Quick Start

Installation

# Clone the repository
git clone https://github.com/your-org/feature-forge.git
cd feature-forge

# Install with uv
uv sync --all-extras

# Or with pip
pip install -e ".[all,dev]"

Basic Usage

Sklearn API

from feature_forge.api import MALMASFeatureEngineer

fe = MALMASFeatureEngineer()
fe.fit(X_train, y_train)
X_test_enhanced = fe.transform(X_test)

Experiment Matrix

from feature_forge.experiment import ExperimentMatrix, ExperimentRunner

matrix = (
    ExperimentMatrix()
    .datasets(["titanic"])
    .methods({"malmas": ["full"], "openfe": ["openfe"]})
    .seeds([0, 1, 2])
)

runner = ExperimentRunner()
results = runner.run(matrix.generate(), run_experiment)

Custom Agent

from feature_forge.agents import Agent, BaseFeatureAgent

class MyAgent(BaseFeatureAgent):
    prompt_filename = "my_prompt.txt"
    agent_name = "my_agent"

Register in pyproject.toml:

[project.entry-points."feature_forge.agents"]
my_agent = "my_package:MyAgent"

Configuration

Environment variables:

export FF_TASK=classification
export FF_LLM__MODEL=deepseek-chat
export FF_LLM__API_KEY=sk-...

Or edit config/settings.yaml.

Running Tests

# All tests
uv run pytest

# With coverage
uv run pytest --cov=feature_forge --cov-report=html

# Linting
uv run ruff check src tests