This project is used a personal self-study into several areas such as data collection & wrangling, feature engineering, data analysis & modeling and visualization. At surface level, it is used as a broad view of in-form players currently in the Premier League to inform an ideal Fantasy Premier League (FPL) team to maximize performance/points output. FPL is a game which assigns points to a player based on their performance output in a given game week, and so, being able to assess performance over time either by modeling or by assessment of surface statistics is vital to maximize returns.
This project is organized into several directories, each serving a specific purpose:
fpl-analytics/
├── notebooks/ # Jupyter notebooks for interactive exploration
│ └── visualize_metrics_fpl.ipynb # Main notebook for testing and visualization of FPL data
│ └── visualize_metrics_understat.ipynb # Main notebook for testing and visualization of Understat darta + FPL
├── src/ # Destination for source code of various modules
│ └── config/ # Initialization to cater to personal user
│ │ └── config.py # Settings regarding color schemes
│ │ └── data_template.json # Settings regarding personal FPL information to extract from (Need to rename as data.json once cloned)
│ └── functions/ # Modules and associated functions by which API data is extracted, consolidated and analyzed
│ │ └── raw_data_fetcher.py # Module for fetching raw data from APIs and assigning to variables
│ │ └── data_builder.py # Module for building relations ahd consolidations of and between raw data
│ │ └── generated_helper_fns.py # Module for extracting helper functions from newly fetched/constructed datasets, for use later on
│ │ └── data_analysis.py # Module for interpreting and transforming data for actionable insights
│ │ └── data_visualization.py # Module for any and all data visualization
│ │ └── data_exporter.py # Module for exporting data to proper formats
│ │ └── helper_utils.py # Standalone module for general helper utility functions
│ │ └── notebook_utils.py # Standalone module for general notebook utility functions
├── tests/ # Sample unit testing modules
├── README.md # Project overview and documentation
└── requirements.txt # List of dependencies required to run the project
This repo primarily works by processing data from various APIs and consolidating it through various modules and visualized via jupyter notebooks. General setup is as follows:
- Clone this repository to your local machine.
- Within src/config, rename data_template.json -> data.json.
- Make the following adjustments to data.json:
- "beacon_team_ids" - a custom feature to include rival IDs you would like to measure your team against, typically those with demonstrated success.
- "personal_team_id: - entry to include your personal FPL team ID (use link here to see how to find your FPL ID).
- "personal_league_ids" - used to track statistics and ranks within leagues your FPL team belongs to (or whichever is entered within "personal_team_id").
- Navigate to notebooks folder and open "visualize_metrics" within jupyter environment. Visualization and consolidation of FPL details are implemented through here.
- To execute, simply hit run all.
Add badges from somewhere like: shields.io
