The R package shinybrms provides a graphical user interface (GUI) for fitting Bayesian regression models using the R package brms which in turn relies on Stan. The shinybrms GUI is a shiny app.
To get an impression of the shinybrms app, have a look at this page. The following text explains how to launch the shinybrms app (and also how to install it, if necessary).
The following two sections describe two ways for launching the shinybrms app, either with or without the installation of shinybrms. The former is recommended as it offers all advantages that R packages have (e.g., offline usage). For both ways, you need to perform the following steps first:
- Install R (see the R homepage).
- Install the R package rstan (see
the “RStan Getting Started” GitHub
page
for instructions; make sure to use
install.packages("rstan", [...], dependencies = TRUE)
with[...]
as advised on the “RStan Getting Started” GitHub page). - If you want to be able to use the cmdstanr backend (or if you need it because the rstan backend doesn’t work as expected), then you need to install cmdstanr as well as CmdStan by following the instructions on the cmdstanr homepage. In general, the rstan backend should be sufficient, though. In the context of shinybrms, the major advantage of the cmdstanr backend is a (generally) faster Stan run.
-
Use one of the following approaches to install the R package shinybrms either from CRAN or from GitHub. The GitHub version might be more recent than the CRAN version, but the CRAN version might be more stable. You also need to decide whether you want to use the example datasets from the R packages lme4, MASS, and rstanarm or not.
- If you want to use the example datasets from the R packages
lme4, MASS, and rstanarm, then the R code for
installing shinybrms from CRAN and GitHub (respectively) is
as follows:
-
To install shinybrms from CRAN:
install.packages("shinybrms", dependencies = TRUE)
-
To install shinybrms from GitHub:
if (!requireNamespace("devtools", quietly = TRUE)) { install.packages("devtools") } devtools::install_github("fweber144/shinybrms", dependencies = TRUE)
-
- If you don’t want to use the example datasets from the R
packages lme4, MASS, and rstanarm, then the R code
for installing shinybrms from CRAN and GitHub (respectively)
is as follows:
-
To install shinybrms from CRAN:
install.packages("shinybrms")
-
To install shinybrms from GitHub:
if (!requireNamespace("devtools", quietly = TRUE)) { install.packages("devtools") } devtools::install_github("fweber144/shinybrms")
-
- If you want to use the example datasets from the R packages
lme4, MASS, and rstanarm, then the R code for
installing shinybrms from CRAN and GitHub (respectively) is
as follows:
-
Launch the shinybrms app by either running the following R code:
library(shinybrms) launch_shinybrms()
or this R code which ensures that the app opens up in the default web browser (helpful, e.g., if you are using RStudio):
library(shinybrms) launch_shinybrms(launch.browser = TRUE)
-
Install the R package brms. You may use the following R code for this:
install.packages("brms")
-
If you want to use the example datasets from the R packages lme4, MASS, and rstanarm, you need to install these packages. You may use the following R code for this:
install.packages(c("lme4", "MASS", "rstanarm"))
-
Launch the shinybrms app directly from GitHub by either running the following R code:
shiny::runGitHub("fweber144/shinybrms", subdir = "inst/shinybrms_app")
or this R code which ensures that the app opens up in the default web browser (helpful, e.g., if you are using RStudio):
shiny::runGitHub("fweber144/shinybrms", subdir = "inst/shinybrms_app", launch.browser = TRUE)