An application of the RPart Decision Tree package to classify music from Echonest's music database by genre, based on a set of attributes that includes tempo, key, speechiness, and duration.

Production team: Megan Barnes, Zakir Gowani, Michael Lizza

$ python MuNal.py

This script presents a TkInter GUI which interfaces with a pickled decision tree file to classify input songs into one of three broad genres (Rap, EDM, Folk). We thank The Echo Nest for access to their music data, as well as Terry Therneau, Beth Atkinson, and Brian Ripley for their RPart decision tree package.

To test if our database generation scripts work on your system:

$ sh app/test_all.sh

• pyen: Available here
• sqlite3
• rpy2: Available here
• urllib2, urlparse
• pickle
• tkinter and ttk

Choose highest versions for each of the packages above, especially for rpy2. Be sure that your R version is up to date (at least version 3) before downloading rpy2. This application was written in Python 2.X and will not run under default Python3.x settings.