Romeo-5/Nuclear-Power-Plant-Accident-Diagnosis-and-Root-Cause-Analysis ? reverse-engineered prompt
Reverse engineered prompt
Build me a Python project that can take nuclear power plant sensor telemetry and help spot accidents early, then explain what might be causing them.
I want it to work with the NPPAD dataset, so it should download and preprocess the data, train a few models for anomaly detection on normal only data, and also train a multi class accident classifier for the 18 operating conditions. Please include a way to run the main experiments from simple command line scripts, save configs in a clean way, and produce evaluation results like AUROC, F1, accuracy, and confusion style summaries.
It would also be great if the accident diagnosis part can show which sensors matter most for each accident, using something like gradient based attribution or SHAP style explanations. If possible, include a physics informed digital twin model too, so the predictions stay closer to reactor physics instead of just fitting the data. Make the project easy to run, with notebooks, tests, and clear setup instructions. If you need to check current library docs online, go ahead.
Are you gonna build this?
make sure you review the code using coderabbit