goxtoolbots

goxtoolbots copied to clipboard

#GoxToolBots

A collection of trading bots implementing simple strategies for goxtool

Goxtool is a trading client for the MtGox Bitcon currency exchange. The user manual is here: http://prof7bit.github.com/goxtool/

Simple trend following bot:

strategy_simple_trend_follower.py

This bot records MtGox trades and estimates the market trend. When upward trend detected, the bot buys. When downward trend detected, bot sells.

Market trend estimation is done by using three moving price averages.

Features:

• Uses three moving average to estimate the trend of the market
• Trades following the market trend
• Download historic data for MtGox trades
• Backtest your strategy on historic date
• Plot the behavior of your bot
• Save/Restore bot state

Usage:
./goxtool.py --strategy strategy_simple_trend_follower.py

Volume gated trend following bot:

strategy_volume_trend_follower.py

Very similar to trend follower, but the trades are gated by the shift in volume. The idea is that significant price shift is always a result of a significant change in trade volumes. So unless an increased trade volume is detected, the trend follower is disabled.

Market trend estimation is done by using three moving price averages.

Usage:
./goxtool.py --strategy strategy_volume_trend_follower.py

Trailing stop loss bot:

strategy_trailing_stoploss.py

This is an implementation of a trailing stop loss trading bot for MtGox bitcoin exchange.

This bot tracks MtGox trades and records the maximum price of BTC in the last three hours. If the current price drops below 15% of the maximum, the bot sells all BTC assets as Market Order and stops. It also send a notification email to the owner.

Features:

• Sell all BTC for USD in case the BTC price dips 15% below maximum value of 3 hours
• Send an email notification when sell BTC
• Download historic data for MtGox trades
• Backtest your strategy on historic date
• Plot the behavior of your bot
• Save/Restore bot state

Usage:
./goxtool.py --strategy strategy_trailing_stoploss.py

Key commands:

Each bot supports these key commands:

• "S" - Sell all BTC
• "B" - Buy BTC for all the USD you have
• "C" - Cancell all outstanding orders
• "D" - Dump bot state (for offline plotting)

Backtesting:

python download-mtgox-data.py  

in the mtgoxdata folder downloads all the data from mtgox trades into a sqlite database. Please be careful and do not overwhelm the exchange with requests. I will try to put up a website to share this data for direct download. Meanwhile you can ping me, I can send it to you.

If you run any of the strategy_logic_*.py python scripts as a standalone porgram, it will default to backtesting its behavior on downloaded historic data. It will use matplotlib to show the prices and points of sells/buys. This is useful for developing your own bots.

Examples:
python strategy_logic_simple_trend_follower.py
python strategy_logic_trailing_stoploss.py

Please not that in the main() function of each strategy logic scripts you can tweak the simluation start and end dates.

Bots states:

Bot automatically dumps its state into a file when the strategy is unloaded. You can also manually produce a dump by pressing "D". You can plot the state of the dump with plot_pickled.py. The filename of the dump is strategy_logic_*.pickle

Email:

Trailing stop loss bot can send you an email notification when it sells. Adjust your email credentials in goxtool.ini to enable email notifications. You will need the following:

[email]
email_to = [email protected]  
email_from = [email protected]  
email_server = smtp.example.com  
email_server_port = 25  
email_server_password = TradingBotEmailPassword  

TA-Lib:

Here's how you can plug in TA-Lib if you want more technical analisys indicators

1.1 Get build essentials
sudo apt-get install build-essential
sudo apt-get install python2.7-dev
sudo apt-get install python-pip

1.2 Install TA-Lib
http://www.zwets.com/ta-lib/
sudo gedit /etc/apt/sources.list
add:
deb     http://www.zwets.com/debs    unstable/
deb-src http://www.zwets.com/debs    unstable/
sudo apt-get update
sudo apt-get install ta-lib-dev

1.2.1 If the above does not work (Goobuntu prohibits 3rd party packages)
Download http://sourceforge.net/projects/ta-lib/files/ta-lib/0.4.0/ta-lib-0.4.0-src.tar.gz/download?use_mirror=iweb
cd ~/Downloads
tar zxfv ta-lib-0.4.0-src.tar.gz
cd ta-lib
./configure --prefix=/usr
make
sudo make install

1.3 Install python wrapper
sudo pip install Cython
sudo pip install numpy
sudo pip install TA-Lib

Support:

Please consider donating to: 16csNHCBstmdcLnPg45fxF2PdKoPyPJDhX