🌦️ SkyNotifAI

This project is a SkyNotifAI built with Flask, TensorFlow, and Scikit-Learn. It allows you to predict the average, minimum, and maximum temperatures for a future date using a pre-trained machine learning model. Additionally, it provides historical temperature data leading up to the specified prediction date.

📂 Project Structure

project_directory/
│
├── app.py               # 🌐 Flask API server
├── README.md            # 📃 Project documentation
├── src/
│   ├── data_preprocessing.py  # 🧹 Data preprocessing utilities
│   ├── model_training.py      # 🏋️ Model building and training
│   ├── model_evaluation.py    # 🧪 Model evaluation functions
│   ├── predict_temperature.py # 🔮 Prediction utilities
│
├── model/
│   └── weather_lstm_model.h5  # 🗂️ Trained LSTM model file
│
└── data/
    └── export2016_2024.xlsx   # 📊 Historical weather data

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/Yasser5711/SkyNotifAI.git
cd SkyNotifAI

2. Install Dependencies

Ensure you have Python 3.7+ and install the required packages:

pip install -r requirements.txt

3. Run the Flask API

Start the Flask server with the following command:

python app.py

The API will be running at http://127.0.0.1:5000.

📊 Usage

1. Predict Future Temperatures

You can make a GET request to the /predict endpoint with a date parameter in YYYY-MM-DD format. This will return the predicted average, minimum, and maximum temperatures for the specified date along with historical data.

Example Request with curl:

curl "http://127.0.0.1:5000/predict?date=2024-08-30"

Example Response:

{
    "date": "2024-12-01",
    "predicted_avg_temp": 15.67,
    "predicted_min_temp": 10.23,
    "predicted_max_temp": 20.56,
    "historical_data": [
        {"date": "2024-08-20", "tavg": 18.3, "tmin": 12.1, "tmax": 24.5},
        {"date": "2024-08-21", "tavg": 17.9, "tmin": 11.8, "tmax": 23.4},
        ...
    ]
}

2. Making Requests with Python

You can also make requests using Python's requests library:

import requests
response = requests.get("http://127.0.0.1:5000/predict?date=2024-08-30")
print(response.json())

🛠️ Customization

Adjusting the Model

The model is built and trained using the code in src/model_training.py. You can adjust hyperparameters like sequence length, learning rate, number of epochs, etc., to experiment with the model's performance.

Modifying Data Preprocessing

If you need to preprocess data differently, you can modify src/data_preprocessing.py. This file handles tasks like loading data, imputing missing values, scaling features, and creating time-based features.

🔍 Project Details

Model Architecture

The model used in this project is a Bidirectional LSTM combined with Conv1D layers to capture sequential patterns in the weather data. The architecture is designed to predict the average, minimum, and maximum temperatures for a future date based on historical weather data.

Data Preprocessing

Data preprocessing involves:

• Handling Missing Data: Imputation using KNN.
• Feature Scaling: Scaling features to a 0-1 range using MinMaxScaler.
• Feature Engineering: Adding seasonal features (e.g., month, day of the year).

Prediction Logic

The prediction logic can be found in src/predict_temperature.py, where the model predicts the future temperature based on a sequence of historical data points.

Instructions to Use:

1. Install Dependencies: Run pip install -r requirements.txt to install the necessary Python packages.
2. Run the API: Use python app.py to start the Flask API server.
3. Make Requests: Use tools like curl, Postman, or Python's requests library to interact with the API.

Additional Tips:

• Experiment: Try adjusting model parameters in src/model_training.py to see how it impacts predictions.
• Error Handling: The API already has basic error handling for missing dates or invalid inputs, but consider adding more detailed validation depending on your use case.