I am trying to convert Spotify songs to vectors, so I can cluster like songs and analyze artist discography
tf-idf stands for term frequency inverse document frequency, and is the way I generate the vectors
The example I'm running my code on is Kanye's discography – I input his albums, and then I query spotify until I get playlists with multiple of his songs in them. This playlist is added to a matrix that I eventually use to define the vectors of songs.
This project will compare the clusters formed from the vectorized versions of Kanye's songs to the implied clusters defined by his albums.