Alpha release - please let me know if you find any bugs!
If you are looking for a full backtesting framework, please check out bt. bt is built atop ffn and makes it easy and fast to backtest quantitative strategies.
ffn is a library that contains many useful functions for those who work in quantitative finance. It stands on the shoulders of giants (Pandas, Numpy, Scipy, etc.) and provides a vast array of utilities, from performance measurement and evaluation to graphing and common data transformations.
>> import ffn
>> returns = ffn.get('aapl,msft,c,gs,ge', start='2010-01-01').to_returns().dropna()
>> returns.calc_mean_var_weights().as_format('.2%')
aapl 62.54%
c -0.00%
ge 36.19%
gs -0.00%
msft 1.26%
dtype: objectThe easiest way to install ffn is from the Python Package Index
using pip or easy_install:
$ pip install ffnSince ffn has many dependencies, we strongly recommend installing the Anaconda Scientific Python Distribution. This distribution comes with many of the required packages pre-installed, including pip. Once Anaconda is installed, the above command should complete the installation.
ffn should be compatible with Python 2.7 and Python 3.
Read the docs at http://pmorissette.github.io/ffn
A special thanks to the following contributors for their involvement with the project:
- Jordan Platts @JordanPlatts
MIT