confint.blm: Confidence Interval for Coefficients of a Bayesian Linear...

In manschmi/blm: Bayesian Linear Regression

Description

Compute the confidence interval confint for a bayesian linear model.

Usage

## S3 method for class 'blm'
confint(object, parm, level = 0.95, ...)

Arguments

object	a blm object.
parm	parameter for which confidence interval should be computed. Either a number or vector of numbers. If missig all parameters are considered.
level	confidence level
...	other arguments (currently ignored).

Details

Deviance is the sum of squares of the residuals, extracted from the maxium a posterior estimate of the blm object.

Value

Pseudo confidence interval of parameter(s) of the blm object. Returns the quantile specified by level for the distribution of param.

Examples

  x <- rnorm(10)
  b <- 1.3
  w0 <- 0.2 ; w1 <- 3
  y <- rnorm(10, mean = w0 + w1 * x, sd = sqrt(1/b))
  model <- blm(y ~ x, prior = NULL, beta = b, data = data.frame(x=x, y=y))
  
  confint(model)
  confint(model, 1)
  confint(model, c(1,2))

manschmi/blm documentation built on May 21, 2019, 11:25 a.m.