🌧️ Rainfall Prediction System

A production-ready machine learning application built using Python to predict rainfall based on historical weather data sourced from Kaggle. The project follows industry best practices such as environment-based configuration, modular code design, automated dataset handling, and an interactive Streamlit web interface.

---

📌 Overview

Accurate rainfall prediction plays a critical role in agriculture, water resource management, and disaster preparedness. This project applies machine learning techniques to analyze historical rainfall data and generate predictions in a reproducible and scalable manner.

---

🚀 Key Features

• Machine learning–based rainfall prediction
• Automated dataset download using Kaggle API
• Secure configuration through environment variables
• End-to-end ML pipeline (data preprocessing → training → evaluation)
• Interactive web interface built with Streamlit
• Production-ready, ATS- and interview-friendly implementation

---

🛠️ Technology Stack

• Python
• Pandas, NumPy – Data processing
• scikit-learn – Machine learning
• Kaggle API – Dataset integration
• Streamlit – Web application
• python-dotenv – Environment management

---

📊 Dataset

• Source: Kaggle
• Contains historical rainfall data used for training and evaluation
• Dataset is automatically downloaded and cached locally during runtime

---

🔐 Environment Configuration

1. Create an environment file:

   cp .env.example .env

2. Add your Kaggle credentials to .env:

   KAGGLE_USERNAME=your_kaggle_username
   KAGGLE_KEY=your_kaggle_api_key
   KAGGLE_DATASET=dataset-owner/dataset-name

---

▶️ How to Run the Project

Step 1: Clone the repository

git clone https://github.com/Dev-Dy/Rainfall-prediction
cd Rainfall-prediction

Step 2: Create and activate a virtual environment (recommended)

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

Step 3: Install dependencies

pip install -r requirements.txt

Step 4: Run the machine learning pipeline

python main.py

This will:

• Download the dataset from Kaggle (if not already present)
• Preprocess the data
• Train the machine learning model
• Display evaluation metrics

Step 5: Run the Streamlit web application

streamlit run streamlit_app.py
---

## 🧠 Machine Learning Workflow

* Dataset retrieval from Kaggle
* Data cleaning and preprocessing
* Feature encoding
* Model training using Random Forest
* Model evaluation with standard metrics
* Real-time prediction through Streamlit UI

---

## 📜 License

This project is open-source and intended for educational and demonstration purposes.