Data Science Hiring Screener

This folder contains a polished candidate response and the code used to reproduce every numerical result.

Files

• output/pdf/red_sox_data_science_hiring_screener.pdf - final seven-page submission
• analysis/run_analysis.py - statistical analysis for Questions 1-4
• analysis/build_submission.py - PDF builder, including both diagnostic charts
• data/mlb_team_runs_per_game_2025.csv - supplied team-run data
• data/mpg.csv - seaborn automobile dataset referenced by the prompt

Reproduce

From this directory, run:

python3 analysis/run_analysis.py
python3 analysis/build_submission.py

The analysis requires Python 3.10+, NumPy, and Pandas. PDF generation also requires ReportLab.

The analysis uses a fixed random seed (20260728) for the game-cluster bootstrap and 10-fold cross-validation split.

External sources

• Automobile data: https://raw.githubusercontent.com/mwaskom/seaborn-data/refs/heads/master/mpg.csv
• FanGraphs article: https://blogs.fangraphs.com/faster-fastballs-produce-worse-swings/