A portfolio of end-to-end data analytics and data science projects spanning pandemic epidemiology, real estate economics, global security, sports statistics, and quantitative finance.
- About This Repository
- Projects Overview
- Technology Stack
- Getting Started
- Repository Structure
- Contributing
- License
- Author
This repository contains a curated collection of data analytics and data science projects completed as part of an ongoing portfolio. Each project follows a rigorous workflow: problem formulation → data acquisition → cleaning → exploratory analysis → modeling → interpretation.
The work spans multiple domains and tools, demonstrating proficiency in both Python and R, statistical modeling, machine learning, geospatial analysis, and financial time-series analytics.
| # | Project | Domain | Language | Focus |
|---|---|---|---|---|
| 1 | 🦠 COVID-19 Pandemic Analysis | Public Health | Python | Pandemic visualization & EDA |
| 2 | 🏠 Cracow Real Estate Pricing | Real Estate | Python | Price prediction with ensemble ML |
| 3 | 💣 Global Terrorism Database | Security / Policy | R | Geospatial EDA & trend analysis |
| 4 | ⌚ Polar Watch — Workout Vitals | Sports Analytics | Python | Statistical modeling of fitness data |
| 5 | 📈 FX Trading Algorithm Analysis | Quantitative Finance | Python | Algorithm performance & risk metrics |
| Project | Key Result |
|---|---|
| COVID-19 | ~35M cases analyzed; rural pop vs. cases correlation: -0.46 |
| Cracow Real Estate | VotingRegressor ensemble outperformed MLP, GBR, and baseline |
| Global Terrorism | 11-year EDA; ISIL, Taliban, Al-Qaida among deadliest groups |
| Polar Watch | Mixed model RMSE 61 vs. 79 (OLS); 283 workouts analyzed |
| FX Trading | 92 trades; 40% win rate; Monte Carlo: 100K simulations |
A multi-notebook end-to-end exploratory data analysis of the COVID-19 pandemic. Combines JHU CSSE case time series with World Bank socioeconomic indicators to uncover trends across time, geography, and population health metrics.
- Global scale: ~35M confirmed cases analyzed as of October 2020
- Multi-level analysis: World, continent, and country-level breakdowns
- Socioeconomic integration: Correlates pandemic metrics with GDP, life expectancy, rural population, and healthcare expenditure
- Reusable architecture: Custom
CovidDataVizclass for reproducible plotting
A complete data science workflow for predicting residential flat sale prices in Cracow, Poland — from web-scraped listing data through cleaning, exploratory analysis, feature engineering, and ensemble regression modeling.
- Ensemble modeling: VotingRegressor combining MLP and Gradient Boosting
- Rich feature engineering: 8+ derived features including log-transforms and ratio features
- Comprehensive preprocessing: KNN imputation, one-hot encoding, min-max scaling
- Interpretability: District-level price analysis reveals central vs. outlying area premiums
An exploratory data analysis and geospatial visualization of the Global Terrorism Database (GTD), covering 2007–2017. Examines attack patterns, casualty distributions, weapon and target types, and the most active terrorist organizations.
- 11-year window: 2007–2017, filtered to confirmed terrorist incidents
- Geospatial mapping: Annotation-style maps of group activity
- Multidimensional analysis: Time, geography, attack type, weapon, target, and perpetrator
- Rich visualizations: Faceted time series, correlation matrices, tile plots, stacked area charts
A statistical analysis of workout data exported from a Polar watch. Heart-rate distributions, caloric expenditure, session duration, and sport-type differences are examined, with a focus on comparing strength training against cardiovascular activity.
- 283 workouts analyzed over ~1 year
- Statistical rigor: OLS regression, linear mixed models, VIF, Goldfeld-Quandt, Shapiro-Wilk, Q-Q plots
- Activity comparison: Strength vs. cardio heart-rate profiles
- Caloric modeling: RMSE 79 (OLS) → 61 (mixed model), R² up to 0.98 by sport
Walk-forward performance analysis of an algorithmic trading system. Examines 92 trades across multiple instruments to evaluate profitability, risk characteristics, trade duration distributions, and statistical properties of returns.
- 92 trades analyzed across multiple FX instruments
- Monte Carlo simulation: 100,000 samples to estimate forward performance
- Distribution analysis: Profit-per-lot characterized by skew, kurtosis, and fitted distributions
- Timing edge: Identified profitable intraday patterns (2pm, 4pm)
| Tool | Purpose |
|---|---|
| Python 3.8+ | Primary language for 4 of 5 projects |
| R 4.0+ | Global Terrorism analysis |
| Jupyter Notebooks | All analysis and modeling |
| Git | Version control |
| Library | Domain | Projects |
|---|---|---|
| Pandas | Data manipulation | All Python projects |
| NumPy | Numerical computing | All Python projects |
| Matplotlib | Visualization | All projects |
| Scikit-learn | ML preprocessing & modeling | Flats in Cracow |
| Statsmodels | Statistical modeling | Polar |
| Scipy | Statistics & distributions | Polar, Trading |
| wbdata | World Bank API access | COVID-19 |
| tabulate | Table formatting | Polar |
| Library | Purpose |
|---|---|
tidyverse (dplyr, ggplot2, tidyr) |
Data manipulation & viz |
| GGally | Matrix & pair plots |
| rworldmap / mapproj | Geospatial visualization |
| ggrepel | Non-overlapping text labels |
| lubridate | Date handling |
| scales | Axis formatting |
Each project includes a requirements.txt (or requirements.R equivalent) for reproducible setup:
| Project | Install Command |
|---|---|
| COVID-19 | pip install -r covid19-pandemic-analysis/requirements.txt |
| Cracow Real Estate | pip install -r cracow-real-estate-pricing/requirements.txt |
| Global Terrorism (R) | See global-terrorism-eda/requirements.txt |
| Polar Watch | pip install -r polar-watch-fitness-analysis/requirements.txt |
| FX Trading | pip install -r fx-trading-analysis/requirements.txt |
- Python 3.8+
- R 4.0+ (for Global Terrorism project only)
- Jupyter Notebook or JupyterLab
# Clone the repository
git clone https://github.com/shsarv/Data-Analytics-Projects-in-python.git
cd Data-Analytics-Projects-in-python
# Install Python dependencies (example for Flats in Cracow)
pip install pandas numpy matplotlib scikit-learn joblib
# Install R dependencies (for Global Terrorism)
install.packages(c("tidyverse", "GGally", "rworldmap", "ggrepel", "mapproj", "lubridate", "scales"))Each project folder contains a dedicated README.md with specific setup instructions.
Data-Analytics-Projects-in-python/
├── README.md # This file
├── LICENSE
├── covid19-pandemic-analysis/ # 🦠 Pandemic visualization
│ ├── data/
│ │ └── download_data.py
│ ├── features/
│ │ ├── make_all.py
│ │ ├── make_cases.py
│ │ ├── make_cases_daily_change.py
│ │ ├── make_cases_since_t0.py
│ │ ├── make_continents.py
│ │ ├── make_coordinates.py
│ │ ├── make_country_stats.py
│ │ ├── make_country_to_continent.py
│ │ ├── make_mortality.py
│ │ ├── make_world_bank.py
│ │ └── utils.py
│ ├── visualizations/
│ │ └── covid_data_viz.py
│ ├── notebooks/
│ │ ├── Data-wrangling.ipynb
│ │ ├── Exploratory-analysis-globally.ipynb
│ │ ├── Exploratory_analysis_fancy_plot.ipynb
│ │ ├── Exploratory-analysis-mortality.ipynb
│ │ └── Exploratory_analysis_socioeconomic.ipynb
│ └── tests/
├── cracow-real-estate-pricing/ # 🏠 Real estate ML
│ ├── 00_Data_Wrangling.ipynb
│ ├── 00_Data_Wrangling.pdf
│ ├── 01_Exploratory_Analysis.ipynb
│ ├── 01_Exploratory_Analysis.pdf
│ ├── 02_Model.ipynb
│ ├── 02_Model.pdf
│ └── img/
├── global-terrorism-eda/ # 💣 Terrorism EDA
│ ├── Global Terrorism.ipynb
│ └── img/
├── polar-watch-fitness-analysis/ # ⌚ Sports statistics
│ ├── Polar.ipynb
│ ├── Polar.pdf
│ ├── mdl_results.txt
│ └── img/
└── fx-trading-analysis/ # 📈 Algorithmic trading
├── Trading Results Analysis.ipynb
├── Trading Results Analysis.pdf
└── img/
Contributions, issues, and feature requests are welcome. Please feel free to open an issue or submit a pull request.
- Fork the project
- Create your feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Distributed under the MIT License. See LICENSE for more information.
Sarvesh Kumar Sharma
Built with ❤️ and a lot of ☕