Biggy Gains

Biggy Gains is a trading bot being developed by myself and my friend Chris just for fun. This repository contains the environment layer used by the bots to interact with their world, as well as the bots themselves. Live trading and live paper trading will be done with Alpaca.

Setup

Create virtual environment:

virtualenv venv

Activate your virtual environment:

# Windows
source venv/Scripts/activate

# Linux/OSX
source venv/bin/activate

Install dependencies:

pip install -r requirements.txt

Run tests:

python -m unittest discover -s tests

Running

Biggy Gains is organized into two main components; Environment and Bot. Environment is the main interaction point between the Bot and the real world. This allows historical back-testing, live paper trading, and live trading to all be done without any code change in the bot being used. All that needs to be done is swap the Environment. Both the Environment and Bot are specified as command line options.

Configuration

The following options are able to be set through environment variables or the command line. Values passed in via the command line always override values in environment variables.

REDDIT_KEY (--reddit-key): Key id used to connect to Reddit. Required for LiveEnvironment
REDDIT_SECRET (--reddit-secret): Key secret used to connect to Reddit. Required for LiveEnvironment
ALPACA_URL (--alpaca-url): Endpoint to make trades through. Paper or live. Required for LiveEnvironment
ALPACA_KEY (--alpaca-key): Key id to connect to Alpaca with. Required for LiveEnvironment
ALPACA_SECRET (--alpaca-secret): Key secret to connect to Alpaca with. Required for LiveEnvironment

Run python main.py --help for full configuration options.

Starting Biggy Gains

Biggy gains may be ran with the following command:

python main.py [configuration parameters] <environment> <bot> <datastore>

Environment is one of:

live: Live environment. Connects to Reddit and Alpaca and supports both real and paper trading
More to come...

Bot is one of:

ben_sentiment_bot: Gathers sentiment data from Reddit and trades the most positive securities
More to come...

Datastore is one of:

memory: In memory datastore with no persistence. Good for testing
More to come...

Tools

Due to the need for labeled data sets and trained models several tools are contained within this repository. Implemented so far are: