rv (version 0.949)

postsim: Generate Posterior Simulations for lm or glm Objects

Description

Generate posterior simulations for a given fitted linear or general linear model, assuming the standard "noninformative" priors on the unknowns.

Usage

postsim(fit)
  ## S3 method for class 'lm':
postsim(fit)
  ## S3 method for class 'glm':
postsim(fit)

Arguments

fit

an lm or glm object

Value

• A (named) random vector for each fitted coefficient.

References

Kerman, Jouni and Gelman, Andrew. Manipulating and Summarizing Posterior Simulations Using Random Variable Objects. Technical report, Columbia University, New York.

Examples

x <- 1:20
  y <- rnorm(length(x), mean=x, sd=10)
  print(summary(fit <- lm(y ~ x)))
  bayes.estimates <- postsim(fit)