From Prof. Oates's 2016 data science class

This document describes how to install Anaconda, which has everything
you'll need for the first part of the class.  The instructions are
written for linux (I'm doing this on a VM running Ubuntu on my Mac
Book), but should work with minor modifications for Windows, etc.

In fact, I encourage users of other operating systems to start
discussions in Slack.

(1) Google "download anaconda" and click on the first result

(2) Scroll down a bit and you should see a set of tabs to download for
Windows, OSX, and Linux.  Click the correct tab for your OS if it is
not already foregrounded.

(3) Choose 32 or 64 bits based on your machine architecture (64 bits
should work and is preferred) and Python 3.5 or 2.7.  Note that
Anaconda installs Python locally so choose the version that you're
most comfortable with, or 3.5 if you don't care.  Click the link based
on these choices.

(4) That downloads a large file that beings with "Anaconda" and
contains indications of the choices you made above.  In my case the
file was named Anaconda3-4.1.1-Linux-x86_64.sh

(5) Make the install script executable:
 
      chmod +x Anaconda3-4.1.1-Linux-x86_64.sh

(6) Run it, and answer a few simple questions:

      ./Anaconda3-4.1.1-Linux-x86_64.sh

(7) After installing, Anaconda will ask if you want to prepend the
anaconda bin directory to your path.  Do that.  If you don't, you are
in danger of not using the correct python and being confused when you
try to load packages.  Be sure to quit all shells and start new ones
so that the new path takes effect.

(8) Test the install.  At a shell:

      python
      import pandas

    If that returns without an error, you're good to go!