bhlm.SDplots: Savage-Dickey plots for outcomes

In arcuo/bachelors_meta: Bayesian Hierarchical Latent Mixture model meta analysis and plots.

Description

Plot outcomes posterior and prior distributions with object from bhlm.

Usage

bhlm.SDplots(bhlm_object, null_hypothesis, outcome_priors_data = NULL,
  outcome_options = NULL, return_plots = FALSE,
  density_estimation = "logspline", cum_prob = NULL, iter = 10000)

Arguments

bhlm_object	Object returned from bhlm, of class bhlm_object.
null_hypothesis	int, point at which to check the Bayes Factor.
outcome_priors_data	data.frame with variables for each outcome (Important: variable names need to be the same as chosen outcome names.). Sample from the prior distribution with ex. rnorm(10000, 0, 1). If not set, automatically samples from the prior distributions (bhlm_object@outcome_priors_c or bhlm_object@outcome_priors_m). Automatic sampling is not yet implemented for priors defined with data vectors.
outcome_options	Choose which outcomes should be plotted. Defaults to bhlm_object@outcome_options.
return_plots	Return ggplot objects in list.
density_estimation	Log estimate the posterior and prior distributions for the plot. Non-log estimated plots are not yet fully implemented.
cum_prob	Print cummulated probability (from log estimated posterior distribution) at chosen point along x (Not yet implemented).
iter	Number of iterations for sampling from prior distribution in automatic sampling.

Value

list of outcomes with each element containing a plot and log estimation of the posterior distribution. T he simulation data is also listed as "data".

Get plot: list$outcome$plot

Get Log estimate: list$outcome$log_est

Author(s)

Hugh Benjamin Zachariae

arcuo/bachelors_meta documentation built on May 5, 2019, 2:41 a.m.