glmbayes-package: glmbayes: Bayesian Generalized Linear Models with iid...
In glmbayes: Bayesian Generalized Linear Models (IID Samples)

glmbayes-package

R Documentation

glmbayes: Bayesian Generalized Linear Models with iid Sampling

Description

glmbayes provides independent and identically distributed (iid) samples for Bayesian generalized linear models (GLMs), serving as a Bayesian analogue to the base glm() function. Supported likelihood families include Gaussian, Poisson, Binomial, and Gamma models with log-concave likelihoods.

Details

The main user-facing interface is glmb(), which mirrors the structure of glm() and supports prior specification through pfamily objects. Lower-level functions such as rglmb() and rGamma_reg() provide direct access to the underlying samplers and can be used in block Gibbs sampling or hierarchical model implementations.

For an introduction to the package, examples, and a complete set of vignettes, see:

README: https://github.com/knygren/glmbayes#readme
All vignettes: browseVignettes("glmbayes")

The package includes extensive documentation on model fitting, prior construction, diagnostics, and optional GPU acceleration using OpenCL.

Releases: Current version 0.9.7 on CRAN (install.packages("glmbayes")). Source is available from GitHub; R-Universe (https://knygren.r-universe.dev/glmbayes) also builds binaries from that source. Prebuilt CRAN and R-Universe binaries do not include OpenCL; GPU support requires a source install once the host OpenCL environment is ready (see vignette("Chapter-16", "glmbayes") for the three-step process).

IID posterior simulation for non-Gaussian GLMs and several non-conjugate linear-model setups uses the likelihood-subgradient envelope method of \insertCiteNygren2006glmbayes. Introductory material and worked examples are in \insertCiteglmbayesChapter00,glmbayesChapterA01glmbayes; estimation and simulation background in \insertCiteglmbayesChapterA02,glmbayesSimmethods,glmbayesChapterA08glmbayes; prior derivations for Prior_Setup() in \insertCiteglmbayesChapterA12glmbayes; GPU/OpenCL topics in \insertCiteglmbayesChapter12,glmbayesChapterA10glmbayes.

OpenCL startup checks

In interactive sessions, attaching the package with library(glmbayes) may emit a short packageStartupMessage when has_opencl() is FALSE (typical for CRAN binaries) but a GPU or OpenCL stack appears available on the host. OpenCL modelling paths require a source install of glmbayes with OpenCL at compile time; has_opencl() then reports whether that build succeeded. The note confirms full CPU use and points to vignette("Chapter-16"). Machines without a detectable GPU stack stay silent. Set options(glmbayes.quiet_opencl_startup = TRUE) to suppress attach notes (recommended for CI and R CMD check).

Author(s)

Kjell Nygren

References

\insertAllCited

Examples

set.seed(333)
## Dobson (1990) Page 93: Randomized Controlled Trial :
counts <- c(18,17,15,20,10,20,25,13,12)
outcome <- gl(3,1,9)
treatment <- gl(3,3)
print(d.AD <- data.frame(treatment, outcome, counts))

## Call to glm
glm.D93 <- glm(counts ~ outcome + treatment, 
               family = poisson())

## Using glmb
## Step 1: Set up Prior
ps=Prior_Setup(counts ~ outcome + treatment,family = poisson())
mu=ps$mu
V=ps$Sigma

# Step 2: Call the glmb function
glmb.D93<-glmb(counts ~ outcome + treatment, family=poisson(), 
               pfamily=dNormal(mu=mu,Sigma=V))

summary(glmb.D93)

## ----Printed_Views------------------------------------------------------------
## Printed view of the output from the glm function 
print(glm.D93)
## Printed view of the output from the glmb function 
print(glmb.D93)

## ----Methods---------------------------------------------------------------
## Methods for class "lm"
methods(class="lm")

## Methods for class "glm"
methods(class="glm")

## Methods for class "glmb"
methods(class="glmb")

## ----summary--------------------------------------------------------------
## summary output for the "glm" class
summary(glm.D93)

## summary output for the "glm" class
summary(glmb.D93)

## ----fitted outputs-------------------------------------------------------
## fitted outputs for the glm function
fitted(glm.D93)

## ----glmb fitted outputs------------------------------------------------------
## mean of fitted outputs for the glm function
colMeans(fitted(glmb.D93))

## ----predictions----------------------------------------------------------
## predictions for the glm function
predict(glm.D93)

## predictions for the glmb function
colMeans(predict(glmb.D93)) 

## ----residuals------------------------------------------------------------
## residuals for the glm function
residuals(glm.D93)

## residuals for the glmb function
colMeans(residuals(glmb.D93))

## ----vcov-----------------------------------------------------------------
## vcov for the glm function
vcov(glm.D93)

## vcov for the glm function
vcov(glmb.D93)

## ----confint--------------------------------------------------------------
## confint for the glm function
confint(glm.D93)

## confint for the glm function
confint(glmb.D93)

## ----AIC/DIC------------------------------------------------------------------
## AIC for the glm function (equivalent degrees of freedom and the AIC)
extractAIC(glm.D93)

## DIC for the glmb function (estimated effective number of parameters and the DIC)
extractAIC(glmb.D93)

## ----Deviance-------------------------------------------------------------
## Deviance for the glm function
deviance(glm.D93)

## Deviance for the glmb function
mean(deviance(glmb.D93))

## ----logLik---------------------------------------------------------------
## Deviance for the glm function
logLik(glm.D93)

## Deviance for the glmb function
mean(logLik(glmb.D93))

## ----Model Frame----------------------------------------------------------
## Model Frame for the glm function
model.frame(glm.D93)

## Model Frame for the glmb function
model.frame(glmb.D93$glm)

## ----formula--------------------------------------------------------------
## formula for the glm function
formula(glm.D93)

## ----formula-------------------------------------------------------------
## formula for the glmb function
formula(glmb.D93)

## ----family--------------------------------------------------------------
## family for the glm function
family(glm.D93)

## family for the glmb function
family(glmb.D93$glm)

## ----nobs-----------------------------------------------------------------
## nobs for the glm function
nobs(glm.D93)

## nobs for the glmb function
nobs(glmb.D93)

## ----show-----------------------------------------------------------------
## show for the glm function
show(glm.D93)

## show for the glmb function
show(glmb.D93)

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