In Japanese:

Jitensha (bicycle) + Shea (share) = Jitenshea

Analyze bikes sharing station data some cities where there are Open Data.

You have two parts in this project:

• a data pipeline and data processing with luigi to get, transform and store data

• Web application to get and visualize some data through a REST API

Open Data from French cities Bordeaux and Lyon:

• Data Grand Lyon Website
• Data Bordeaux Website

Some luigi tasks can be called every 10 minutes for instance to gather the bicycle-sharing stations data. Another one is called every day to aggregate some data. You can use cron-job to carry out this stuff.

A configuration file sample can be found at the root directory config.ini.sample. Copy it into the jitenshea directory , rename it into config.ini and update it.

It is used for the database access, some tokens for API, etc.

For the wep application, you can launch bower to install Javascript and CSS dependencies.

PostgreSQL database with PostGIS. You must have the shp2pgsql command.

• Python 3.6
• pandas
• requests
• luigi
• sh
• psycopg2
• sqlalchemy
• flask-restplus
• daiquiri

See the conda_env.sh script to create a conda environment with the dependencies.

Note: flask-restplus and daiquiri should be install via pip.

Install the extras dependencies, e.g. pip install -e ."[dev]", and run pytest.