Code, design records and run-level results for "Will the soft sensor still work after the control loop is changed? Screening validated predictors for vulnerability to planned control policy changes"

Python code, dated design records (with SHA-256 checksums) and run-level results supporting a manuscript submitted to Journal of Process Control. The study screens validated soft-sensor predictors for loss of gain after planned control policy changes in three simulated processes: the Tennessee Eastman process, the IndPenSim dataset and Benchmark Simulation Model no. 1. See README.md for the folder structure and the order in which to run the scripts.