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
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 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.
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
concrete_transactions : examples/optimised
example_transactions : examples/0xeth/multiple_tokens
example_domain : examples/0xeth/multiple_tokens_domain