Python scientific tools from scratch on a mac

pyscratch.bash

# This was the order I installed things for a working scientific computing environment on my Mac.

brew install python
pip install virtualenv virtualenvwrapper

# Also add these to your ~/.bash_profile:
export PATH=/usr/local/bin:/usr/bin:$PATH
source /usr/local/bin/virtualenvwrapper.sh

# Not necessary, but quite nice
brew install bash-completion git

# Pre-requisites
brew install gfortran
brew install freetype
brew install zmq
brew install pyqt

# If you're going to do it with virtualenv+virtualenvwrapper, setup your virtualenv
mkvirtualenv sciduck

# Numerical tools installation, in order + some dependencies
pip install numpy f2py pillow python-dateutil pytz six freetype-py pygments
# scipy, pandas, matplotlib *must* be installed after numpy
pip install scipy pandas matplotlib
pip install ipython[notebook]

# You should now be able to run ipython as
# $ ipython qtconsole --pylab=osx
# However, it also works when using just
# $ ipython qtconsole --pylab
# It will then show "Using matplotlib backend: MacOSX" in the IPython window