Fractal is an ultimate DeFi research library for strategy development and fractaling created by Logarithm Labs.
pip install fractal-defi
Run MLFlow locally or self-hosted:
pip install mlflow
mlflow io
Than just import pre-built strategies and run the pipeline:
from typing import List
from sklearn.model_selection import ParameterGrid
from fractal.strategies import GMXV2UniswapV3Basis, BasisTradingStrategyHyperparams
from fractal.loaders import BinanceDayPriceLoader, LoaderType
from fractal.core.pipeline import DefaultPipeline, MLFlowConfig, ExperimentConfig
from fractal.core.base import Observation
# Load data and build observations
def build_observations() -> List[Observation]:
...
# Build a grid of parameters to search
def build_grid() -> ParameterGrid:
...
# Define MLFlow and Experiment configurations
mlflow_config: MLFlowConfig = MLFlowConfig(
mlflow_uri='http://127.0.01:5000',
experiment_name='my_strategy'
)
experiment_config: ExperimentConfig = ExperimentConfig(
strategy=GMXV2UniswapV3Basis,
backtest_observations=build_observations(),
window_size=24,
params_grid=build_grid(),
debug=True
)
pipeline: DefaultPipeline = DefaultPipeline(
experiment_config=experiment_config,
mlflow_config=mlflow_config
)
pipeline.run()See our discord dev channel for updates.