Arun-Chettri-1702/TimeGAN-Anomaly-Detection ? reverse-engineered prompt

Reverse engineered prompt

Build me a notebook that does anomaly detection on the machine data in this folder using TimeGAN.

I want to load the train, test, and label text files, clean and prepare the time series, train a TimeGAN model on the normal training data, then use the generated sequences or reconstruction style scoring to find anomalies in the test set. Please include the full flow from data loading to preprocessing, training, scoring, and a simple evaluation against the provided labels, with clear charts so I can see where the anomalies are.

Keep it easy to run from top to bottom in one notebook, and use the existing files in the repo if they are useful. If anything about the dataset or current library versions needs checking, look up the current docs online and make sensible choices. I also want the results explained in plain language so someone new to anomaly detection can follow what is happening.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab