Everyone Everywhere Music Analysis Tool
By Brendan Graham
November 27, 2019
This app calls the Spotify API and leverages the “Features” that Spotify uses to classify songs. This project started when we were joking that we could use AI to write a new Everyone Everywhere song. That conversation reminded me of a
blog post I had seen where the author used the spotifyR
package to get song data and analyze the songs. I thought it would be interesting to analyze my own band’s songs in the same manner, and also provide anyone a way to search for bands and see how Spotify classifies their songs in terms of energy, danceability, valence, liveness, etc…
Check out the app here
Upon load, the app displays an analysis of Everyone Everywhere tracks. A user can pick a feature, such as energy, and the app will rank tracks according to energy, and show a distribution of track energy for every album associated with the band. The lower left chart uses k-means to cluster artist tracks, and the lower right radar charts show how albums compare to each other across several features.
View the project on GitHub.