Exploring the hubway data challenge with django and d3.js

• python-dev: Development headers for the python programming language.
• libpq-dev: This is required by the psycopg2 python library.
• postgresql-client: Specifically, the 'psql' program for creating databases.

Recommended installation is with virtualenv. The following instructions install virtualenv, activate the environment, and start the system.

1. Install virtualenv in your system's python path:

> easy_install virtualenv

2. Create a new virtual environment (this one is called 'env'):

> virtualenv env

3. Activate the virtual environment:

> . env/bin/activate

4. Install the pre-requisites:

> pip install -r requirements.txt

5. With all the requirements satisfied, create the database:

> psql -c "create role hubway with password 'hubway' login;" -U postgres
> psql -c "create database hubwayapp with owner hubway;" -U postgres

7. Set the environment variable used to connect to the database (assuming a bash shell):

> export DATABASE_URL=postgres://hubway:hubway@localhost/hubwayapp

6. With the database created, create the database schema and load the initial data (1% of all trips):

> hubway/manage.py syncdb

7. With the database created and the data loaded, you are ready to start the development server:

> hubway/manage.py runserver

The application is deployed on heroku, and uses the heroku toolbelt to perform deployments.

When running your app, the heroku toolbelt looks for a local file ".env". This contains the environment variables required for the application. It contains two local settings:

1. DATABASE_URL
2. PORT

An example of this file:

DATABASE_URL=postgres://hubway:hubway@localhost/hubwayapp
PORT=8000

Create this file, and you will be able to deploy the application locally with foreman:

> foreman start

This will start the production configuration of the application on your local system.

In order to use all the data (the above steps only install 1% of the hubway challenge data), you must truncate the trip table, and load the full_data.json data fixture:

> psql -c "truncate table hubway.trips;" -U postgres hubwayapp
> DATABASE_URL=postgres://hubway:hubway@localhost/hubwayapp hubway/manage.py loaddata full_data

This is a large set of data, and takes a while to load. There are 552,073 trips in the data, so give it a few minutes to load. It takes 10 minutes or more on my laptop.