An interactive and modular Streamlit dashboard for exploring, visualizing, and predicting FIFA player statistics.
This project showcases advanced skills in data visualization, caching, state management, and AI model integration, following professional engineering standards and a clean modular architecture.
This dashboard allows users to:
- Load and clean FIFA player data from local or remote sources.
- Explore the dataset through interactive filters and Plotly visualizations.
- Analyze trends in player performance, value, and potential.
- Predict player potential using a pre-trained Machine Learning model.
Built with a focus on:
- Modular Streamlit architecture
- Optimized caching and session state management
- Professional UI/UX principles
- Deployment-ready structure
proyecto_fifa/ │ ├── app.py # Main Streamlit app controller │ ├── modules/ # Modularized page components │ ├── page_intro.py # Page 1 - Data loading & cleaning │ ├── page_data_viz.py # Page 2 - Data visualization & filters │ └── page_model_inference.py # Page 3 - Predictive model inference │ ├── assets/ │ ├── dataset_description.html # HTML dataset dictionary (from Kaggle) │ └── model_fifa.pkl # Pre-trained model (generated locally) │ ├── data/ │ └── players_21.csv # Local FIFA dataset │ ├── create_dummy_model.py # Script to train and save a regression model ├── requirements.txt # Dependencies for reproducibility └── README.md # Project documentation
git clone https://github.com/superTay/fifa-player-analytics.git cd fifa-player-analytics --- ## ⚙️ Installation & Setup ### 1️⃣ Clone the repository ```bash git clone https://github.com/superTay/fifa-player-analytics.git cd fifa-player-analytics 2️⃣ Install dependencies pip install -r requirements.txt 3️⃣ Generate the model (if not already created) python create_dummy_model.py 4️⃣ Run the Streamlit app streamlit run app.py 🧠 Architecture & Functionality 🏁 Page 1 — Introduction Loads and cleans the dataset with @st.cache_data. Displays sample data and metadata using st.expander. Shows dataset dictionary embedded via HTML. Initializes st.session_state['df'] for global access. 📊 Page 2 — Data Visualization Includes interactive filters (multiselects, sliders). Uses @st.cache_data for unique value caching. Stores filtered DataFrame in st.session_state['df_fil']. Generates responsive Plotly charts (scatter, histogram, box, bar). Displays results dynamically based on user filters. 🤖 Page 3 — Predictive Model Loads a pre-trained regression model (LinearRegression) via joblib. Allows users to input player attributes (age, overall, value_eur, etc.). Predicts future potential rating interactively. Uses @st.cache_resource for efficient model loading. 💾 Caching & State Management Purpose Implementation Dataset caching @st.cache_data Model caching @st.cache_resource Session control st.session_state['df'], st.session_state['df_fil'] Filtering Persistent across interactions Performance Optimized data loading and resource reuse 🎨 UI/UX Principles Dark theme aesthetic for data visualization. Clear visual hierarchy with emoji section titles. Sidebar-based navigation for intuitive flow. Use of st.expander, st.columns, and st.form for a clean layout. Responsive design for desktop and wide-screen displays. 📈 Tech Stack Area Technology Frontend Streamlit Data Pandas, NumPy Visualization Plotly Express Machine Learning Scikit-learn (Linear Regression) Model Handling Joblib Deployment Streamlit Cloud / Render / Docker-ready 🧮 Model Overview A lightweight Linear Regression model trained with: y ~ overall + age + value_eur Predicts a player’s potential rating (potential). Demonstrates clean separation between training (in create_dummy_model.py) and inference (in page_model_inference.py). Cached and reusable for future sessions. 🚀 Future Enhancements Replace dummy model with an advanced Random Forest or XGBoost. Integrate real-time data via an API (e.g. sofifa.com). Add KPIs (average overall, mean market value, etc.). Extend the model for player clustering or similarity search. Include authentication for personalized dashboards. 🧠 Key Learning Outcomes This project demonstrates proficiency in: 🧩 Streamlit architecture design (multi-page modularization) ⚙️ Session state & caching optimization 🎨 Data visualization and UI/UX 🧠 Model inference integration 🧾 Professional software documentation & Git version control 🧑💻 Author Christian Marzal Della Rovere 🎓 Full Stack AI Developer (in progress) 🌐 LinkedIn • GitHub 🏷️ GitHub Topics #streamlit #data-visualization #machine-learning #fifa21 #dashboard #ai #python #fullstack ⭐ If you find this project useful, consider giving it a star on GitHub — it helps others discover it and supports the project’s visibility!