esports_data_analysis

This is the final project for ECE 143, Winter 2021 taken at UCSD. Our group, Group 26, consists of 4 members as follows alphabetically:-

• Anirudh Swaminathan : Github @Anirudh-Swaminathan : Email aswamina [at] ucsd [dot] edu
• David Liau : Github @GenjideGaulle : Email dilau [at] ucsd [dot] edu
• Hejin Liu : Github @Hejin-Bill : Email hel025 [at] eng [dot] ucsd [dot] edu
• Zhifeng Li : Github @Hallizz : Email zhl478 [at] ucsd [dot] edu

Dependencies

Install dependencies

pip install -r requirements.txt

Packages were used:

• pandas
• numpy
• seaborn
• matplotlib pyplot
• sklearn

Folder Organization

• images : contains images generated, or used by our code. Also contains a mp4 file generated by clustering visualization
• notebooks : contains all Jupyter Notebooks + the Final_notebook.ipynb which mirrors all the visualizations provided in the presentation
• scratch : contains python source codes that we used to figure the dataset out and experiment on
• src : the final python source files that are used for generating visualizations for the presentation

Usage

All the data we used can be found in the following link:https://www.kaggle.com/karangadiya/fifa19

We stored our dataset outside the GitHub repository to avoid pushing large objects frequently. Feel free to download the data.csv file and change the directory to access it appropriately.

Visualizations

All the code and clustering can be found in the notebooks file, and the images we generated are stored in the images file.

Clustering

We clustered all of the FIFA teams into 3 groups based off of the following fields: player salary, player age and player overall rating. These three groups segment clubs into the following categories: Championship-Contenders, Middle-of-the-Pack, and Developmental-Rosters. A video of our cluster visualization can be found within the "images" folder, or the interactive plot can be generated in "src/plot.py".