cause_grid_adapt: CAUSE posterior density estimation

In jean997/cause: CAUSE: Causal Analysis Using Summary Effect Estimates

Description

Uses an adaptive grid approximation to estimate the posterior distribution of q, gamma, and eta or a subset of those

Usage

cause_grid_adapt(
  X,
  param_ests,
  params = c("q", "gamma", "eta"),
  priors = list(function(q) {     dbeta(q, 1, 10) }, function(b) {     dnorm(b, 0, 0.6)
    }, function(b) {     dnorm(b, 0, 0.6) }),
  n_start = c(10, 20, 20),
  max_post_per_bin = 0.001
)

Arguments

X	An object of class cause_data
param_ests	param_ests Object of class cause_params
params	Vector of parameter names. Should be some subest of c("q", "gamma", "eta") given in any order.
priors	List of prior functions corresponding to parameters in params. Each element should be a funciton that takes one argument and returns a log likelihood.
n_start	Vector of length length(params) giving the number of starting bins for each dimension
max_post_per_bin	Maximum posterior probability for each grid cube. This parameter controls the fineness of the approximation.

Value

An object of class cause_post.

jean997/cause documentation built on Dec. 25, 2021, 10 p.m.