AgroNomics - Crop Price Predictor 🌾

Empowering Farmers with Data-Driven Insights

A web-based machine learning application that predicts crop prices to help farmers make informed selling decisions and reduce dependency on middlemen.

🎯 Problem Statement

• Farmers suffer significant losses due to unpredictable crop prices
• Lack of forecasting tools tailored to local agriculture
• Decisions based on guesswork instead of data-driven insights
• Heavy dependency on middlemen for price information

💡 Solution

AgroNomics provides a comprehensive web-based ML tool that:

• Forecasts crop prices for upcoming periods
• Uses historical datasets and regression models
• Offers a simple, intuitive interface accessible to farmers
• Covers multiple crop categories across Indian states

✨ Key Features

🔮 Price Prediction

• Predicts crop prices based on location, season, and commodity type
• Covers 9 major crop categories with 200+ commodities
• State and district-wise predictions across India

📊 Comprehensive Database

• Grains: Wheat, Rice, Maize, Millets, Barley, etc.
• Pulses: Arhar Dal, Masur Dal, Green Peas, etc.
• Vegetables: Onion, Potato, Tomato, Cabbage, etc.
• Fruits: Banana, Mango, Apple, Orange, etc.
• Spices: Turmeric, Ginger, Garlic, Black Pepper, etc.
• Cash Crops: Cotton, Sugarcane, Coffee, Tobacco, etc.

🌍 Geographic Coverage

• All 28 Indian states and 8 Union Territories
• District-level granularity for precise predictions
• 700+ districts covered

🌦️ Seasonal Intelligence

• Summer (Zaid): March-June
• Monsoon (Kharif): June-September
• Post-Monsoon: October-November
• Winter (Rabi): December-March

Scroll down to see web interface and model details

🛠️ Tech Stack

Technology	Purpose
Python 3.13	Core logic & scripting
Pandas, NumPy	Data processing & computation
scikit-learn	ML models (Random Forest, DT)
Flask	Web backend & REST API
Joblib	Model serialization
HTML/CSS/JS	Frontend interface

• Dataset: https://www.kaggle.com/datasets/syedjaferk/agriculture-commodity-data-2019

---

🚀 Installation & Setup

Prerequisites:

• Python 3.11+
• pip (Python package installer)

1. Clone the repository

   git clone https://github.com/yourusername/AgroNomics.git
   cd AgroNomics

2. Create a Virtual Environment

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

3. Install Dependencies

   pip install -r requirements.txt

4. Add Model Files Run Agronomics.ipynb Place your model.pkl and encoder.pkl files inside the project directory. (Ensure these files are present for predictions to work.)
5. Run the Flask App python app.py Your app will be available at: http://127.0.0.1:5000

---

📁 Project Structure

    .
    └── AgroNomics/
        ├── assets/
        │   ├── agridata.csv    (original dataset from Kaggle)
        │   ├── cleaned_agridata.csv   (cleaned dataset for model building)
        │   └── crops.docx
        ├── static     (frontend styling & logic)/
        │   ├── css/
        │   │   └── style.css
        │   └── js/
        │       └── script.js
        ├── templates   (frontend structure)/
        │   ├── about.html
        │   ├── copyright.html
        │   ├── home.html
        │   └── result.html
        ├── requirements.txt
        ├── AgroNomics.ipynb  (final model code)
        ├── model.plk      (final model)
        ├── encoder.plk   (encoding for model)
        ├── app.py         (Flask API)
        ├── LICENSE
        └── README.md

---

🏗️ System Architecture

User Input → Flask Backend → ML Model → Price Prediction → Web Interface

• User Input: State, District, Crop Category, Crop, Month, Season.
• Backend (Flask): Processes input and calls the ML model.
• ML Model (Random Forest): Predicts the crop price using trained data.
• Price Prediction: Returns modal price per kilogram.
• Web Interface: Displays the results in a clean and user-friendly manner. Print Result.

🎮 Usage

1. Open the homepage.

2. Select State, District, Crop Category, Crop, Month, and Season.

3. Click "Predict Price" to view the estimated price (₹/kg).

   

🧠 Model Details

• Dataset: https://www.kaggle.com/datasets/syedjaferk/agriculture-commodity-data-2019
• Algorithm: Random Forest Regression with fallback prediction system
• Features: Category, Crop, State, District, Month, Season
• Encoding: Ordinal Encoding for categorical variables
• Fallback System: Rule-based predictions when ML models unavailable
• Accuracy: ~88% on cleaned dataset (~40k entries)

📊 Model Performance

Metric	Value
MAE	3.25
RMSE	5.64
R²	0.88

Top Features (by importance): Crop > Category > State > District > Month > Season

• 

---

🔧 Troubleshooting

• Model Loading Error: Ensure model.pkl and encoder.pkl are in the project directory.
• Missing Templates: All HTML files must be in the templates/ folder.
• Port in Use: Stop other processes or use python app.py --port=5001.

---

Name	Role	Contact
Sangini Garg	Developer	GitHub