Start here!

Try this 10-minute tutorial. When it loads, type tutorial and press Enter to start.

def hi():
  print("hello world")

Set up your computer

This is our recommended way to install Python on your system.

Install Anaconda

  • Please download the Anaconda installer. We recommend Python 3.
  • Choose Install for me only
  • By default, Anaconda will prepend itself to your PATH – leave this as is
  • When Anaconda has finished installing, open a terminal (Linux, OSX), or the Anaconda Prompt (Windows)
  • Type conda update conda, hit enter, and then type "y" (and hit enter)
  • Type conda update anaconda, hit enter, and then type "y" (and hit enter)

Run the Jupyter Notebook

With Jupyter, you intersperse code, output, explanatory text, and figures in one big file called a "notebook." Notebooks are a convenient format to explore a language and to share examples of code.

  • To run a notebook, open the Terminal (Linux, OSX) or Anaconda Prompt (Windows) and type jupyter notebook.
  • The notebook will open a new tab in your default browser. Do not close the terminal, as this will also shutdown the notebook.
  • When it has loaded, click on "New" (at the top right) and then "Python3" to create a new notebook.

Python 2 vs Python 3

Python 3 (released in 2008) is the newest version of Python, and most features have not changed. Most packages have been updated to Python 3 by now (2016). So, if your lab does not have a preference, I recommend using Python 3.

There are a couple key differences for novice programmers:

  • In Python 2, you can print with print 42 or print(42). In Python 3, you need to use parentheses, as in print(42).

  • In Python 2, division of two integers like 5/2 will evaluate to 2. (Python will drop the remainder if both numbers are integers.) Python 3 does exact division ('2.5', in this example). If you use Python 2 and do not want this behavior, add this line at the top of each program:

    from __future__ import division.

Text editors and IDEs

For creating large projects in Python, we recommend using a text editor in combination with the Jupyter notebook. Popular choices include:

For even larger projects, a well-engineered IDE (Interactive Development Environment) may be better than a text editor. Typically, IDEs include drag-and-drop support for debugging and refactoring. Popular choices include:

Practice

Past topics of the Python Working Group are linked here. Suggestions? Send us an email at dlab-frontdesk@berkeley.edu.

Software Carpentry

These exercises are really useful to get acquainted with Python. Here's a link to the Jupyter notebook version and the webpage version.

Python for Social Sciences

This is a free online book by Jean Mark Gawron. It's free and online at this link.

Resources

Check the Learning Resources page for more learning materials. Others in the community may also have relevant materials.