Run OpenBUGS on a Mac

I had to use the good old `OpenBUGS` for some analyses that cannot be done in `JAGS`. Below are the steps to install `OpenBUGS` then to run it from your Mac either natively or from `R`. This tutorial is an adaptation of [this post](https://sites.google.com/site/mmeclimate/-bayesmet/openbugs-on-mac-os-x) and [that one](http://www.davideagle.org/r-2/bayesian-modeling-using-winbugs-and-openbugs/running-openbugs-on-mac-using-wine).

1. If not done already, install Homebrew. This program will make the installation of any other programs on your Mac so easy!

2. Install Wine which will allow you to run any Windows programs (.exe) on your Mac. To do so, start by opening Terminal, then type in the command: brew install wine

3. Next, download the Windows version of OpenBUGS here

4. To install OpenBUGS, still in Terminal, go to the directory where the file was downloaded and type (you might need to unzip the file you downloaded first): wine OpenBUGS323setup.exe

5. OpenBUGS is now installed and ready to be used! You can run it by first going to the directory where OpenBUGS was installed. On my laptop, it can be achieved via the command: cd /Applications/OpenBUGS323

6. Then, you just need to tye in the following command in the Terminal, and you should see an OpenBUGS windows poping up: wine OpenBUGS

Now we would like to run OpenBUGS from R.

7. Install the package R2OpenBUGS by typing in the R console:

if(!require(R2OpenBUGS)) install.packages("R2OpenBUGS")

## Loading required package: R2OpenBUGS

8. Now let’s see whether everything works well by running the classical BUGS school example:

Load the OpenBUGS Package

library(R2OpenBUGS)

Load the data

data(schools)

Define the model, write it to a text file and have a look

nummodel <- function(){
for (j in 1:J){
  y[j] ~ dnorm (theta[j], tau.y[j])
  theta[j] ~ dnorm (mu.theta, tau.theta)
  tau.y[j] <- pow(sigma.y[j], -2)}
mu.theta ~ dnorm (0.0, 1.0E-6)
tau.theta <- pow(sigma.theta, -2)
sigma.theta ~ dunif (0, 1000)
}
write.model(nummodel, "nummodel.txt")
model.file1 = paste(getwd(),"nummodel.txt", sep="/")
file.show("nummodel.txt")

Prepare the data for input into OpenBUGS

J <- nrow(schools)
y <- schools$estimate
sigma.y <- schools$sd
data <- list ("J", "y", "sigma.y")

Initialization of variables

inits <- function(){
  list(theta = rnorm(J, 0, 100), mu.theta = rnorm(1, 0, 100), sigma.theta = runif(1, 0, 100))}

Set the Wine working directory and the directory to OpenBUGS, and change the OpenBUGS.exe location as necessary:

WINE="/usr/local/Cellar/wine/2.0.4/bin/wine"
WINEPATH="/usr/local/Cellar/wine/2.0.4/bin/winepath"
OpenBUGS.pgm="/Applications/OpenBUGS323/OpenBUGS.exe"

The are the parameters to save

parameters = c("theta", "mu.theta", "sigma.theta")

Run the model

schools.sim <- bugs(data, inits, model.file = model.file1,parameters=parameters,n.chains = 3, n.iter = 1000, OpenBUGS.pgm=OpenBUGS.pgm, WINE=WINE, WINEPATH=WINEPATH,useWINE=T)

R will pause. You might get a weird message starting by err:ole, just ignore it. When the run is complete, a prompt will reappear, then just type the following command to get the result:

print(schools.sim)

## Inference for Bugs model at "/Users/oliviergimenez/Desktop/nummodel.txt", 
## Current: 3 chains, each with 1000 iterations (first 500 discarded)
## Cumulative: n.sims = 1500 iterations saved
##             mean  sd 2.5%  25%  50%  75% 97.5% Rhat n.eff
## theta[1]    12.2 7.9 -1.3  7.5 11.2 16.4  32.1  1.0    62
## theta[2]     9.1 6.5 -4.0  5.1  9.4 13.2  21.4  1.0   150
## theta[3]     7.8 7.7 -9.4  3.6  8.5 12.6  21.1  1.0   360
## theta[4]     8.8 6.6 -4.5  4.5  9.2 13.3  20.4  1.0   110
## theta[5]     6.8 6.9 -8.2  2.3  7.5 11.4  17.7  1.0   410
## theta[6]     7.3 7.2 -8.6  2.7  8.2 11.8  18.9  1.0   190
## theta[7]    11.5 6.4 -0.3  7.5 11.2 15.7  25.0  1.1    42
## theta[8]     9.7 7.6 -4.7  5.1  9.6 14.4  25.1  1.0   130
## mu.theta     9.2 5.2 -1.2  5.8  9.3 12.5  18.2  1.0    88
## sigma.theta  5.9 5.6  0.2  1.7  4.4  8.5  20.2  1.1    51
## deviance    60.7 2.2 57.2 59.2 60.1 61.9  65.6  1.0   120
## 
## For each parameter, n.eff is a crude measure of effective sample size,
## and Rhat is the potential scale reduction factor (at convergence, Rhat=1).
## 
## DIC info (using the rule, pD = Dbar-Dhat)
## pD = 2.8 and DIC = 63.4
## DIC is an estimate of expected predictive error (lower deviance is better).

When run natively, WinBUGS and OpenBUGS have nice debugging capabilities; also, you can see what is going on, I mean the program reading the data, generating inits, and so on. To get the OpenBUGS window with a bunch of useful info, just add debug=T to the call of the bugs function, and re-run the model

schools.sim <- bugs(data, inits, model.file = model.file1,parameters=parameters,n.chains = 3, n.iter = 1000, OpenBUGS.pgm=OpenBUGS.pgm, WINE=WINE, WINEPATH=WINEPATH,useWINE=T,debug=T)

## arguments 'show.output.on.console', 'minimized' and 'invisible' are for Windows only

You will have to close the OpenBUGS window to get the prompt back.