jmelett / pyfx

Forex Trading Platform written in Python

pyFxTrader

Scope

• The goal is to provide a backtesting and live trading tool, which can run multiple strategies on up to 10 currency pairs.
• Strategies can monitor up to three timeframes (e.g. H1, H2 and D1) and calculate buy/sell actions based on them.

Strategy

Indicators

Besides building your own, you can also use a wide range of indicators via TA-Lib or numpy.

Timeframes

• OANDA supports most commonly used timeframes. You can find all supported values here.

TODO

• Add Matplot/Plot.ly support. See also following tutorial.

• Implement proxy class for backtesting, which will first check if data is available locally and only then fetch/save via API.
• Implement usage of ETags to reduce traffic/latency.
• Prepare Makefile.

Requirements

sudo apt-get install virtualenv python-setuptools libssl-dev:

sudo apt-get build-dep python-matplotlib

Additionally get TA-Lib from http://www.ta-lib.org/hdr_dw.html

Installation

git clone [email protected]:jmelett/pyfx.git
cd pyfx
virtualenv env
source env/bin/activate
pip install -e .
envdir .my_envs python ./_cmd.py -h

Troubleshooting

In case you encounter the error `ImportError: libta_lib.so.0: cannot open shared object file: No such file or directory`

Following command* will do the trick: `export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH`

* See also http://stackoverflow.com/questions/11813279/python-wrapper-for-ta-lib-import-failure