Code, field dictionary, de-identified snapshot and external validation for the audit of the International Judo Federation public results database
收藏资源简介:
Replication material for the manuscript "When the field does not contain what its name promises: An audit of completeness and recording conventions in a public database of 94,811 international judo contests" (Cantão, E. W.; preprint: https://doi.org/10.51224/SportRxiv.1091). Contents: (1) ijf_judo_audit_code_and_outputs.zip: retrieval code (00), audit and sensitivity code in R and Python (10, 14, 15), figure code (16), dictionary and snapshot builder (17), external validation code (18, 19), the field dictionary and the CSV outputs that feed the tables of the manuscript (the external validation outputs carry no athlete names); (2) ijf_judo_audit_snapshot_deidentified.zip: the contests table (94,811 rows) and the actions table (166,292 rows) as retrieved from the public interface of the International Judo Federation in August 2026, with athlete identifiers replaced by salted hashes; (3) README.md with the instructions to run everything. The International Judo Federation publishes no reuse licence for its records; the snapshot is shared for replication of this audit only, the source remains publicly accessible at https://judobase.ijf.org, and the source is dynamic and may differ in a new retrieval. The official Olympic judo results books used in the external validation (Rio 2016, Tokyo 2020, Paris 2024) are public documents of the federation (https://www.ijf.org/documents) and are not redistributed here. The code and the field dictionary are released under CC BY 4.0.



