Data and code for Residuals, rankings and outcome policies in paired closed-loop simulation: evidence from Tennessee Eastman and a quadruple-tank test

Start with README.md. The single cce_data_CCE1.zip archive contains the complete baseline and the amplitude and noise extensions; no files from a previous version are needed for assembly. Fixed-input reproduction package for “Residuals, rankings and outcome policies in paired closed-loop simulation: evidence from Tennessee Eastman and a quadruple-tank test”. Version CCE1 preserves the RC11 baseline and adds three quadruple-tank intervention amplitudes (0.25, 0.5 and 1.0 cm) under two numerical solvers, and 120 Tennessee Eastman sampled histories from three severities, ten paired measurement-noise seed groups and four profiles. The archive includes exact sampled projections, pairing, event and completion metadata, frozen analysis specifications, metric code, complete result tables, cutoff margins, conditional intervals, three-state outcomes and execution records. Shared comparators, repeated histories, multiple windows and solver refinements are not independent validation samples. Packaging hash checks preserve the scientific inputs; historical execution records retain their actual dates. Admission-model fitting, thermal raw trajectories and full external Tennessee Eastman dynamic validation remain outside the raw reproduction scope. Third-party simulator code, binaries, literature PDFs and dependency distributions are excluded. Licenses apply by file: MIT for study-authored code and CC BY 4.0 for study-generated data and documentation. See LICENSE.md and release/license_file_map_CCE1.csv inside the archive. Third-party notices, quoted terms and attributed reference facts retain their existing rights.