BTC Aggregated Orderbook

BTC Aggregated Orderbook repository contains C++ and Python implementation which fetches orderbook data from multiple cryptocurrency exchanges, aggregates the data, and visualizes the cumulative size of buy and sell orders over time.

Features

• Fetches orderbook data from multiple exchanges (Bybit, OKEx, Binance, etc.)
• Aggregates the orderbook data into a single DataFrame
• Calculates cumulative sums for buy and sell sides
• Dynamically updates and visualizes the cumulative size using Matplotlib
• Analyzes the current price and determines which side (buy or sell) has more cumulative size within a specific price range

Prerequisites

• Python 3.x
• Pandas library
• Matplotlib library
• Requests library
• PyQT5 library (for visualization)
• Depending on your system, you may need some additional libraries for C++ implementation.

Installation

1. Clone the repository:

   git clone https://github.com/suleymanozkeskin/btc_aggregated_orderbook.git

2. Install the required libraries:

   pip install pandas matplotlib requests pyqt5

Usage

1. Run the main.py script to fetch orderbook data from the exchanges and save it to a CSV file:

   python main.py

2. Run the analyze.py script to visualize the cumulative size of buy and sell orders over time:

   python analyze.py

3. Alternatively, you can run analyze_PyQT.py to visualize the orderbook data using PyQT5:

   python analyze_PyQT.py

The script will continuously update the plot and display the current price and side with more cumulative size within a specific price range.

Contributions

Contributions to this project are welcome! If you find any issues or have suggestions for improvements, please open an issue or submit a pull request.

Authors

• Suleyman Ozkeskin - suleymanozkeskin.com
• Egehan Usluer - (https://github.com/EgehanU)

To Do

• Add more exchanges
• Add more features
• Add statistical analysis
• Depending on your needs, subscribe to the exchanges' websocket APIs to get real-time orderbook data instead of fetching it periodically.
• Currently, data is being saved to a CSV file and then read from it. This is not the most efficient way to do it. Instead, you can use a database to store the data and read it from there. Or, you can use the memory to store the data and read it from there. This might be a faster approach.
• Add a GUI to the C++ implementation for easier usage and visualization. Currently, the C++ implementation is only a console application and does not visualize the data as Python implementation does.

License

This project is licensed under the MIT License.