---
title: "Installation"
output: rmarkdown::html_vignette
vignette: >
%\VignetteIndexEntry{Installation}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
---
```{r, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>"
)
```
`bml` fits Bayesian multiple-membership multilevel models via **JAGS**, so installation has two parts:
1. Install JAGS (system software)
2. Install the `bml` R package (from CRAN or GitHub)
## 1. Install JAGS
Install the latest JAGS version for your operating system from:
-
After installing JAGS, restart R so `rjags` can find it.
## 2. Install `bml` R package
### Option A: Install from CRAN (Recommended)
The stable release version is available on CRAN:
```{r, eval = FALSE}
install.packages("bml")
```
### Option B: Install development version from GitHub
For the latest development features, install from GitHub using `{remotes}`:
```{r, eval = FALSE}
install.packages("remotes")
remotes::install_github("benrosche/bml")
```
If you want to build vignettes locally during installation:
```{r, eval = FALSE}
remotes::install_github("benrosche/bml", build_vignettes = TRUE)
```
## Troubleshooting
- `Error: (converted from warning) ...` while installing. `{remotes}` treats an installation warning as an error on your machine. You can opt out:
```{r, eval = FALSE}
Sys.setenv(R_REMOTES_NO_ERRORS_FROM_WARNINGS = "true")
remotes::install_github("benrosche/bml")
```
- JAGS not found / `rjags` fails to load: Make sure JAGS is installed (not just the R packages), then restart R. On Windows, confirm you installed the 64-bit JAGS build if you are using 64-bit R.