This tutorial is a simple application that used weather data we've collected early.
In this tutorial, we used Recurrent Neural Networks (RNNs) to build time series forecast model. This is covered in three parts:
Part1: Forecast weather temperature of one city. | Run in Google Colab | View source on GitHub |
Part2: Forecast weather info (barometer, humidity, temperature, visibility, wind) for any Saudi city. | Run in Google Colab | View source on GitHub |
Part3: Forecast weather temperature for multiple Saudi cities on a certain date. | Run in Google Colab | View source on GitHub |
This tutorial uses Saudi weather dataset which is an hourly weather data from 2017 to 2019
for all kingdom main cities:
- Qassim
- Hail
- Madina
- EP
- Riyadh
- Mecca
- Tabuk
- Assir
- Northern boarder
- Jazan
- Najran
- Baha
- Jawf
This dataset contains 8 different features such as:
- Date
- Time
- Temperature
- Weather description ( clear - sunny - .... )
- Wind speed
- Humidity
- Barometer (atmospheric pressure)
- Visibility (how much be able to see or be seen)
The requirements.txt file contains all Python dependencies. You can install them by running this command:
pip3 install -r requirements.txt
- D-tale - Library to provide an easy way to view & analyze Pandas data structures.
- Tensorflow - Open source library to develop and train ML models.