A multi-process ML-ready framework for effective algorithmic trading

LiuAlgoTrader

LiuAlgoTrader is a scalable, multi-process ML-ready framework for effective algorithmic trading. The framework simplify development, testing, deployment, analysis and training algo trading strategies. The framework automatically analyzes trading sessions, hyper-parameters optimization, and the analysis may be used to train predictive models.

LiuAlgoTrader can run on a laptop and hedge-on-the-go, or run on a multi-core hosted Linux server and it will automatically optimize for best performance for either.

LiuAlgoTrader uses Alpaca.Markets brokerage APIs for trading, and can use either Alpaca or Polygon.io for stocks' data. The framework is evolving to support additional brokers and data-providers.

LiuAlgoTrader is a full trading platform with a breath of tools to manage automated investment portfolios.

See LiuAlgoTrader in Action

LiuAlgoTrader comes equipped with powerful & user-friendly back-testing tool.

Watch a $4,000 Daily Profit using LiuAlgoTrader Framework for Day Trading.
Watch Trend-Following strategy beating SP-500 using LiuAlgoTrader out-of-the-box tools for Swing Trading,
Sample tear-sheet using LiuAlgoTrader sample Trend Follow strategy.
Make 30% trading pair volatility using LiuAlgoTrader.

Quickstart

Prerequisite

Paper, and/or a funded account with Alpaca Markets. OR Polygon.io subscription optional (Starter plan and above),
Installed Docker Engine and Docker Compose

Install & Configure

Step 1: To install LiuAlgoTrader just type:

pip install liualgotrader

Having issues installation? check out the installation FAQ page

Step 2: To configure the frame work type:

liu quickstart

and follow the installation wizard instructions. The wizard will walk you through the configuration of environment variables, setup of a local dockerized PostgreSQL and pre-populate with test data.

Note for WINDOWS users

Try the samples

LiuAlgoTrader quickstart wizard installs samples allowing a first-time experience of the framework. Follow the post-installation instructions, and try to back-test a specific day.

Additional samples can we found in the examples directory.

Back-testing

While Liu is first and foremost a trading platform, it comes equipped with full back-testing capabilities, providing command-line tool & jupyter notebook for analysis, and a browser-based UI covering both functionalities.

Machine Learning

These features are still work in process:

Design & Planning,
LSTM sample
Attention (Transformer) : WIP

Analysis & Analytics

The framework includes a wide ranges of analysis Jupyter Notebooks, as well as streamlit applications for analysis for both trading and back-testing sessions. To name a few of the visual analytical tools: