Skip to content

Latest commit

 

History

History
53 lines (34 loc) · 2.56 KB

File metadata and controls

53 lines (34 loc) · 2.56 KB

NRCD resource paper & analysis

CIKM resource paper and reproducible analysis scripts for the National Running Club Database (NRCD).

The standalone **nrcd** Python package — PyPI nrcd GitHub nrcd — is for athletes/researchers standardizing their own data. Paper analysis uses scripts/ and does not require running package tests.

Data

Download CSV exports from Zenodo: Zenodo 22077479

Place files in data/ (see data/README.md).

Analysis (reproduce paper statistics)

pip install -r requirements.txt
python scripts/run_all.py --paper-only   # main paper numbers

Key scripts:

Script Purpose
scripts/run_all.py Orchestrates validations and writes results/dataset_stats.json
scripts/load_data.py Merge approved CSV tables
scripts/standardization.py XC standardization pipeline (paper formulas)
scripts/validation_*.py Formula validation experiments

Optional API backfill (scripts/enrich_api.py) requires the separate [nrcd](https://pypi.org/project/nrcd/) package: pip install nrcd[apis].

Citation

arXiv

NRCD: An Open Database of Collegiate Running with Unified Performance Standardization
Jonathan A. Karr Jr, Ryan M. Fryer, Ben Darden, Nicholas Pell, Kayla Ambrose, Evan Hall, Ramzi K. Bualuan, and Nitesh V. Chawla.
arXiv:2608.14776, 2026. https://arxiv.org/abs/2608.14776

@article{karr2026nrcd,
  title={NRCD: An Open Database of Collegiate Running with Unified Performance Standardization},
  author={Karr Jr, Jonathan A and Fryer, Ryan M and Darden, Ben and Pell, Nicholas and Ambrose, Kayla and Hall, Evan and Bualuan, Ramzi K and Chawla, Nitesh V},
  year={2026}
}

Dataset: Zenodo 22077479


Repository by ORCID Jonathan Karr Email, with the help of Cursor.