chains.

• saved: The number of saved samples across all chains.

• params: The number of nodes with MCMC samples.

All of these will be returned if types = NULL, a subset can be returned by specifying (for example) types = c("burn", "thin").

Note

If the post object was thinned after MCMC completed using post_thin(), then the "burn" and "thin" dimensions will be improperly calculated. post_thin() performs post-MCMC thinning of mcmc.list objects, and is solely for developing long-running post-processing code, so this is okay.

Examples

# load example mcmc.list
data(cjs)

# get all relevant dimensions
post_dim(cjs)

# get only the number of chains
post_dim(cjs, "chains")

# get the thinning and burn-in intervals
post_dim(cjs, c("burn", "thin"))

postpack documentation built on Dec. 28, 2022, 1:23 a.m.