The mcglm package fits multivariate covariance generalized linear
models (Bonat and Jorgensen, 2016).
mcglm is an R package designed to fit Multivariate Covariance
Generalized Linear Models. It allows you to specify a distinct linear
predictor for each response variable, offering exceptional flexibility
for analyses involving multiple outcomes.
With mcglm, you can model a wide range of response types — continuous,
discrete (such as counts and binary), limited, and even zero inflated
responses, whether continuous or mixed.
Its main strength lies in the ability to capture complex relationships between variables through multiple covariance structures, enabling more realistic and robust multivariate modeling.
This package was developed as part of the Wagner’s Ph.D. thesis, combining academic rigor with practical value for the statistical modeling community.
Use the devtools package (available from
CRAN) to
install automatically from this GitHub repository:
library(devtools)
install_github("bonatwagner/mcglm")
- Wagner Hugo Bonat (author and main developer)
This R package is develop using
roxygen2 for documentation and
devtools to check and build.
Also, we adopt the Gitflow
worflow in
this repository.
Please, see the instructions for contributing to collaborate.