A virtualized Python setup on OS X (Mountain) Lion

Written by Arvinder Singh / July 06, 2012 / 4 mins read / Filed under Python, / Osx, / Numpy, / Scipy, / Matlibplot, / Virtualenv

Update 2012-07-26: Tested the install successfully for Mountain Lion.

Yesterday I replaced my slow old hard drive with a shiny new SSD drive. In the process I decided to reinstall Python on my machine and document the process for future reference.

I have a strong preference for command line installer such as homebrew, versioning system git and isolating installations into virtual environments. Come follow along.


Start by installing Xcode. You can download it for free from Apple app store.

Then install Command line tools for Xcode from https://developer.apple.com/downloads/


From the website

Git is a free and open source distributed version control system designed to handle everything from small to very large projects with speed and efficiency.

Git is easy to learn and has a tiny footprint with lightning fast performance. It outclasses SCM tools like Subversion, CVS, Perforce, and ClearCase with features like cheap local branching, convenient staging areas, and multiple workflows.

  • Download and run installer from http://git-scm.com/.
  • Restart bash/zsh sessions
  • Run provided shell script to add PATH and MANPATH to your ~/.MacOSX/environment.plist file


Homebrew allows you to install Unix tools on OS X. A much better replacement of fink.

/usr/bin/ruby -e "$(/usr/bin/curl -fsSL https://raw.github.com/mxcl/homebrew/master/Library/Contributions/install_homebrew.rb)"
	brew doctor

If your system has no issues, brew would return

Your system is raring to brew.


Lion already comes installed with version 2.7.1. However we will build virtual environment to isolate and install more than one versions of python as well as corresponding modules.


readline: The GNU Readline library provides a set of functions for use by applications that allow users to edit command lines as they are typed in.

sudo easy_install readline

gdbm: GNU dbm is a library of database functions that use extensible hashing and work similar to the standard UNIX dbm. These routines are provided to a programmer needing to create and manipulate a hashed database.

pkg-config : pkg-configis a helper tool used when compiling applications and libraries and helps you insert the correct compiler options on the command linerather than hard-coding values on where to find libraries. It is language-agnostic.

sqlite: SQLite is a software library that implements a self-contained, serverless, zero-configuration, transactional SQL database engine.

brew install readline sqlite gdbm pkg-config

pip: A better package manager for Python

sudo easy_install pip


Virtualenv lets us install isolated Python environment, each with its own set of packages and dependencies. It is similar to rvm in Ruby.

Lets install virtualenv and virtualenvwrapper (a tool to easily manage virtualenv enviroments).

sudo pip install virtualenv
sudo pip install virtualenvwrapper

To load virtualenv, we have to source /usr/local/bin/virtualenvwrapper.sh script. We can automate by adding to ~/.bashrc or in my case ‘~/.zshrc’ depending on the shell you use.

To use virtualenv refer http://www.doughellmann.com/docs/virtualenvwrapper/command_ref.html#command

Python tools

mkvirtualenv ML
workon ML

As numpy and scipy will fail without fortran compiler

brew install gfortran

Install NumPy

pip install numpy

Install SciPy

pip install scipy

Although OSX comes with the prerequisites for matplotlib installed, setup.py is not able to locate them. Therefore we’ll use

export LDFLAGS="-L/usr/X11/lib"
export CFLAGS="-I/usr/X11/include -I/usr/X11/include/freetype2 -I/usr/X11/include/libpng12"

Now install matplotlib

pip install matplotlib


R is a free software environment for statistical computing and graphics.

brew install R

Link R

ln -s "/usr/local/Cellar/r/2.15.1/R.framework" /Library/Frameworks

Enable rJava support

R CMD javareconf JAVA_CPPFLAGS=-I/System/Library/Frameworks/JavaVM.framework/Headers

Last we’ll install iPython. If you like the default python shell I recommend a good ~/.pythonrc . Here is a good starting point by Seth House.

pip install ipython

Now Python shell is ready.


After launching terminal, to move into virtualenv - ML in our case

workon ML

Listing installed packages installed in the virtualenv (pip packages stored here for env ML ~/.virtualenvs/ML/)

pip freeze

Install/Upgrade/Uninstall packages

pip install packageName
pip install -U packageName
pip uninstall packageName

Leave virtualenv

deactivate ML