psalmschoi/Rapid-pulse-diagnosis-and-prognosis — reverse-engineered prompt

I want this repo turned into a clean, reproducible battery diagnosis and prognosis workflow that I can run without having to figure out the paper from scratch.

Please wire it up so I can take the dataset from Mendeley, prepare the data, extract the pulse based signals, DCIR, and dVdQ features, calculate PIOV, run the clustering step, and then train and compare the capacity prediction notebooks. I also want the plotting scripts to work so I can quickly see single cell pulse behavior and capacity retention across cells. If the deep learning notebooks need SOC_Point_Data.csv in a specific place, make that clear and set things up so it just works.

A simple end to end run path would be great, plus a short README update that explains what to run first, what outputs get created, and how to reproduce the main results from the paper. If something is missing or outdated, look up current docs online if you need to and make reasonable fixes.