With 24 movies (and counting) in the James Bond franchise, this project hypothesized that the films' soundtracks contained auditory markers which could potentially identify which actor portrayed Bond.

Data was collected and processed with help from the following sources:

• Spotify API
• Spotipy
  • Python library for the Spotify api
• Internet Movie Database
• Wikipedia's List of Bond Films
• Wikipedia's List of Pitch Classes
• PostgreSQL
• Psycopg2
  • Python library integrating PostgreSQL

Data was collected in Python, and then processed in R

• Python
  • keys.py - Not included in the repository. Stores values for 'client_ID', 'client_secret' and 'oauth_token' required to access the Spotify API.
  • films_scrape.py - Fetches the list of films from Wikipedia and should be called first.
  • Album_IDs.py - Inserts the album_id into the database, and should be called second.
  • GetTracks.py - Inserts data for both Albums and individual tracks into the database. Must be run after above scripts.
  • PitchClass.p - May be called at anytime before processing in R
• R
  • Analysis.R - Calls all other necessary R scripts.
  • Packages Used:
    • RPostgreSQL
    • caret (and its dependencies)
    • klaR (for Kernel-based Naive Bayes Classification)
    • ggplot2
    • ggfortify
    • dplyr