Crop Recommendation System

A 3-tier application for predicting crop recommendations based on agronomic features.

Architecture

┌─────────────────────────────────────┐
│      React Frontend (Port 3000)      │
└──────────────┬──────────────────────┘
               │
┌──────────────▼──────────────────────┐
│  Java/Jersey Backend (Port 8080)     │
└──────────────┬──────────────────────┘
               │
┌──────────────▼──────────────────────┐
│   Python Flask ML Service (5000)     │
├──────────────────────────────────────┤
│  Scikit-learn Crop Classifier        │
│  - Logistic Regression               │
│  - 22 Crop Classes                   │
│  - 95.91% Accuracy                   │
└──────────────────────────────────────┘

┌──────────────────────────────────────┐
│   PostgreSQL Database (5432)         │
│   - Prediction History               │
│   - User Accounts                    │
└──────────────────────────────────────┘

Project Structure

Predictor/
├── predictor.ipynb              # ML model development notebook
├── Crop_recommendationV2.csv    # Training data
├── crop_label_pipeline.joblib   # Trained model artifact
├── label_map.json               # Label mapping (code -> crop name)
│
├── ml-service/                  # Python Flask ML API
│   ├── src/
│   │   ├── models/
│   │   │   └── predictor.py     # CropPredictor class
│   │   └── utils/
│   │       └── validation.py    # Input validation
│   ├── app.py                   # Flask REST API
│   ├── requirements.txt          # Python dependencies
│   ├── .env.example              # Environment variables
│   └── .gitignore
│
├── backend/                     # Java/Jersey REST API
│   ├── src/main/java/com/crop/
│   │   ├── resources/           # JAX-RS endpoints
│   │   ├── services/            # Business logic
│   │   ├── dao/                 # Database access
│   │   ├── models/              # DTOs and entities
│   │   └── config/              # Configuration
│   ├── src/main/resources/      # Config files
│   ├── pom.xml                  # Maven build config
│   └── .gitignore
│
├── frontend/                    # React SPA
│   ├── src/
│   │   ├── components/          # React components
│   │   ├── pages/               # Page components
│   │   ├── App.js               # Main app
│   │   └── index.js             # Entry point
│   ├── public/
│   │   └── index.html           # HTML template
│   ├── package.json             # Node dependencies
│   └── .gitignore
│
├── database/                    # Database setup
│   ├── init.sql                 # Schema initialization
│   └── migrations/              # Database migrations
│
├── scripts/                     # Startup scripts
│   ├── run-ml-service.sh       # Start ML service
│   ├── run-backend.sh          # Start backend
│   └── run-frontend.sh         # Start frontend
│
├── docs/                        # Documentation
│   ├── SETUP.md                # Environment setup
│   ├── API.md                  # API documentation
│   └── ARCHITECTURE.md         # Architecture details
│
└── README.md                   # This file

Quick Start

Prerequisites

• Python 3.9+ (for ML service)
• Java 11+ (for backend)
• Node.js 16+ (for frontend)
• PostgreSQL 12+ (for database)

1. Start ML Service

chmod +x scripts/run-ml-service.sh
./scripts/run-ml-service.sh
# Runs on http://localhost:5000

2. Start Backend

chmod +x scripts/run-backend.sh
./scripts/run-backend.sh
# Runs on http://localhost:8080

3. Start Frontend

chmod +x scripts/run-frontend.sh
./scripts/run-frontend.sh
# Runs on http://localhost:3000

4. Configure Database

psql -U postgres < database/init.sql

Services

ML Service (Python Flask)

REST API for crop prediction using Scikit-learn model.

Endpoints:

• GET / - Health check
• GET /predict - Usage instructions
• POST /predict - Single or batch prediction

Example Request:

curl -X POST http://localhost:5000/predict \
  -H "Content-Type: application/json" \
  -d '{
    "N": 50, "P": 40, "K": 50, "temperature": 25.5,
    "humidity": 80, "ph": 7.0, "rainfall": 100,
    "soil_moisture": 40, "soil_type": "Loamy",
    "sunlight_exposure": "High", "wind_speed": 2.5,
    "co2_concentration": 400, "organic_matter": 5,
    "irrigation_frequency": 3, "crop_density": 100,
    "pest_pressure": 2, "fertilizer_usage": 50,
    "growth_stage": "Vegetative", "urban_area_proximity": 5,
    "water_source_type": "Rainwater", "frost_risk": "Low",
    "water_usage_efficiency": 0.8
  }'

Backend (Java/Jersey)

Application server providing REST endpoints and business logic.

• Calls ML service for predictions
• Manages user accounts
• Stores prediction history
• Handles database operations

Frontend (React)

User interface for crop recommendation.

• Input form for agronomic features
• Prediction results display
• Prediction history
• User authentication

Development

ML Service Development

cd ml-service
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python app.py

Backend Development

cd backend
mvn clean install
mvn jetty:run

Frontend Development

cd frontend
npm install
npm start

Configuration

Each service has a .env.example file. Copy and configure:

# ML Service
cp ml-service/.env.example ml-service/.env

# Backend
cp backend/.env.example backend/.env

# Frontend
cp frontend/.env.example frontend/.env

Model Details

• Algorithm: Logistic Regression with class balancing
• Test Accuracy: 95.91%
• Macro F1 Score: 0.9585
• Cross-Validation (Stratified 5-fold): 0.9569 ± 0.0072
• Features: 23 agronomic features (numeric and categorical)
• Output Classes: 22 crop types

Documentation

• docs/SETUP.md - Detailed environment setup
• docs/API.md - Complete API documentation
• docs/ARCHITECTURE.md - System architecture and data flow

License

MIT License

Contact

For issues or questions, please open an issue in the repository.