You will need to have the following installed:
- Python - R - IPython, and notebook
Download and install R here:
http://cran.r-project.org/bin
To obtain Python + IPython + notebook app (if you don't have these already), I recommend the Anaconda Python distribution package here:
http://continuum.io/downloads
Now, you will also need to install the rpy2 resource package.
Mac OS/Linux specific instructions to get rpy2:
Type the following:
pip install rpy2
If you don't have pip installed (try typing pip at the terminal to test if you have it installed), get it! It's a very useful command that can be used to install various Python packages
Windows specific instructions to get rpy2:
Go here:
http://www.lfd.uci.edu/~gohlke/pythonlibs/#rpy2
and download and install the appropriate binary executable for your system (i.e. 32-bit sys: rpy2‑2.x.x.win32‑py2.7.exe; 64-bit sys: rpy2‑2.x.x.win‑amd64‑py2.7.exe)
All Users:
After that, download the following zip here to a directory of your choice:
https://github.com/donhmai/rlangipythonsetup/archive/master.zip
Unzip the files.
Mac OS/Linux specific instructions to set up rkernel.py:
cd to the directory you unzipped the files via terminal.
Type:
bash setup_rkernel_maclinux.sh
**Windows specific instructions to set up rkernel.py:**
Double-click on setup_rkernel_win.bat. It will prompt you for administrator access privileges. Click yes: these privileges are required to set the R_HOME and R_USER system environment variables that rpy2 needs in order to run properly. After that, follow the instructions as prompted by the command-line that opens up.
**All Users:**
If you don't see rkernel.py in the current directory anymore, that means that the script ran successfully!
To run the R Interpreter in IPython's notebook app, just type:
```shell ipython notebook --profile rkernel ```
When you create a new notebook, writing a series of R commands, such as:
X <- c(0,1,2,3,4)
Y <- c(3,5,4,6,7)
XYlm = lm(Y~X)
XYcoef = coef(XYlm)
print(summary(XYlm))
par(mfrow=c(2,2))
plot(XYlm)
should result in the commands getting executed as they would in a standard R interpreter. Congrats!
Note: to run the regular Python interpreter (and not R) in notebook instead, just run:
```shell ipython notebook ```
This guide was based off Fernando Perez's "hack" to convert a Python interpreter session into a R interpreter session by prepending "%%R" prior to ea. command getting run