webclinic017 / trading-system-4 Goto Github PK

Several .py scripts are included in this respository.

IB_API_prod.py
Connects to an Interactive Brokers gateway to collect tick data in real-time and pushes batches to a MySQL database.

Query.py
Queries data from the database to a pandas dataframe

Format_DF.py
Pre-processes the queried data and formats it for further analysis.

IndicatorsDisplaced_MP.py
Performs feature engineering on the formatted data. The script is modular such that new indicators and statistical measures can be easily added. This time series data is cross-sectional so historical displaced data is included in each observation in time. The script utilizes a process pool to run calculations across all available CPU cores.

Backtest.py
The backtest script is still very basic in terms of functionality (e.g. only one position can be open at a time). However, this provides a glimpse into the performance of the strategy on out-of-sample data.

Currently working on deep learning time-series modeling in conjunction with cross-sectional time-series.