/pycon-pandas-tutorial

PyCon 2015 Pandas tutorial materials

Primary LanguageJupyter NotebookMIT LicenseMIT

Welcome to Brandon’s Pandas Tutorial

The first instance of this tutorial was delivered at PyCon 2015 in Montréal, but I hope that many other people will be able to benefit from it over the next few years — both on occasions on which I myself get to deliver it, and also when other instructors are able to do so.

If you want to follow along with the tutorial at home, here is the YouTube recording of the 3-hour tutorial at PyCon itself:

https://www.youtube.com/watch?v=5JnMutdy6Fw

To make it useful to as many people as possible, I hereby release it under the MIT license (see the accompanying LICENSE.txt file) and I have tried to make sure that this repository contains all of the scripts needed to download and set up the data set that we used.

Quick Start

If you have both conda and git on your system (otherwise, read the next section for more detailed instructions):

$ conda install --yes ipython-notebook matplotlib pandas
$ git clone https://github.com/brandon-rhodes/pycon-pandas-tutorial.git
$ cd pycon-pandas-tutorial
$ build/BUILD.sh
$ ipython notebook

Detailed Instructions

You will need Pandas, the IPython Notebook, and Matplotlib installed before you can successfully run the tutorial notebooks. The Anaconda Distribution is a great way to get up and running quickly without having to install them each separately — running the conda command shown above will install all three.

Note that having git is not necessary for getting the materials. Simply click the “Download ZIP” button over on the right-hand side of this repository’s front page at the following link, and its files will be delivered to you as a ZIP archive:

https://github.com/brandon-rhodes/pycon-pandas-tutorial

Once you have unpacked the ZIP file, download the following four IMDB data files and place them in the tutorial’s build directory:

  • ftp://ftp.fu-berlin.de/pub/misc/movies/database/actors.list.gz
  • ftp://ftp.fu-berlin.de/pub/misc/movies/database/actresses.list.gz
  • ftp://ftp.fu-berlin.de/pub/misc/movies/database/genres.list.gz
  • ftp://ftp.fu-berlin.de/pub/misc/movies/database/release-dates.list.gz

To convert these into the CSV files that the tutorial needs, run the BUILD.py script with either Python 2 or Python 3. It will create the three CSV files in the data directory that you need to run all of the tutorial examples. It should take about 5 minutes to run on a fast modern machine:

$ python build/BUILD.py

You can then start up the IPython Notebook and start looking at the notebooks:

$ ipython notebook

I hope that the recording and the exercises in this repository prove useful if you are interested in learning more about Python and its data analysis capabilities!

Brandon Rhodes