Predictify

Overview

A Data analysis tool to scrape your Spotify History usage and let a ML-Model predict your next songs

Authentication API

Official Documentation Authorization Code Flow

Usable possible APIs

Recently Played Tracks: /me/player/recently-played Official Spotify Documentation

Get Track: /tracks/{id} Official Spotify Documentation

Get Track's Audio Features (Deprecated): /audio-features/{id} Official Spotify Documentation

Get Track's Audio Analysis (Deprecated): /audio-analysis/{id} Official Spotify Documentation

Get Artist: /artists/{id} Official Spotify Documentation

Docker usage

cd inside the projects directory:

cd predictify

To run predictify inside a container, first make sure to build the image:

make dockerfile

Create a seperate data directory (e.g. data-docker):

mkdir data-docker

Note

To detatch the container to run it in the background add the --detach directly after the run command. Then run the following docker command, to run the container in the foreground:

docker run \
    --name predictify \
    --network=host \
    --volume $(pwd)/data-docker:/app/predictify/data \
    --volume $(pwd)/config:/app/predictify/config \
    predictify:unstable

GDPR Data

If you have gdpr data, create a folder: data/gdpr_data and add all .json files containing your play history into it. In order to extract it, run the script: python3 src/runtime.py --export

Authors

Chris Kiriakou Dominik Agres