A Datafeed System for Financial Data.
Python Shell
Latest commit 6219327 May 30, 2014 @yinhm do not using update_multi for put_xminute when markets minutes are no…
…t 1440, need to fix timestamp_to_index first.

README.md

A Datafeed System for Financial Data.

Datafeed is an fast, extensible quotes data storage build on Python/HDF5. IMIGU (http://imigu.com) has been using Datafeed on production more than a year.

Datafeed is licensed under the Apache Licence, Version 2.0 (http://www.apache.org/licenses/LICENSE-2.0.html).

Components

  • Quotes store server.
  • Client to interactive with quote server(get/put).
  • Datafeeds providers client, including Yahoo, Google, etc.

Python version support

Officially 2.7. Python 3 may works but not tested.

Dependencies

  • NumPy: 1.5.0 or higher
  • h5py: 2.0 or higher
  • tornado: 2.0 or higher

Install dependent packages with pip:

pip install -r pip.txt

Optional packages

  • pandas: dividend/split and more
  • python-dateutil: <2.0, RSS parsing
  • pycurl: url fetch
  • pywin32: only needed if you want to run TongShi client

INSTALLATION

git clone git://github.com/yinhm/datafeed.git

Run

cd datafeed
cp config_example.py config.py
python server.py

Client

from datafeed.client import Client
c = Client()
c.get_report("SH000001")

TODO

  • Documentation

License

Apache Public License (APL) 2.0

Thanks

  • Big thanks to my company ( http://jeebo.cn ) allow me to open source Datafeed.