Skip to content

Latest commit

 

History

History
31 lines (25 loc) · 2.07 KB

README.md

File metadata and controls

31 lines (25 loc) · 2.07 KB

big-man-betweenness

This repository contains all working material pertinent to the Big Man Betweenness notebook, which was submitted to the 2023 NFL Big Data Bowl competition.

Reproducibility

For those seeking to reproduce our work, we recommend the following:

  • Install the specific version of R and required packages outlined in requirements.txt
  • Clone a remote copy of the repository to your local machine and navigate to the root directory:
git clone https://github.com/bscod27/big-man-betweenness.git
cd big-man-betweenness.git
  • Download, unzip, and store the competition files in the data folder
  • Execute the scripts detailed below

Scripts

The following scripts were used in this project and executed in chronological order:

  1. batch_build.R - takes the raw data, constructs frame-by-frame networks for each play, calculates network measures of interest, and creates weekly "builds" for downstream analysis
    • Note: this script has been programmed to be executed via the command line and fed a positional argument specifying the week/batch of interest; this was done to enable parallel computation, as the program is computationally expensive
  2. analysis.R - wrangles the builds into analysis-friendly form and produces the results
  3. gifs.R - creates the igraph network animations
  4. sample_space.R - generates sample_space.png, which is shown in the report

We use relative path notation in our scripts, as we assume the user sets their working directory to the root of the current folder structure.

Folders

Data, visuals, and other pertinent material are found inside the following folders:

  • data - placeholder folder for the competition data, wrangled datasets, and images used to construct animations
  • images - contains all static images
  • gifs - contains all animations; also contains Python documentation for the supplementary plotly animation