b-rich/irkernel

R kernel for IPython

Native R kernel for IPython

This is still experimental and unreliable. Your code should be safe, since IPython handles saving and loading notebooks in another process, but you'll lose all your variables if it crashes.

##Installing

You'll need zmq development headers to compile rzmq, as well curl headers for R devtools. On a recent Ubuntu/Debian installation, you can get these with:

sudo apt-get install libzmq3-dev libcurl4-openssl-dev

or with homebrew:

brew install zmq
# or upgrade
brew update
brew upgrade zmq

If using MacPorts on OS X, make sure an X server is intalled, open a terminal and do the following:

• sudo port install zmq

• Direct the compiler to use MacPorts libraries using

   export CPATH=/opt/local/include
   export LIBRARY_PATH=/opt/local/lib
  

• Start R in the same terminal, and proceed as below:

We need development versions of several packages from Github for now, due to recent fixes. First, you need to make sure you have the devtools R package available. If you don't, at the R console type:

install.packages("devtools")

Then, you can install the necessary development dependencies with:

library(devtools)
install_github('armstrtw/rzmq', pull=8, ref=NULL)
install_github("takluyver/IRdisplay")
install_github("takluyver/IRkernel")

You'll also need IPython. If you already have a Python environment set up, install IPython using your preferred tools. If not, installing Anaconda is the quickest way to get everything you need.

Running the notebook

ipython notebook --KernelManager.kernel_cmd="['R', '-e', 'IRkernel::main(\'{connection_file}\')']"

You can also substitute 'qtconsole' or 'console' for 'notebook' in this command.