estimatesmethods: 'estimates' and 'estimate' objects and methods

In rferrali/rogali: A series of functions for the corrupt_audit project

Description

estimates and estimate objects and methods

Usage

## S3 method for class 'estimate'
as.matrix(x, burn = 0, thin = 1,
  what = c("params", "probs"))

## S3 method for class 'estimates'
as.matrix(x, burn = 0, thin = 1,
  what = c("params", "probs"))

as.mcmc.estimate(x, ..., burn = 0, thin = 1, what = c("params",
  "probs"))

as.mcmc.list.estimates(x, ..., burn = 0, thin = 1, what = c("params",
  "probs"))

as_tibble.estimate(x, ..., burn = 0, thin = 1, what = c("params",
  "probs"), details = T)

as_tibble.estimates(x, ..., burn = 0, thin = 1, what = c("params",
  "probs"), details = T)

Arguments

burn	number or percent of iterations to discard
thin	thinning interval. Keep every thin-th iteration.
what	returns either parameter values or posterior probabilities Pr(Z = 1)
...	use to select variables a la select
details	a logical indicating whether the returned tibble should contain additional columns such as the iteration number and and id variable for the chain.

Details

An estimates object is merely a list of estimate objects. Each estimate object contains parameter samples for a given chain. They have a fairly complicated structure that I describe in more details below. Those objects should never be used directly, but rather be converted to more familiar data types. In particular, the as.mcmc and as.mcmc.list methods should be used to perform additional diagnostics using the coda package.

Value

An estimate object contains the following fields:

beta

a kBeta x 2 x nSamples array with draws of beta coefficients

gamma

a kGamma x nSamples matrix with draws of gamma coefficients

pZ

a nMun x nSamples matrix with posterior Pr(Z = 1)

rferrali/rogali documentation built on May 26, 2019, 7 p.m.