readme

Clusterfy

Clusterfy extracts songs from a user’s Spotify playlists and applies k-means clustering to group songs based on fundamental features such as tempo and key signature. It then provides a visualization of this data by extracting the first 3 principal components from each song’s features and plotting the songs on a 3D chart. Finally, it recommends a playlist based on these clusters and inserts it into the user's Spotify account. 


How to Use:

1) Start Clusterfy
	
- Run the command "python music_clustering.py"
- Go to "http://localhost:5000/" in your web browser


2) Request an OAuth Token from https://developer.spotify.com/web-api/console/post-playlists/
	
- Fill in your Spotify username and press "Get OAuth Token"
- Check "playlist-modify-public" and "playlist-modify-private" and press "Request Token"

*Note: OAuth tokens expire after a certain period of time and you will have to request a new one*


3) Enter User Information into Clusterfy
	
- Enter your Spotify username into the text field "Username"
- Copy and Paste the OAuth token from the previous step into the text field "Auth Token"
- Wait for Clusterfy to finish processing your songs


4) Add Playlists
	
- Check out our cool data visualization of your songs!
- If you want to add a playlist built around one of the clusters, click on the corresponding button on the right side of the plot