Spotify Artists Rabbit Hole

• Retrieves songs from n artists related to other n artists and so on starting from a base artist.
• Builds a graph with the relationship of every artist displayed within streamlit.
• Extracts most popular tracks with metadata of retrieved artists

---

Usage

Create data

Set environment variables

SPOTIFY_USERNAME
SPOTIFY_CLIENT_ID
SPOTIFY_CLIENT_SECRET

pip install -r requirements.txt

Run create_data.py to create the datasets

python api/create_data.py --artist "The Beatles" \
                          --max-artists 4 \
                          --n-related 2 \
                          --n-tracks 2 \
                          --output-dir sample_data

2 output files will be created inside the output directory

── sample_data
   ├── graph.pkl
   └── songs.csv

Streamlit

Locally

pip install -r requirements.txt

Change serverAddress to localhost in .streamlit/config.toml if its set to 0.0.0.0

streamlit run streamlit_app/app.py

Docker

Build the image

docker build -t st_rabbit_hole:latest .

Run the container

docker run --rm -p 8501:8501 st_rabbit_hole:latest

---

Examples