Installation¶
Requirements¶
- Python 3.9 or higher
- pip or conda package manager
Basic Installation¶
Install FeatCopilot using pip:
This installs the core package with support for:
- Tabular feature engineering
- Time series feature extraction
- Feature selection methods
- Scikit-learn integration
Installation with LLM Features¶
FeatCopilot supports two LLM backends:
Option 1: GitHub Copilot SDK (Default)¶
This installs the GitHub Copilot SDK for native Copilot integration.
Option 2: LiteLLM (100+ Providers)¶
This installs LiteLLM, supporting OpenAI, Azure, Anthropic, Google, GitHub Models, and 100+ other providers.
Full Installation¶
For all features including both LLM backends and time series analysis:
This includes:
- GitHub Copilot SDK
- LiteLLM (100+ providers)
statsmodelsfor advanced time series features
Development Installation¶
For contributing to FeatCopilot:
# Clone the repository
git clone https://github.com/thinkall/featcopilot.git
cd featcopilot
# Install in development mode
pip install -e ".[dev]"
Development dependencies include:
pytest- Testing frameworkblack- Code formattingruff- Lintingmypy- Type checking
Verify Installation¶
import featcopilot
print(featcopilot.__version__)
# Output: 0.1.0
from featcopilot import AutoFeatureEngineer
print("Installation successful!")
Dependencies¶
Core Dependencies¶
| Package | Version | Purpose |
|---|---|---|
| numpy | >=1.21.0 | Numerical operations |
| pandas | >=1.3.0 | Data manipulation |
| scipy | >=1.7.0 | Scientific computing |
| scikit-learn | >=1.0.0 | ML utilities & transformers |
| pydantic | >=2.0.0 | Configuration validation |
| joblib | >=1.1.0 | Parallel processing |
Optional Dependencies¶
| Package | Version | Purpose |
|---|---|---|
| github-copilot-sdk | >=0.1.0 | GitHub Copilot SDK (default LLM backend) |
| litellm | >=1.0.0 | LiteLLM (100+ LLM providers) |
| statsmodels | >=0.13.0 | Advanced time series |
Troubleshooting¶
ImportError: No module named 'featcopilot'¶
Ensure you've installed the package:
LLM features not working¶
- Install LLM extras:
pip install featcopilot[llm] - Set up your LLM provider API key (see Authentication)
- Ensure your API key is valid and has sufficient quota
Dependency conflicts¶
Try creating a fresh virtual environment: