palist

Auth flow

• Front end prompts user login w/ Spotify
• Pass user 'id' attribute to back end for data analysis
  • Auth scopes listed here. We will likely use user-library-read, user-follow-read, user-read-recently-played, and user-top-read

During testing, back end can prompt user login with Spotipy as done in scratch/test_auth.py

Get started with Docker for development

In palist/docker, run the following commands to access bash in a docker image with all Python dependencies in requirements.txt installed and with this repo mounted at usr/src/app :

docker build -t palist-app . 
docker run -it -p 8888:8888 -v [PATH/palist]:/usr/src/app palist-app /bin/bash
jupyter notebook --ip 0.0.0.0 --no-browser --allow-root # (to run jupyter)

We Dockerize our app once it is further along (a simple guide).

Starting the Flask app

From the root directory of the project, run:

cd musicanalyzer
export FLASK_APP=musicanalyzer.py
python -m flask run

Finally, navigate to the local url where the app is being served (127.0.0.1:5000)

Notes

• API credentials are here, connected to Sarah's spotify acct

Backend/Data Analysis Tasks

• Make Dockerfile (for 2/23)
• Set up test auth workflow (for 2/23)
• Explore audio features (for 3/1)
  • Look at how to combine features for groups of tracks (while capturing variation of taste)
    • Similarity scores (like R^2) based on how well users correlate in the histogram of each of their audio feature --> NMDS plot