Millennium Quantitative Research Playground Project Setup and Usage

Installation

1. Clone the repository:

   git clone https://github.com/jerrychen04/millennium-data-quality.git
   cd millennium-data-quality

2. Create a conda environment: If you do not have conda installed, install it here. Make sure you cd into the root of this repository before running this cmd:

   conda env create -f environment.yml

3. Activate the conda virtual environment:

   Using conda (make sure to activate everytime you create a new shell instance):

   conda activate data-quality

Running the Main Script (sample mean reversion strategy implementation) as .py file

To run the main script, execute the following command:

python backtester/main.py

Running sample research notebooks:

To run a .ipynb, like bab.ipynb, simply select the data-quality kernel and run the scripts. If not visible, select under > Select Another Kernel > Python Environments. [IMPORTANT] When running on cached data, make sure date ranges align with your order generator.

Running Tests

To run the unit tests, use the following command. Note: tests are not freshly maintained at the moment:

python -m unittest discover -s unit_tests

To Create A Strategy

Go to order_generator.py and create a new instance of OrderGenerator in a new file for cleanliness. See example code in main.py to run strategy. Write corresponding unit tests as needed. Run research runs in .ipynb and restart kernel when making package changes.