layout: default title: IPython Notebook Install Guide active_tab: resources —
IPython Notebooks are a platform for writing and running Python code to view output and produce visualizations instantaneously. It provides a number of feedback tools that will help you quickly create the best version of your code.
We’ll be using Python for the vast majority code we write in this course. If you’ve never written in Python, don’t worry - we’ll get you up to speed with the Python Bootcamp assignment. It’s a high-level language with a great variety of applications, extremely readable, and will let us use IPython Notebooks.
Python is likely already set up on your computer, but in order to confirm that, open up a command line (Terminal on OS X, Command Prompt on Windows, CLI on Linux) and run after the prompt ($):
python --version
You should receive some output with the version, like below.
Python 2.7.10
Either 2.7 or 3.4+ is fine – different versions have different library support, but both will work for our purposes. If you don’t have Python, you need to download it. A guide to do so for a number of common operating systems is here.
Package managers are great tools to set up software and its dependencies on your computer. You’ll use them a few times throughout the course, and familiarizing yourself with what they offer will give you a broader range of tools to use in your final project.
The exciting news is that the package manager we’ll use to start, called pip, comes bundled with the Python installation that you just confirmed or completed. As long as the version that you saw when you typed python was above 2.7.9 or 3.4, you are guaranteed to have pip. To double-check, you can type pip --version into your command line. If it’s installed, you’ll see your version and possibly where you got it. If pip isn’t installed, you’ll see an error message which tells you some variation of ‘command not found.’ If you see an error, you should first try updating your version of Python by downloading the most recent copy, and if that doesn’t resolve it, download pip directly.
You can use pip to install a very useful Python tool called virtualenv. It is a tool to keep the dependencies required by different projects in separate places, by creating virtual Python environments for them. (It also solves a problem created by the latest version of Mac OS X “El Capitan”, which doesn’t let users install anything in the Python directory, even using the sudo command). To install virtual environments run this command once:
sudo pip install virtualenv
To create a new virtual environment, give the environment a name like nets213 and run the command (you only need to do this once):
cd ~/
virtualenv nets213
Note that it will install it in a specific directory, so whenever you want to activate it, be sure to go back to that directory. To activate your virtual environment run this command:
cd ~/
source nets213/bin/activate
You will then see that the prompt on your Terminal window is prefixed with (nets213). That shows you that you’re in the nets213 virtual environment.
You’re in the home stretch of IPython Notebook set-up! Now you can return to your command line and run (nets213) $ pip install ipython[notebook]. That will install ipython notebook and its dependencies in your nets213 virtual environment.
Type ipython notebook to launch a new browser tab that acts as a server for your notebooks. Here, on the dashboard, you can create new notebooks and modify existing notebooks and Python scripts.
You’re all set!
You may need to give pip extra permissions to get IPython installed, so if you see any error messages or exceptions mentioning permissions or installation failing, run $ sudo pip install ipython[notebook]. When you are successful, you’ll see some packages download and the command line wait for your next input.
At this point, some users have seen an error: “IPython 4.0.1 does not provide the extra ‘notebook.’” If you see this, you can resolve it by running:
$ pip install -Iv ipython[notebook]==3.2.0 instead of $ pip install ipython[notebook]
conda install ipython-notebook