netanelbi / lanturn

Pre-requisites

Run archive node, serving Ethereum HTTP JSON-RPC on port 8545

Install common software dependencies: python, npm

Python Modules: pip3 install GitPython pandas tqdm scikit-learn

Simulation Clients

The following scripts by default launch 44 simulation networks.

cd eth_clients/hardhat

bash setup_hardhat.sh

bash launch_hardhats.sh

eth_clients/hardhat directory has detailed instructions on managing sim clients

Download data:

Download and uncompress the zip file from the URL: https://lanturn.s3.us-east-2.amazonaws.com/dataset.zip

Copy the files from dataset/data/ to data/ directory of this repository. This data is used for deriving the prices for Binance (CEX settlement of non-ETH currencies), although optimizations and simulations can run without this data, in which case the settlement will occur only on dexes.

problem-dataset dir contains the input transactions (user transactions and templates) and domain for optimization.

baseline-dataset dir contains the flashbots baseline for the above problems.

Sample optimization command:

python3 optimize.py -t <example_transactions> -d <example_domain> --dexes <contract_list> --n_parallel_gauss 44 --capital 1000 --reorder

where contract_list can be a sublist of ['sushiswap', 'aave', 'uniswapv3', 'uniswapv2', 'uniswapv3-jit'] depending on the contracts and experiments involved.

python3 runme.py script automatically launches simulation clients (setup has to be called manually first) and runs an example optimization.

If you want to run simulation standalone, sample command: python3 simulate_client.py -f <concrete_transactions> -p <PORT_ID> where PORT_ID = <PORT of sim client> - 8601

Example files

concrete_transactions : examples/optimised

example_transactions : examples/0xeth/multiple_tokens

example_domain : examples/0xeth/multiple_tokens_domain