🚌 Transit Optimization Assistant

This tool helps NGOs, transit agencies, or researchers analyze GTFS (bus stop) data to:

• 🧮 Detect redundant bus stops (too close together)
• 🧭 Identify underserved areas (no stop within 400m)
• 🤖 Use LLMs (via Groq) to explain what action to take
• 🗺️ Visualize results on an interactive map
• 📄 Export results as CSV for planning or reporting

---

📁 File Structure

BUSDATA/
├── .env                         # 🔐 Groq API key
├── app.py                      # 📌 One-click Streamlit UI
├── busDataset_fixed.xlsx       # ✅ GTFS stops with lat/lon
├── gap_detection.py            # 🧭 Detect coverage gaps
├── llm_reasoning.py            # 🤖 LLM suggestions
├── map_visualizer.py           # 🗺️ Optional standalone map generator
├── redundant_stops_directional.py  # 🔁 Redundant stop detector (with direction support)
├── redundant_stops_directional.csv # 📝 Output of redundant stop pairs
├── redundant_stops_with_llm.csv    # 🤖 LLM merge suggestions
├── requirements.txt            # 📦 Dependencies
├── transit_map.html            # 🗺️ Optional static map preview
├── utils.py                    # 🔧 Shared helper functions
│
└── gtfs_data/                  # 📥 Raw GTFS input files
    ├── routes.txt
    ├── stop_times.txt
    ├── stops.txt
    └── trips.txt

---

✅ Files to Run (In Order)

👉 Option 1: One-Click UI (Streamlit)

Just run:

streamlit run app.py

Inside the web app:

1. 📤 Upload busDataset_fixed.xlsx
2. 🧮 Click Detect Redundancies
3. 🤖 Click Generate LLM Suggestions
4. 🧭 Click Find Gaps
5. 🗺️ Click Show Map
6. 📄 Download CSVs as needed

---

👉 Option 2: CLI Mode (Manual Workflow)

# 1. Detect close stops (same route + direction)
python redundant_stops_directional.py

# 2. Generate merge suggestions via LLM
python llm_reasoning.py

# 3. Find underserved areas
python gap_detection.py

# 4. Generate interactive map (optional)
python map_visualizer.py

---

⚙️ Requirements

Install all dependencies:

pip install -r requirements.txt

Contents of requirements.txt:

streamlit
streamlit-folium
pandas
folium
openpyxl
python-dotenv
shapely
requests

---

🔐 .env Example

Create a .env file in the root with your Groq API key:

GROQ_API_KEY=your_groq_key_here

---

📦 Output Files

File	Description
redundant_stops_directional.csv	Pairs of stops <125m apart, same route + direction
redundant_stops_with_llm.csv	Suggested merges with LLM explanations
coverage_gaps.csv	Areas >400m from nearest stop
transit_map.html	Optional static map (if not using Streamlit)