Repository files navigation

Update: Here's thewinning code from Kevin.

These are instructions for recommended preparation for students in the PyCon 2015 Kaggle Tutorial.

If possible, try to check back Wednesday night for any minor updates to the data set orenvironment.yml. I'll add a note here if anything is updated.

The tutorial will be based on the datahere.

If you don't have Anaconda or Miniconda installed in your laptop, you can download them from here:

• Miniconda: Python distribution with conda package manager.[Recommended]
• Anaconda: Free enterprise-ready Python distribution with 270+ data andscientific packages.

For step-by-step instructions, visitAnaconda Install

Alternatively, command line download instructions for UNIX systems:

$ wget http://bit.ly/miniconda$ bash miniconda

If you have Anaconda, you are already setup to go.

If you have Miniconda you'll need the following libraries.

$ conda install numpy pandas scipy matplotlib scikit-learn nltk ipython-notebook seaborn

It's useful to have your dependencies in environments. Conda handles environments natively and can help you manage your Data Science projects.

The theenvironment.yml file in this repository (by downloading it, pulling the repository withgit clone https://github.com/dchudz/pycon2015-kaggle-tutorial.git, or even forking and then pulling your own copy).

Once you have either Miniconda or Anaconda, you can just run the following commands to setup your environment (from inside the directory withenvironment.yml):

$ conda env create$ source activate kaggletutorial

Note: Windows users should runactivate kaggletutorial instead.

$ ipython notebook

The tutorial will include lots of time for working on your own and in groups, so feel free to add any additional tools (e.g. for machine learning, text processing, data visualizaton, and data manipulation) you like to your environment.