Description Usage Arguments Details Value Examples
Evaluates posterior of Stan model at specified evaluation point
1 2 3 | evaluate.posterior(ikde.model, eval.point, burn.iter = 1000,
sample.iter = 1000, control = NULL, refresh = NULL,
display.output = FALSE, show.trace = FALSE)
|
ikde.model |
An object of class ikde.model, does not necessarily have to be built |
eval.point |
A list of parameter names and the point to evaluate the posterior |
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 list of ikde.model objects created by create.restricted.models to estimate posterior density. Each ikde.model is fit, then conditional posterior density is estimated at the specified point.
A real number indicating value of the log-posterior at the specified evaluation point
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | 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)
ikde.model <- build.model(ikde.model)
stan.fit <- fit.model(ikde.model)
stan.extract <- rstan::extract(stan.fit)
eval.point <- list(beta = apply(stan.extract$beta, 2, mean),
sigma_sq = mean(stan.extract$sigma_sq))
# Only an estimation, may not exactly match presented result
evaluate.posterior(ikde.model, eval.point)
# [1] -1.889711
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.