The Forex environment is a forex trading simulator featuring: configurable initial capital, dynamic or dataset-based spread, CSV history timeseries for trading currencies and observations for the agent, fixed or agent-controlled take-profit, stop-loss and order volume.
The environment features discrete action spaces and optionally continuous action spaces if the orders dont have fixed take-profit/stop-loss and order volume.
A concatenation of num_ticks vectors for the lastest: vector of values from timeseries, equity and its variation, order_status( -1=closed,1=opened),time_opened (normalized with max_order_time), order_profit and its variation, order_drawdown /order_volume_pips, consecutive_drawdown/max_consecutive_dd
discrete action 0: 0=nop,1=close,2=buy,3=sell
discrete action 0 parameter: symbol
(optional) continuous action 0 parameter: percent_tp, percent_sl,percent_max
The reward function is the average of the area under the curve of equity and the balance variation.
The datasets used for the tests were generated with a MQL4 program located in the agents folder (Pending documentation).
Install Python, pip, OpenAI Gym and other dependencies:
sudo apt-get install -y python3-numpy python3-dev cmake zlib1g-dev libjpeg-dev xvfb ffmpeg libboost-all-dev libsdl2-dev python3-pip git gcc make perl
pip3 install graphviz neat-python gitpython gym neat-python matplotlib
git clone https://github.com/harveybc/gym-forex
Set the PYTHONPATH venvironment variable, you may add the following line to the .profile file in your home directory to export on start of sessions. Replace with your username.
export PYTHONPATH=/home/username/gym-forex/:${PYTHONPATH}
cd gym-forex
nano agents/config
Configure the population size and other parameters according to your computing capacity or requirements, start with the defaults.
nano res
For example:
#!/bin/bash
git pull
python3 agents/agent_NEAT.py ./datasets/ts_5min_1w.CSV ./datasets/vs_5min_1w.CSV config_20
After editing, change the permission of the file to be executable:
chmod 777 res
./res