Earth Quakes Prediction Using Machine Learning Algorithms

System(s) Requirements

To set up this project on your computer, you will need to have the following application and libraries installed on your computer

• Docker (Docker Desktop)
• Python3
• Keras
• Flask
• Numpy
• Pandas
• Jupyter
• Sklearn
• Folium

Installation & Setup

To install, you need to clone this repository by copying the repository https url. You can do this by clicking on the Code button displayed on this repo menu. The repository link should look similar or be exactly this https://github.com/ogunsakin01/earth-quakes-analysis-and-prediction.git. Then open the terminal interface on your machine and navigate to the desired directory you wish to set up the project. After that, then run this command below to clone the repository

git clone https://github.com/ogunsakin01/earth-quakes-analysis-and-prediction.git

On successful completion of this command, a folder names "earth-quakes-analysis-and-prediction" will be created in the directory location with the source code of this research, then you can run the command below to switch to the root directory of the project in your terminal

cd earth-quakes-analysis-and-prediction

Running without Docker

Before doing anything else, ensure python and its associated terminal helper command pip is installed and working on your computer. You can confirm that by running this command in the terminal

python3 --version && pip3 --version

or run the command below to confirm another version of python and pip is installed

python --verson && pip --version

If Python is installed and working, then you can proceed to the next step by installing the required libraries and packages using in this research experiment. You can do that by running the command below

pip3 install --trusted-host pypi.python.org -r requirements.txt

This command will install the following libraries and packages to your computer flask, numpy, pandas, jupyter, scikit-learn, keras, and folium

Accessing the Jupyter notebook

To access the main experiment, run the command below in your terminal while still in the root folder of the project.

jupyter notebook experiment.ipynb

This will start up the jupyter notebook server and auto redirect to a browser tab with the experiment open and ready to use. If that does not happen, copy and paste this url into your browser http://localhost:8888/notebooks/experiment.ipynb

Running with Docker (API)

This project also provides a second setup that opens up the experiment, and it's output over different endpoints To successfully set up and run this project, you need to have installed Docker desktop on your computer. After installing docker desktop, you need to run this command to build your docker container

docker build -t my-api .

then run the command below to start the API

docker run -p 80:80 my-api

or run both commands together

docker build -t my-api . && docker run -p 80:80 my-api

This will start the Docker container and map port 80 on the container to port 80 on your local machine. Ensure no other process in using port 80 on your computer before running these commands, if you do, ensure to disable the processes to keep the prot free.

Accessing Analysis & Predictions Endpoints

Visit http://localhost in your web browser to confirm everything works fine.

• http://127.0.0.1/get-clean-tectonic-plates-data - Gets cleaned tectonic plates dataset
• http://127.0.0.1/visualise-earthquake-on-tectonic-plate - Displays visualisation of earthquake on tectonic plate
• http://127.0.0.1/get-prediction/2024-12-10 -Get prediction using Random Forest model
• http://127.0.0.1/get-lstm-prediction/2024-12-10 - Get prediction using LSTM model
• http://127.0.0.1/visualise-prediction/2024-12-10 - Visualise prediction from Random Forest model
• http://127.0.0.1/visualise-lstm-prediction/2024-12-10 - Visualise prediction from LSTM model
• http://127.0.0.1/get-clean-earthquake-data - Get cleaned earthquake dataset

Limitation & Challenge of Current Docker Setup

You have to run docker build -t my-api . and docker run -p 80:80 my-api everytime you make updates to the files to view your new updates