summary.rgamma_reg: Summarizing Bayesian gamma_reg Distribution Functions

In glmbayes: Bayesian Generalized Linear Models (IID Samples)

Summarizing Bayesian gamma_reg Distribution Functions

Description

These functions are all methods for class rGamma_reg or summary.rGamma_reg objects.

Usage

## S3 method for class 'rGamma_reg'
summary(object, ...)

## S3 method for class 'summary.rGamma_reg'
print(x, digits = max(3, getOption("digits") - 3), ...)

Arguments

object	an object of class "rGamma_reg" for which a summary is desired.
x	an object of class "summary.rGamma_reg" for which a printed output is desired.
digits	the number of significant digits to use when printing.
...	Additional optional arguments

Value

summary.rGamma_reg() returns an object of class "summary.rGamma_reg", a list containing summaries of posterior draws for the dispersion and precision parameters. Components include:

call	the matched call from the fitted object.
n	number of posterior draws.
coefficients1	matrix of prior means and standard deviations for precision and dispersion.
coefficients	matrix of posterior means, posterior standard deviations, Monte Carlo errors, and empirical tail probabilities.
Percentiles	matrix of posterior percentiles for dispersion draws.
implied_disp_point	dispersion point estimate implied by the Gamma prior on precision, computed as rate / shape.

print.summary.rGamma_reg() prints the summary object and returns x invisibly.

Examples

## summary.rGamma_reg: dGamma prior (dispersion-only; coefficients fixed)
## All five functions (rGamma_reg, rglmb, rlmb, glmb, lmb) use summary.rGamma_reg when
## prior is dGamma.
##
## This example uses the Boston data: Prior_Setup() for hyperparameters and
## ps$coefficients as fixed beta for dGamma / rGamma_reg-style runs.
data("Boston", package = "MASS")

predictors <- setdiff(names(Boston), "medv")
Boston_centered <- Boston
Boston_centered[predictors] <- scale(Boston[predictors], center = TRUE, scale = FALSE)

form <- medv ~
  crim + zn +
  indus + chas + nox + age + dis + rad + tax + ptratio + black + lstat + rm

ps.boston <- Prior_Setup(form, gaussian(), data = Boston_centered)
rate_dg <- if (!is.null(ps.boston$rate_gamma)) ps.boston$rate_gamma else ps.boston$rate

y <- ps.boston$y
x <- as.matrix(ps.boston$x)
wt <- rep(1, length(y))

## 1. rGamma_reg
out1 <- rGamma_reg(
  n = 1000,
  y = y,
  x = x,
  prior_list = list(beta = ps.boston$coefficients, shape = ps.boston$shape, rate = rate_dg),
  offset = rep(0, length(y)),
  weights = wt,
  family = gaussian()
)
summary(out1)

## 2. rglmb
out2 <- rglmb(n = 1000, y = y, x = x,
  pfamily = dGamma(shape = ps.boston$shape, rate = rate_dg, beta = ps.boston$coefficients),
  weights = wt, family = gaussian())
summary(out2)

## 3. rlmb
out3 <- rlmb(n = 1000, y = y, x = x,
  pfamily = dGamma(shape = ps.boston$shape, rate = rate_dg, beta = ps.boston$coefficients),
  weights = wt)
summary(out3)

## 4. glmb
out4 <- glmb(form, data = Boston_centered, family = gaussian(),
  pfamily = dGamma(shape = ps.boston$shape, rate = rate_dg, beta = ps.boston$coefficients))
summary(out4)

## 5. lmb
out5 <- lmb(form, data = Boston_centered,
  pfamily = dGamma(shape = ps.boston$shape, rate = rate_dg, beta = ps.boston$coefficients))
summary(out5)

glmbayes documentation built on Aug. 5, 2026, 1:07 a.m.