This project is an interactive web-based tool for computing and visualizing the shortest path using Dijkstra's algorithm on raster map data. It supports environmental planning by helping users identify cost-effective and potentially more environmentally friendly powerline routes.

• ✅ Interactive map with raster visualization (.tif)
• ✅ Click-to-select start and goal points
• ✅ Dijkstra algorithm with 8-directional movement
• ✅ Path visualization and export to .txt
• ✅ Environmental analysis support via LLM (Streamlit + Meta-LLaMA)

• Dash (Plotly) – Interactive UI
• NumPy – Matrix computations
• Rasterio – Raster data loading
• Streamlit – Environmental impact assistant
• OpenAI / LLaMA API – Natural language evaluation
• GeoJSON / TXT – Data input/output

📁 root
├── dash_app.py             # Dash UI and Dijkstra pathfinding
├── dijkstra.py             # Dijkstra algorithm (8 directions) + save to txt
├── landuse_cutout.tif      # Raster map (not included here)
├── jihocesky_median_bezoutliers.geojson  # Area map data (geojson)
├── proposed_route.txt      # Example input route
├── villages_layers.txt     # Area environmental layers
├── main.py                 # Streamlit LLM-based analysis tool

git clone https://github.com/kruzek/GreenHack.git
cd GreenHack

pip install -r requirements.txt

(Make sure you have a Python 3.8–3.11 environment.)

python dash_app.py

Go to http://127.0.0.1:8050 in your browser.

streamlit run main.py

1. Load the landuse_cutout.tif raster.
2. Let the user select start/goal points.
3. Apply Dijkstra's algorithm with cost weighting and 8 directions.
4. Visualize the path on a heatmap.
5. Allow download of path as .txt.

• Loads villages_layers.txt and proposed_route.txt
• Takes user messages and proposed path
• Uses Meta-LLaMA 3 to provide a structured environmental impact analysis
• Suggests improved or shortened routes based on impact

Created for the GreenHack 2025 challenge by the GreenHulk team.