🌱 Soil Analysis

A Python project that predicts soil fertility and provides AI-driven improvement recommendations for better crop yield.
It combines Random Forest classification with AI-based nutrient analysis to suggest actionable soil improvement steps.

Live Project link: https://huggingface.co/spaces/pymit/Soil-Analysis

---

Features

• Predict soil fertility using Random Forest based on N, P, K, and pH values.
• AI-powered suggestions for improving soil nutrients (Nitrogen, Potassium,Phosphorus and pH).
• Automatically calculates whether each nutrient is ample or deficient.
• Provides real-world solutions: fertilizer type, compost, or soil amendments with suggested amounts.
• Integration with MLflow for experiment tracking and logging metrics.
• Streamlit frontend for interactive input and visualization.

---

Installation

1 Clone the repository:

git clone https://github.com/mitanshhh/Soil-Analysis.git

2 Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # Linux / macOS
venv\Scripts\activate     # Windows

3 Install required dependencies:

pip install -r requirements.txt

4 Create a .env file and add your API key for ChatGroq:

GROQ_API_KEY=your_api_key_here

Usage

Run the Streamlit app:

streamlit run main.py

---

Enter Nitrogen, Phosphorus, Potassium, and pH values for your soil.

Click Predict Fertility.

View the predicted fertility status.

Check AI-driven improvement recommendations for each nutrient.

Example Output

[
  ["Nitrogen", "deficient", "Apply urea at 100 kg/ha or add nitrogen-rich compost such as cow dung manure..."],
  ["Potassium", "deficient", "Incorporate potassium sulfate at 75 kg/ha or potassium-rich organic amendments..."],
  ["ph", "ample", "Use elemental sulfur or acidifying organic matter to lower pH by approximately 3.2 units..."]
]

Project Structure

Soil-Analysis/
├── ML/
│   └── soil_analysis.pkl          # Trained Random Forest model
|   └── ml_model.py                # Python Code for RFC
├── Dataset/
│   └── soil_data.csv              # Soil dataset
|   └── Crop_recommendation.csv    # Actual dataset used for analysis and data manipulation
├── main.py                        # Streamlit frontend + backend integration
├── data_analysis.py               # Python code used for data analysis
├── .env                           # Environment variables (GROQ_API_KEY)
├── README.md                      # Project documentation
└── requirements.txt               # Project dependencies

Keypoint

MLflow Integration

Tracks precision, recall, and F1-score for your Random Forest model.

Logs model versioning and metrics for reproducibility.

License: "MIT License"

Acknowledgements:

- Streamlit for interactive UI
- LangChain for AI integration
- MLflow for experiment tracking