Description Usage Arguments Details Value Examples
Evaluates marginal likelihood of Stan model at the posterior mean using iterative kernel density estimation
1 2 3 |
ikde.model |
An object of class ikde.model, does not necessarily have to be built |
burn.iter |
Number of warmup iterations |
sample.iter |
Number of sampling iterations |
control |
Control parameters used in the Markov chain. See ?rstan::stan for details. |
refresh |
How frequently should progress be reported, in numbers of iterations |
display.output |
Boolean indicating whether output from rstan::stan should be printed |
show.trace |
Boolean indicating whether to show trace plots |
Uses evaluate.likelihood, evaluate.priors, and evaluate.posterior to form an estimate of marginal likelihood at the posterior mean.
A real number indicating value of the log-marginal-likelihood at the posterior mean
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 | data(lm.generated)
X <- lm.generated$X
y <- lm.generated$y
data <- list(N = list(type = "int<lower=1>", dim = 1, value = nrow(X)),
k = list(type = "int<lower=1>", dim = 1, value = ncol(X)),
X = list(type = "matrix", dim = "[N, k]", value = X),
y = list(type = "vector", dim = "[N]", value = y))
parameters <- list(beta = list(type = "vector", dim = "[k]"),
sigma_sq = list(type = "real<lower=0>", dim = 1))
model <- list(priors = c("beta ~ normal(0, 10);",
"sigma_sq ~ inv_gamma(1, 1);"),
likelihood = c("y ~ normal(X * beta, sqrt(sigma_sq));"))
ikde.model <- define.model(data, parameters, model)
# Only an estimation, may not exactly match presented result
ikde(ikde.model)
# [1] -388.9264
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.