End-to-end statistical analysis of 1,898 NYC food delivery orders from Food Hub, uncovering demand drivers and building predictive models to reduce delivery time and maximise revenue.
| Analysis | Insight |
|---|---|
| Weekend vs Weekday demand | Weekend orders cost 12% more on average (p < 0.01) |
| Cuisine popularity | American & Japanese account for 58% of orders |
| Delivery time drivers | Food prep time explains 41% of variance in total delivery time |
| Revenue threshold | Orders > $20 account for 68% of total revenue |
git clone https://github.com/Jash-stack/nyc-restaurant-delivery-analytics
cd nyc-restaurant-delivery-analytics
pip install -r requirements.txt
jupyter notebook- Exploratory Data Analysis — distribution plots, correlation heatmaps, outlier detection
- Hypothesis Testing — t-tests for weekday vs weekend cost/time differences
- Regression Modelling — OLS to quantify delivery-time drivers (prep time, cuisine, cost)
- Customer Segmentation — K-Means clustering on order behaviour patterns
- Revenue Analysis — Pareto breakdown of high-value orders and cuisines
├── tests/
│ └── test_analysis.py # Unit tests for data integrity & statistical functions
├── .github/workflows/ci.yml # CI: ruff lint + pytest
├── pyproject.toml # Package metadata + tool config
└── README.md
pip install pytest pytest-cov ruff scipy
pytest tests/ -v
Jash Shah · MS Data Science, Stevens Institute of Technology · LinkedIn