Yhack 2019 project for Daniel Knopf, Sam Ephron, and Nalu Tripician, focusing on using machine learning to find the most karaoke-able song in a user's Spotify library :)

Verse-a-tility searches through a user's Spotify playlists and creates a new playlist called "Your Top Ten Karaoke Songs! Found by Verse-a-tility". The playlist contains the ten songs with audio stats that most closely match the average of the audio stats of the 230 songs that were scraped from the Google results for "Best karaoke songs," with extra emphasis on danceability, energy, and low instrumentalness.

TO USE: Run spotify.py and type in your spotify username

Possible future plans:

Make executable Username input usability for nontech people Investigate Standard deviations for weighting the average Maybe: How karaokey is your music?