Density

Density estimates how full different parts of Columbia are, based on the number of devices connected to the WiFi (data graciously provided by CUIT in coordination with ESC).

Contributing

Local Dev

We recommend using Vagrant for development environment set-up. If you have Vagrant, running vagrant up should set-up and provision the virtual machine for you.

Inside the virtual machine, run:

cd /vagrant
source config/settings.dev
flask run --host=0.0.0.0

and go to https://localhost:5000. You should see your own local Density app!

Dependencies

Density is build with Python 3.6 and Flask. We use a conda environment, defined via config/environment.yml, for dependency management. When you run source config/settings.dev, the appropriate conda environment will automatically be activated.

Density runs on Postgres 9.6. To access the database, you can just run:

sudo -u postgres psql density

You can find a copy of the database schema in config/schema.sql.

Testing

We use py.test for testing and flake8 for linting. All tests are defined in density/tests. To run tests locally, in the app root directory you should run:

source config/settings.dev
py.test
flake8

We have Travis CI set-up to enforce passing tests.

Deployment

Density is currently deployed on ADI's server via Docker (defined in the Dockerfile). To build the Docker image locally, install Docker and run:

docker build -t density .
docker run --net=host -d density

Project Layout

.
├── config
│   ├── bootstrap.sh      -- bootstrap script for Vagrant
│   ├── dump.sql          -- a database dump for local development
│   ├── environment.yml   -- a file defining our conda environment
│   ├── schema.sql        -- our database schema
│   └── settings.dev      -- various configuration settings for development
├── density
│   ├── config.py         -- loads configuration from environment variables
│   ├── data.py           -- stores various constants
│   ├── db.py             -- app interface to the Postgres database
│   ├── __init__.py       -- the bulk of the app logic
│   ├── predict.py        -- in-progress work for data science work
│   ├── static/
│   ├── templates/
│   └── tests/
├── API.md                -- API documentation
├── Dockerfile            -- Dockerfile for production
├── README.md
├── run.py                -- File for Flask to run
├── setup.cfg
└── Vagrantfile