Jitenshea: bicycle-sharing data analysis

It's an archived repo.

MOVED TO https://gitlab.com/Oslandia/jitenshea

In Japanese:

Jitensha (bicycle) + Shea (share) = Jitenshea

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

You have three parts in this project:

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

• Some statistics and Machine Learning to analyze the time series and predict the bikes availability for each station.

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

Data

Open Data from the 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.

Contributions for other cities are welcomed! e.g. Nantes, Paris, Marseille, etc.

Configuration

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.

Requirements

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

All Python requirements are specified in the setup.py file.

Create a new virtualenv, then do pip install . in the project root directory.

For Javascript and CSS dependencies, you have to install yarn then launch yarn install.

Development

• Install the extras dependencies, e.g. pip install -e ."[dev]".
• Launch test with pytest.

Logo

Logo was designed by Sylvain. Thanks to him!