Related issues in the result output section

Hello,

I hope this message finds you well. I wanted to start by expressing my admiration for your outstanding work and thanking you for open-sourcing your project. While using your code, I encountered a question that I hope you could kindly assist me with.

In your paper's "Abstract" section, you mentioned: "More challengingly, how can we do this in a way that captures complex inter-sensor relationships, and detects and explains anomalies which deviate from these relationships?" After going through some parts of your project's source code, I'm curious about how I could adjust the result output section of the code to obtain multiple sets of sensor IDs that might exhibit abnormal relationships.

I'm looking forward to your response. Thank you very much for your time and consideration.

Hi, if you want to use custom data, you could adjust your input data format similar to the demo data in https://github.com/d-ailin/GDN/tree/main/data/msl. The sensor IDs might require you extract and process from the intermediate results, such as

GDN/main.py

Line 158 in 9853899

test_scores, normal_scores = get_full_err_scores(test_result, val_result)

.

Hi, if you want to use custom data, you could adjust your input data format similar to the demo data in https://github.com/d-ailin/GDN/tree/main/data/msl. The sensor IDs might require you extract and process from the intermediate results, such as

GDN/main.py

Line 158 in 9853899

test_scores, normal_scores = get_full_err_scores(test_result, val_result)

.

Thank you very much for your answer. I will try it first. Thank you again.