Retail-IQ = ML sales forecast system. Modular architecture handle big time-series data, feature engineering, auto reports.
Retail-IQ/
├── data/
│ ├── raw/ # Unmodified input CSVs
│ └── processed/ # Cleaned and featured datasets
├── docs/ # Project documentation and reports
├── notebooks/ # Jupyter notebooks for experimentation
├── outputs/
│ ├── figures/ # Generated plots and visualizations
│ ├── models/ # Serialized model files (.pkl, .json)
│ └── logs/ # Processing and error logs
├── src/
│ └── retail_iq/ # Core Python package
│ ├── config.py # Centralized path and config management
│ ├── preprocessing.py # Data loading and cleaning
│ ├── features.py # Feature engineering logic
│ └── visualization.py # Plotting utilities
├── tests/ # Unit and integration tests
├── pyproject.toml # Project metadata and dependencies
└── requirements.txt # Legacy dependency list
# Using uv (recommended)
uv venv
source .venv/bin/activate # or .venv\Scripts\activate on Windows
uv pip install -e .Open notebooks/eda.ipynb, run all cells, generate EDA reports.
- ML: Scikit-learn, XGBoost, Statsmodels
- Data: Pandas, NumPy
- Visuals: Matplotlib, Seaborn
- API: Flask
- DevOps: Pathlib (robust paths), Setuptools (package mgmt)
