RICH

Introduction

Arbitrage is a challenging data science problem characterized by rapidly fluctuating price discrepancies across token pairs, necessitating real-time solutions. To address this challenge, we model it as a (k)-hop negative cycle detection problem in token graphs and introduce RICH: Real-time Identification of negative Cycles for High-efficiency arbitrage.

Repository Overview

This repository contains the implementation of RICH, including the source code and sample datasets. The code consists of:

• A preprocessing script for graph conversion.
• A compiled executable for negative cycle detection.
• Scripts for running experiments on sample datasets.

Compilation and Execution

1. Convert Graph to CSR Format

Before running RICH, the input graph must be converted into a compressed sparse row (CSR) format. Run the following command:

python convert_to_csr.py original_graph_path csr_graph_path

2. Compile the Code

Compile the source code by running:

mkdir build
cd build
cmake ..
make

3. Run the Executable

After compilation, run the binary for negative cycle detection:

./DPWithFilterFull

If the at-most k variant is preferred, then use:

./DPWithFilterFullAtMostk

Then, specify the dataset, k, number of coloring process.

Data

This repository includes sample datasets to test RICH. The datasets contain transaction-based graphs where nodes represent tokens and edges represent token exchange interactions.

Usage Example

To execute RICH on a dataset, follow these steps:

1. Convert the input graph into CSR format.
2. Compile the source code.
3. Run the compiled executable on the processed graph.

Contact

For questions or collaboration inquiries, feel free to contact us.

Disclaimer

RICH is developed exclusively for academic research. Its deployment in real-world trading scenarios must account for regulatory and ethical considerations. Any practical use is at the user's own risk and should carefully consider multiple market factors such as slippage and execution delays.

---