Derived well tables, analysis code and figures for: Group-identifier leakage in unconventional production models: ordinary covariates identify the well pad

Data and code archive for the article "Group-identifier leakage in unconventional production models: ordinary covariates identify the well pad". Derived state tables with well-pad labels attached for 29,670 horizontal wells in four United States shale plays (New Mexico, Ohio, Colorado, Pennsylvania), twenty result CSVs, the analysis code, and the five figures as submitted (1000 dpi PNG, with a vector PDF of each). run_all.sh rebuilds every result table and all five figures from the deposited tables alone, about two hours on one core. DATA_DICTIONARY.md documents every column of every result file; verify_manifest.py checks the archive against MANIFEST.sha256. The ~3 GB of raw regulator downloads the state tables are built from are not included; SOURCES.md records the origin, retrieval date and SHA-256 of each. Licensed CC BY 4.0.